ChatGPT
approach to Artificial General Intelligence AGI
By Kristo Ivanov, prof.em., Umeå University
July 2023 (rev. 240428-2135)
<https://ia802703.us.archive.org/13/items/chat-gpt-agi/ChatGPT-AGI.html>
<https://www8.informatik.umu.se/~kivanov/ChatGPT-AGI.html>
CONTENTS
1. Link
to a general disclaimer
3. Initial overall considerations
AI in
The Design of Inquiring Systems
On
Statistics and Experimental Inference
Apparent
Obsolescence and Desperate Logic
Driving
Forces behind Technology
Ego
Inflation or Philosophy behind “Powerful Experiences”
David
Noble and “The religion of technology”
Autism,
or “Mathematical computer-oriented Minds”?
Artificially
intelligent artistic production?
Return
to The Design of Inquiring Systems
6. References to ChatGPT in other own essays
From:
Information and Theology
From: Computerization
as Design of Logic Acrobatics
From: Computers as Embodied Mathematics and Logic
From:
The Russia-NATO-Ukraine Information
Crisis
7. The open letter: "Pause Giant AI
Experiments"
8. Comment to the open letter (proposal for moratorium)
9. Case
study: “Creating safe AGI that benefits all of humanity”
10. Conclusion: the meaning of hype
11. Conclusion: beyond the hype
12. Concluding humor
2. Historical background
The Wikipedia
report on artificial intelligence - AI in July 2023,
seen as an introduction to artificial
general intelligence - AGI, in the last paragraph of its introductory section
observes that the term artificial intelligence has been criticized for
overhyping AI’s true technological capabilities. The present text is intended
to specify a few details of my criticism. The second paragraph of the very same
introductory section writes about AI’s “several waves of optimism”:
Artificial intelligence was founded as an academic
discipline in 1956, and in the years since it has experienced several waves of
optimism, followed by
disappointment and the loss of funding (known as an "AI winter"), followed
by new approaches, success, and renewed funding. AI research has tried and discarded
many different approaches, including simulating the brain, modeling human problem solving, formal logic, large databases of knowledge, and imitating animal behavior. In
the first decades of the 21st century, highly mathematical and statistical machine learning has
dominated the field, and this technique has proved highly successful, helping
to solve many challenging problems throughout industry and academia.
I
understand that we are experiencing its last, or unfortunately only the latest,
wave of optimism. I suggest that the cause of these waves of optimism is the
basic misunderstanding of what intelligence and
consequent artificiality are
or should be, allied to the forecasted marketing power and hoped economic
profitability, and for whom, of the hype of AI and now AGI. This is related to
a misunderstanding of the essence of logic and mathematics, and consequently a
misunderstanding of statistics that
is reduced to big
data and to a controversially and problematically
described mathematical
statistics, as I tried to show that they are embodied
in computers. The result suggests an ongoing cultural crisis as a
societal analogy to the scientific scandal of lobotomy, a
cut off of the emotions and feeling, valuational and intuitional dimensions of
thought and psyche of a great deal of the population. It portrays difficulties
that in part were discretely addressed long time ago in articles about The artificiality of science and,
The metaphysics of design: A Simon-Churchman "debate (in Interfaces, vol.
10, No. 2, April 1980).
The
fist mentioned article on the artificiality of Science is Churchman's review of
Herbert Simon's book on The Sciences of the Artificial (1969) that academically gave to the latter a prestige
in the field of artificial intelligence, impressing the academic community of
his being a "genius" in the most disparate fields of knowledge, and
concluding with his being awarded the Nobel prize in Economics. All based on a
misunderstanding of the essence and function of logic and mathematics as
mentioned above.
The
latter article refers to the work of a pioneer of AI and Nobel Memorial Prize
in Economic Sciences, Herbert
Simon, and of prof. West Churchman,
philosopher and systems scientist. It also raises the relevance of Churchman’s
timely article “An analysis of the concept of simulation” (in Symposium on Simulation Models by A.C.
Hoggatt & F.R. Balderston, eds., Cincinnati: South-Western, 1963, pp. 1-13)
as well as its application by Ian I. Mitroff, “Fundamental issues
in the simulation of human behavior” (Management Science, Vol. 15, No. 12, August 1969). The main issue
in also covered in Churchman’s “Real
time systems and public information” (in Proceedings of the AFIPS/ACM 1968 Fall Joint Computer Conference,
pp. 1467-1468, among others in pp.1467-1498). I am sure that very few people
will recognize that the same issue is partly covered in “The researcher and the
manager: A dialectic of implementation” in Management
Science, Vol. 11, No. 4, Feb. 1965 (more on this below). This is relevant
if we conceive the manager as analog to the questioner and user of the answer,
and the researcher as analog to the producer of the answer, based on his
database and elaboration of the data.
It is finally symptomatic that the whole of
discussions going on about AI and AGI apparently ignore completely the violent
hype that started in the (now “obsolete”?) seventies around Hubert Dreyfus’s views on artificial intelligence. There is no interest in What Computers Can’t Do. This will not treated here because they were based on “modern”
continental philosophers such as Merleau-Ponty, Martin Heidegger and
phenomenology that I perceive as arising from a misunderstanding or abuse of
the philosophy of Immanuel Kant, as discussed in my
essay on the ongoing computerization of society. The reason that the earlier hype around
Dreyfus is ignored in the present renewed
hype of AI-AGI is related to why humanity continues with divorces and wars, and
why ChatGPT has not been adduced to help in the solution of the Russia-Ukraine
conflict or the 2023 Israel-Hamas war. More on this below, but before that, it
is necessary to orient readers who do not yet know what the acronym ChatGPT
stands for, orienting them to Wikipedia’s
explanation of the term. Common educated
people’s difficulty to even understand the explanation is part of the “mystery”
that surrounds the hype of artificial intelligence as of most modern
mathematized science and technology, isolating them from democratic control.
3. Initial overall
considerations
In
the above section it can be seen that the referenced texts are dated up to the
year 1980. I I am well aware that they may be considered by some as being
outdated because of the great and rapid technological development in these more
than 40 years. Nevertheless, as I wrote in the latest revision of my general
disclaimer: On
occasion of my retirement I intended to carry out in full scale my research
that had been neglected because of managerial duties at my university, and to
summarize my experiences and reflections in one or several publications. The
more I studied and reflected upon my experiences, the more I got convinced that
the main problem was not the lack of debates, books, publications, experiences,
reflections, but rather the lack of wish, will, courage, that is "lack
of time" to select readings, to read, to think, or to
understand in depth and to act upon knowledge that is still valid and already
available.
This
means that I perceive the present attitude to AGI as revealing a deep cultural
crisis, first of all, in the western world, with certain affinities with the
message of Oswald Spengler’s The Decline
of the West. Just because the western world is or was the most “advanced”
(towards nowhere?). The decline is properly illustrated by the breakdown of
social communication, which in science I exemplified by the reciprocal ignoring
of scientific insights within increasingly narrows fields of science where the
concept of “field” itself is being obliterated and promotes both the narrowing
and consequent speed of perceived advancement. The process of narrowing and
reciprocal ignoring the wholeness of reality also implies a loss of democratic social control of
scientific activities that often are driven by narrow and often secretive
disciplinary, economic and military interests. Scientists themselves who are
active as e.g. physicists in fields that earlier were related as in the general
denomination “physics”, not to even mention the general educated public, do not
understand and cannot evaluate their earlier colleagues work in many highly
specialized “sub-fields”, despite the assumed existence of one “scientific
method”, which bestows to all the prestigious name of science. All this while,
paradoxically, Democracy with capital D as the final “guarantor” has come to be
a substitute of the earlier perception of God as I explain in my essay of the conflict between
Russia and Ukraine.
The
loss of democratic control, which in its essence is seen as "general"
for implying the whole society, includes a loss also in the community of
scientists dealing with both science in general, as well as within each field
and subfield that results from progressive specialization that is misunderstood
as a sign or progress. In his book review mentioned above, Churchman observes
that "the debate about artificial intelligence has been raging ever since
the introduction of that rather unfortunate terminology". He sees Simon's
introduction of the term sciences of the
artificial in one more "benign" sign of his "unremitting
enthusiasm", which is matched by the community of scientists sharing his
attitude of being "strongly positivistic in his thinking". I myself
have met scientists who do not even know that positivism in this context means logical positivism, and
justify themselves by boldly acknowledging that, yes, they are proud for being
positive in their attitude to their work (without even differentiating between
positivistic an positive), and logical in their thinking. Similar improbable
accounts have I heard about AI-enthusiasts who counter the criticism of the
system not accounting for induction by
countering it with the question “what is it?”
Churchman
goes further in explaining why databases with statements in the indicative mood
do not take into account the goals implied in their application to particular
uses related to unforeseen (whose?) goals. And he observes that it is not
perceived that natural science itself is artificial. This artificiality of
science in its gradual mathematization and logification leading to
"Newton's syndrome" (belief in the possibility of going from the
simple to the complex, or "more of the same") is something that was
already affirmed in the far-reaching insights of Jan Brouwer
about the foundations of logic and mathematics, as outlined in my essay on Computers as embodied mathematics and logic. It should be seen as a foreknowledge for understanding
the whole text that follows. I have come to the conclusion, supported by the
text below, that the basic problem of the use (not the design) of AI and AGI is
the lack of understanding of its essence being formal science as represented by
mathematics and logic. The latter’s function and limitations are ignored when
they are further equated to intelligence and
further to a misunderstood human
intelligence, which plunges us into a hodgepodge of philosophical
and psychological controversies that are also ignored and soon forgotten in
face of the possibility of making money and obtain power over nature and other
humans.
Quite
often this is done under the mantle of an appeal to “pragmatism”, but then
there is no reference to what pragmatism is or should be since it would lead to
a discussion of the philosophy
of pragmatism its history and its variants. I may be repeating
myself if I claim that one, if not the basic
problem in understanding AI and the more so in understanding AGI is that most
common, even educated, citizens including scientists do not need to understand
what the foundations of logic related to mathematics and to what a theory is related to the abused word model, and the even more abused term conceptual framework. For instance, in
mathematized science, best exemplified by a physics that is mathematical
physics, mathematics and logic would be meaningless without a supporting theory
and its concepts with closely
associated sensory presence (e.g. force, mass,
speed, acceleration) which have been
painfully developed under hundreds of years. Today in quantum physics (cf. my
problematization of it) new concepts
have been derived mathematically in
order to support empirical observations in complex experiments that have no
direct bodily sensory presence, e.g. “particles” that are not particles but
rather mathematical constructs. On the other hand, AI and AGI are pure logical
and mathematical manipulations that are theoretically supported only for
application in well-defined and limited fields that have a theory, but instead
they are mixed with manipulation of common language and use or abuse of “tools”
of mathematical statistics divorced from controversies about foundations of
statistics. More on this below. Nothing of this kind of knowledge is available
to the common educated citizen who is now going to be drowned in the hype of
advertising on AI, and not even to many scientists who work in well-defined and
theoretically supported fields, sometimes unaware that the concept of field
itself is put in question.
All
this is ignored or forgotten as source of debates about AI, and even less about
AGI in their "waves of optimism" culminating with the ongoing latest
hypes. But it does not detract from the possibility of legitimate profitable
applications, legitimate only when (if) one (who?) knows the consequences of
what one is doing (to oneself and others) in the short and long run, and
whether it is ethical (and what that means with and without religion) besides
of being profitable (in the short and long run, and for whom?).
AI in
The Design of Inquiring Systems
I do
not know a better illustration of the meaning of AI and AGI than in the seminal
work of the earlier mentioned West Churchman with a background in pragmatism in
his book The Design of Inquiring Systems (1971) to which I will usually refer further to with
its initials DIS, and that I extended with a general DIS Word and Subject Index. I remember seeing some year ago, in the “Talk”
section of the Wikipedia article on artificial intelligence
somebody referring to this book. Older Wikipedia “talks” are not archived in
Wikipedia, but I remember that it was swiftly disposed of because the author
was seen as an “outsider”, without Wikipedia’s primary editors of the article
displaying an understanding of what it all was about. This is a less-known
attitude of a censorship by primary groups of editors with unconscious “vested
interests” in a particular subject of Wikipedia, as I illustrate in an article on
Wikipedia-democracy. I resume the reasons for the impossibility of
“talks” or debates about such issues at the end of the conclusions of my
special essay about Information and Debate.
In chapter
4 (p. 79) of the above mentioned DIS book the author presents an elaborated
version of an earlier paper On the Design of Inductive Systems (1969) written in co-authorship with Bruce
Buchanan with material based on a research project conducted
by E.A.
Feigenbaum, Joshua Lederberg,
Bruce Buchanan and others, together with the author, at Stanford University.
Bruce Buchanan, however does not seem to have taken very seriously his
co-authorship with Churchman, since in his later works mentioned below he
clearly joins the logical-empiricist track of Herbert Simon. The latter, in
philosophy and methodology of science, stands on the lucrative opposite side of
West Churchman, as suggested in the only text I know that indirectly and in
rhetorical elegant way spells this opposition, the earlier mentioned paper
by W. Ulrich's The metaphysics of design: A Simon-Churchman "Debate".
What
is happening today is that AGI, or AI in the spirit of ChatGPT, seems to
consist of smart elaborations of the chapter 4 mentioned above, which in turn
is a modernization of what Churchman in a book published so early as in 1948
already had called Theory of Experimental Inference, and Buchanan, without daring to enter into the
problems of philosophy of science, tries to extend from MYCIN (and cf. DENDRAL. They
are considered by Churchman in DIS pp. 89, 93, 98) further in "Artificial
intelligence as an experimental science" in James H. Fetzer (ed.) Aspects of Artificial Intelligence (1988, pp. 195, 207, 209-250,). In fact, much of the
AI hype today is based on a simplification, if not outright exploitation of
misunderstood induction and experimental
inference, thereby ignoring the reason and motivation for Churchman's
progress since 1948 up to 1971 and 1979, from the above-mentioned Theory of Experimental Inference up to DIS and The Systems Approach and its Enemies. It is also an exploitation of
misunderstood logic and mathematics equated to “analytical - abstract thought”,
equated further to reason confounded with “intellect”, whatever it is or should
be, forgetting the subtle hodgepodge of Kantian differentiation between reason
and understanding, pure and practical reason, and their relation to the
synthesis in his Third Critique, of judgment, and aesthetics that today is
reduced to “design”. And the whole is
alternatively reduced to “Intelligence quotient” IQ vs. “Emotional quotient” EQ
(see below) seasoned with the expectation that computers in the future will
take care of IQ, and humans of the EQ, whatever they are or should be. We can
wonder whether capital intensive industry and associated governmental economy
with take care of computers and IQ, leaving EQ to churches and agencies for
social welfare.
On Statistics
and Experimental Inference
The
problems are akin to those relating to the critiques of
"Big Data", and even those ignore the earlier criticism of abuse
of statistics such as Edgar S. Dunn's Social Information Processing and Statistical Systems:
Change and Reform (1974),
with the author honoring me with a positive review of my dissertation on Quality-Control of Information. Swedish readers have the opportunity to read Olle Sjöström’s
work, starting with a PhD dissertation (1980, which as usual in computer
context may be considered as “obsolete”) on Svensk Samhällsstatistik: Etik, Policy och Planering [Swedish Social Statistics: Ethics, Policy and
Planning]. It starts (p. 154) its English summary asking rhetorically:
Has the growth
in the “production” of statistical data, together with the associated
publications, press releases and so forth, led to improved information for
decision makers, better informed citizens and increased knowledge about current
social conditions?
The
matter is developed further in the book (1980, recalling why “history and
critique is not obsolete”) with the title Svensk
Statistikhistoria: En undanskymd Kritisk Tradition [History of Swedish
Statistics. A Hidden Critical Tradition], with an extremely rich multi-language
bibliography (p. 240-245), and an initial English abstract that states:
This essay
delineates a critical tradition of statistics in Europe and Sweden, lost or
ignored during the last decades. European statistics develops from
philosophical empiricism and optimistic endeavours in the 17th century to the
exploration of social conditions (John
Graunt) and to comparisons among European national states (Hermann
Conring). […] This book claims that it is necessary to awake
and renew such a critical statistics tradition in order to benefit from modern
statistical techniques and computer resources in the future.
One
of the consequences of the loss of a critical tradition in dealing inductively
with big data is the reduction of statistics to only mathematical statistics, the loss of the distinction and relation
between production or rather measurement, and consumption or rather use and
implementation of data or rather information and knowledge. Consequently one
uses standard statistical “tools” on “data“ that is fed into the computer
without needing to know the presuppositions for the “consumption of data” by
the automated use of the “tools” (more on “tools” below). Never mind about the
foundations of statistics as discussed, for instance, in West Churchman’s
discussion of “objective probability” in Prediction and Optimal Decision (chap. 6). It is only a question of automated
computerized application of ready-made formulas on available data. One further
consequence and example may be the unquestioned new AI hype of
"biotechnology" in How Artificial Intelligence is Revolutionizing Drug
Discovery (March
20, 2023), raising "hundreds of
millions of dollars to pursue their AI-driven drug discovery
pipelines". Which outsiders and laymen would or could question such
specialist "scientific" affirmations, when it is obvious that
"hundreds of millions of dollars" can always produce discoveries, the
more so when there can be no cost-benefits analysis (taste here the extent of the
difficulties), and forgetting that the relevant concept of cost here is opportunity cost.
Intelligence,
however, is much more than
(misunderstood) experimental inference, and the very same Fetzer keeps
returning cyclically in time to an endless inconclusive struggle about the
ineffable "much more", as in Computers and Cognition: Why Minds are not Machines (2000), and Artificial Intelligence: Its Scope and Limits (2012).
Until Fetzer, like myself, gets older and older, and cannot yet incorporate the
last hype-revival of ChatGPT in a new book. The less so when experimental
inference, statistics and mathematics are melted down, as in Q-learning, in
an apparently prestigious mixture that ignores the meaning of “inquiring
system” and that is impossible to understand by the average well-educated
democratic citizen.
Apparent
Obsolescence and Desperate Logic
Of
course, it is characteristic for the AI field that "older" works from
the 1960ies are in practice ignored or dismissed for not portraying later
modern advances. But not only from the 60ies: also those from the 80ies are
considered to be old. By whom?: by those whose opinions and work today will
therefore also be considered obsolete after a couple of years. This is the more
so when there is no consciousness of the foundations and thereby the meaning of
logic and mathematics, that were discussed up to the beginnings of the past
century as I survey in my Computers and embodied mathematics and logic. Not to mention the basis of mathematical or formal
logic in the work of Gottfried Leibniz (1646-1716).
And in my later essay the consequent problems of computerization or
digitalization of society, represented here by the AI-hype, I have a section of
a "Case
study of desperate logic".
It is
a "desperate logic" which could include William J. Rapaport's section
of the book Aspects of Artificial
Intelligence (p. 81) with the desperate title of "Syntactic
semantics: Foundations of computational natural-language understanding",
that revives my text on "Symptoms:
Syntax. Semantics, Pragmatics". It reminds of older
desperate logic as Sheldon Klein’s Culture,
Mysticism & Social Structure and the Calculation of Behavior (European Conference on AI, ECAI, 1982, pp. 141-142).
It also reminds of the present hype of ChatGPT in terms of its use of large
language models LLM, which may be seen as a hodgepodge of syntax, semantics and
pragmatics that tends to invalidate all these categories, as they would also be
invalidated by the hype-attempt to create a Data Science akin to an Information
Science if one could unravel the meaning and difference between them, in
oblivion of earlier mentioned The Design
of Inquiring Systems - DIS. It is also a "desperate logic" as
suggested in the nowadays mostly dimmed but formerly hyped book Gödel, Escher, Bach: an Eternal Golden Braid (1979). It is a subject (whatever it is) that is
constantly and inconsequentially revived, as lately "mystically" on
the Internet as About the "Uncalculable World" and
"Mizutani Portfolio". My impression is that the latest
modern example of desperate logic is found in the mindblowing undefinable field
(whatever field means) of prompt engineering as a
“process of structuring text that
can be interpreted and understood by a generative AI mode”, mentioned also later in this essay.
I
myself feel tempted to feel pride in having devised as analogy that, despite my
exceptionally numerous links, I do not remember having read or heard anywhere.
It is today’s view of “old obsolete knowledge” of 400, 100, 40 or 10 years ago,
the account of the painstaking work of, say, Leibniz, Galilei and Newton, not
to mention the world’s sacred books and the philosophies of, say, Plato and
Aristoteles, all this, invites to conceive a bold, if yet ultimately and
necessarily imperfect analogy:
|
It is as if those pioneers were like hunters who,
risking their own lives and deepening their skills, had succeeded in
imprisoning and domesticating a big wild beast, or putting it in a cage.
Later come we, and in relative comfort examine its “capabilities”, kill and
dismember its body, eating gradually the flesh, and exploiting its remains,
until arriving at the bones and analyzing their material composition. Until
realizing that the beast is dead, and we know nothing about its origin, the
effects of its life on the environment, or its offspring and its effects:
“Desperate hunting”. |
Or one of the
latest logical-mathematical geniuses who in
"neuroscience", and based on experiments with masses of neurons in
monkeys' brains, finds that
"the human brain would be the center of the universe and tries to explain
the history, the culture and the human civilization based on recently
discovered principles regarding human brain function". Finally, this brain
science or neuroscience is converging
with AI in the latest hype of worldwide European and USA-projects: the USA
National Institute of Health’s Brain
Research Through Advancing Innovative Neurotechnologies (BRAIN, here and here, with the Wikipedia
summary here). It goes from an understanding of the “structure of the mouse
brain” over to “the structure of the human brain” and “Developing and applying
brain-inspired computing technology”. What is mentioned is “brain
imaging data structure (BIDS) conversion” and “neuroimaging data and metadata”, without worries about difference between data,
metadata, information, and system, or between structure and function (DIS,
chap. 3). But guess whether “To date, more than 1,100 awards totaling approximately $2.4 billion have
been made by the BRAIN Initiative”
will not produce some interesting and successful results and applications. In
front of this perspective, it is a small consolation to read the insightful and
benevolent reflections of Abraham Kaplan about what seem to be “brain
mythology” in The Conduct of Inquiry (1964, p. 325, my italics):
Suppose it were
definitely established that the direct cause of schizophrenia is the
concentration of a certain substance in the blood, or the formation of a
certain structure in nerve cells or their connections. The effect of this
cause, so far as it concerns psychiatry, would still be statable only in terms
of macro behavior (for instance, that resulting from delusions of persecution).
Nor would it follow that treatment
necessarily requires the laying on of hands, whether by medicine or surgery.
For the possibility would still be open that the substance or structure itself
is produced or destroyed by various patterns of interpersonal behavior or
communication (as in the familiar case of ulcers, so far as producing then is
concerned, at any rate). Psychophysical dualism always encounters the
problem of interaction between the two realms. What I find as exciting about
such development as cybernetics and related disciplines is precisely that they
have “helped to pull down the wall between the great world of physics and the
ghetto of the mind”. [Kaplan’s ending quotation is from Phillip G. Frank, The Validation of
Scientific Theories, N.Y., 1961/2018, p. 155.]
The
hype of cybernetics in the sixties is today in 2023 the hype of AI and AGI.
Neither Frank nor Kaplan seem to have been able to imagine that instead of
“pulling down the wall” what would happen is that the “world of physics”
incorporates or outright swallows the “ghetto of the mind”, the more so for not
considering a ghetto sufficiently dignified.
The
important and serious thing seems to be the necessity of a serious study,
probably possible only in terms of analytical psychology, of what is happening
nowadays with the mind or psyche of modern man in the shadow of
computerization, digitalization or logification. This to the point of
disrupting the basis of human relationships, as illustrated further below with reference to Reason and Gender. It is then more than fair, however, to remark that
such kind of analytical thinking, typical in formal science such as mathematics
and logic, has been important for the particular development of the West. It
may be seen as rooted in the Judeo-Christian tradition, as noted in
controversial studies of the Jewish Ashkenazi intelligence (cf. here and here) and
its performance in the West (cf. the "archetypal" Albert Einstein).
This happens, however, in parallel to the gradual secularization or loss of its
religious basis and Christian influence. Here is also the place to acknowledge,
of course, the immense importance of the formal way of thinking in science,
technology and industry. It stands behind the western mainly material wellbeing
with the remark, however, that it may have been what Goethe meant with what
came to be called a "Faustian bargain".
An explanation of the meaning of such a Faustian bargain can be sought in Carl
Jung’s insight (Collected Works, Vol. 6, p.77 §113) in
a quotation that I already reproduced in an essay of Information and Theology:
Through
the shifting of interest from the inner to the outer world our knowledge of
nature was increased a thousandfold in comparison with earlier ages, but
knowledge and experience of the inner world were correspondingly reduced. The
religious interest, which ought normally to be the greatest and most decisive
factor, turned away from the inner world, and the great figures of dogma
dwindled to strange and incomprehensible vestiges, a prey to every sort of
criticism. [...] Modern rationalism is a process of sham enlightenment and even
prides itself morally on its iconoclastic tendencies.
So,
Leibniz is not even mentioned in e.g. James Fetzer’s edited book on Aspects of Artificial intelligence, not
even in Clark Glymour's contribution with the title "Artificial
Intelligence is philosophy", despite his writing (p. 205) that
"The trouble with artificial intelligence work is that it can use the
senses to overwhelm the intellect".
I
propose that it is much more than this: the advent of incursion of the latest
if not last AI-Hype is the misunderstanding and abuse of mathematics and logic
in the process of dissociation of the human mind in the process that I present
in my text on the
meaning of human-computer interaction HCI. In this
process can be seen (by those who know it) an oblivion of the historical issue
of Geisteswissenschften or “sciences of
the spirit” related mainly to the name of Wilhelm Dilthey, an
issue which later seems to have reverberated in Rudolf Steiner’s
Anthroposophy. The dissociation of the human mind can rhetorically be seen as
tragic ongoing analogy to the historical scientific and political scandal of Lobotomy,
which also happened to dissociate the brain from its so-called emotional side,
as formal science does with the undefined thought or "cognition" in
the undefined mind that eschews all relation with analytical psychology. The metaphysics of design: A Simon-Churchman “debate”. I guess that many readers will think that it is too
farfetched and controversial to be reminded that a "cultural
lobotomy" also was the background of the Holocaust,
(cf. the identification
of inmates in Nazi concentration camps, which later
became worldwide personal
identification number) based on the "intelligent" superiority of a
"race". It all accompanied by eugenics in several "advanced" western countries,
with controversial
interpretation of the Swedish national intellectual heroes, Nobel
Prize winners Gunnar
and Alva Myrdal's stand on the issue.
It is
a cultural lobotomy that may also stand in the background of the ongoing
equalization of humans and animals, with humans reduced to only animals, while
believing that we are heightening animals up to humans. This is related to the
fact that the undervaluation of humans increases the incapacity of seeing the
difference between human and animals, but also between humans and machines, as
evidenced in the apparently unproblematic frequent reference to the famous
"Turing
Test". It reminds us that the less one understands
what a human being is, the easier it is to equal him to a machine or to
“nature”, including to see as Nobel prize laureate in economics Herbert Simon
did “behaving
human beings as simples as ants”
(more on this below). All this is in turn related also to the philosophy
of veganism as well as of
the heightening of "Mother Nature"
to a new goddess for a pantheistic solution of the issue of pollution and climate warming, up to the point of youngsters
declaring to relinquish having a polluting offspring.
The
step is short to equalize humans and machines, with the supposed final
attainment of machines that are not machines, with own
"consciousness" if not "spirit", whatever it is perceived
to be, and a superhuman if not godly intelligence. But if computers do not
become superhuman, humans can perceive and behave and be slow computers and not
only as animals as in the Holocaust of the second world war. This is also the
rationale for the expectation that not only many manual workers but even
so-called intellectuals and scientists who nowadays already think as slow
computers in the computerized West, will in the future be replaced by advanced
AI/AGI/ChatGPT. The greater the number of people at work, even “intellectual”
work, with tasks that are designed to require to act as only slow computers,
the easiest will be to replace them with slow or faster “AI” computers. Not
only that: it is the same cultural lobotomy, the separation (and not only
differentiation) between logic and psyche, that enables democratic societies
and especially politicians to believe that it is possible to counter
criminality with only law, police and prisons, instead of relying honestly on
"love thy neighbor". Or to believe in
countering credit
card frauds with a progressive build-up of the computer
"security" and behind card security
codes. Or to substitute computer security for quality of
information as defined in my doctoral dissertation on Quality-Control of Information, and originally
suggested in my early paper on the
concept of security that today reappears under the avatar of AI-Safety.
In
lack of an understanding of the human psyche that would require study of the
history of philosophy and e.g. analytical psychology with and understanding of
theology, there is a hodge-podge "models" to support the ad-hoc concept
of "Emotional
intelligence" (EI) possibly measured by a corresponding
emotional quotient - EQ, analog with old abused IQ. Even forgetting that the
problems mentioned above may lie behind the reportedly increasing frequency of
the diagnoses of Autism,
including Asperger. It
is interesting to note the vagaries of people, especially when discussing the
arcane modern questions of whether AI will be able to “feel”. The flood of
mass-media speculations is exemplified in a program of 44 minutes at the
Swedish Radio on March 15, 2024 with the title Kan AI ha känslor? [Can
AI have feelings?] with the under-title “Today, humans are increasingly
interacting with AI. AI can make us humans feel joy and even love, but can AI
have emotions and what would they mean if AI had real emotions?” The program,
with two guests researchers Fredrik Svenaeus and Karim Jebari,
showing the practical impossibility of expressing and sharing psychological and
philosophical knowledge when philosophy that historically included psychology
is reduced to simple reference to “phenomenology” (appreciated in universities
for its theology without God), and theology
is programmatically and silently excluded from the conversation. This is
ultimately also the reason of why those who happen to study The Design of Inquiring Systems are not
able to figure out how an application of e.g. the ignored philosopher Hegel’s
and others’ philosophical thought (p. 180ff, see below) can
be related to Carl Jung’s analytical philosophy as it appears in the discussion
of “progress” (pp 203-205) and of “the mind” (p.261ff.), but the more so in the
conclusion that (p.272):
[I]t
may be safe to say that at the present time inquiry has become part of the
unconscious life of most people: they are unaware of the ways in which they
function as inquiring systems. Nor is there a strong inclination for them to
give expression to this function so that its nature appears at the conscious
level. As a consequence we are suffering now the most dangerous symptoms of an
inability to bring to the conscious level an important human function. We
assign to the experts and the politicians the roles of designing and creating
the environment in which we live because we can see no way in which we can play
any role whatsoever in these activities. Appalled as we may be at the events
that are occurring in the world about us as the output of blind technology and
politics, we each in our own frustrated way feel that we can do nothing about
it.
Medical
brain lobotomy, seen here as an archetypal analogy to the reduction of
intelligence to logic, mathematics and formal science divorced from emotions,
feelings and intuition, was supposed to deserve the highest level of
international official scientific recognition by means of the Nobel prize for
physiology and medicine, year 1949. It also raises the memory of the Nobel
prize in economy to Herbert Simon, while West Churchman arguably decreased his
chances of a similar official recognition because of his opposition to Simon's
view of intelligence in the artificiality of science and his view of "behaving
human beings as simples as ants", (probably not intended to
be applied to the author himself). As with the case of the Turing Test mentioned above, the step may be short to considering
certain “non-intelligent” humans as machines or ants to the point of allowing
for attitudes such as in the Holocaust. Churchman also opposed
some tenets of an earlier recipient of the prize in Economic Sciences,
Kenneth
Arrow. It belongs to this story the confession of a
colleague of mine who as member of the awarding
institution for prize in physics, the Royal Academy of Sciences,
that an internal report he had written about a case of this kinds of problems
in his discipline had been classified as secret for the next 50 years. Talking
about democratic freedom of expression and openness in research.
There
is a symptomatic case of a Swedish doctor and author, grand old man and
champion of intellectual debate in journalistic contexts, mainly in social
criticism and satire in the last more than 60 years: P. C. Jersild. One
of his best known books, A Living Soul, symptomatically relevant for our “lobotomized
brains”, deals with a living,
thinking “and feeling” (?) human brain floating in a container of liquid. Also
symptomatic is his answer at the end of a long interview of the Swedish Radio
in a program broadcasted on December 9 2023, about what he thought on the ongoing hype of AI-AGI.
After his many well-formulated answers and thoughts about “everything” during
the interview, including about his atheism and commitment to voluntary
euthanasia despite of being a
physician, he finally, with a certain hesitance, confessed that it was a
difficult matter and that he had not yet a formed opinion. Guess why. It is not
easy to know what a computer really is,
the less so as related to formal science and the human psyche, that is, again: human-computer
interaction, and “intelligence”.
Driving Forces
behind Technology
So
much for democratic freedom of expression at the highest levels of the highest
sciences, reminding the barely advertised that "Democracy is a myth"
because of negligence of its presuppositions, as I explore in a text of the information on the
conflict between Russia and Ukraine. Science as
scientific "fields" in general and computer science in particular,
and artificial intelligence more particularly, nowadays eschews all democratic
control because it is only controlled by narrow isolated self-controlled
communities of adepts who, as in universities and in supposedly democratic
Wikipedia, are the only who are seen as entitled to understand what it all is
about. And in the meantime the ongoing new wave of AI-hype is advertised by
means of a plethora of "visionary" programs of doubtful quality in
mass media and in discussions in social media: a reprise of the Dot.com bubble, possibly a revival of the historically archetypal Tulip mania. Let's remark that many people made big profits by
all those "visions", as they are making by Cryptocurrency.
Ultimately,
however, the question may boil down to which are the motivating forces that
drive modern scientific, technical and industrial view of western and already
also parts of eastern one, the latter’s antiquity having already been the
object of my reflections
in the text in Logic and Rape as related to Science and Civilization in China. It is a view that may be related to a peculiarity of
the Judeo-Christian civilization as related to “smartness” in the controversy
around the “Ashkenazi Jewish intelligence” (see here, here
published here, and
here). It is a view that is driven by
mathematics and logic of western mathematized technoscience, as synthetized
also in mathematical logic as embodied in computers. It is a matter that has
been studied in the philosophy of science and of technology that is touched
upon in an essay of mine on Trends in philosophy of technology. I believe that theologically this field was best
surveyed by editing of Carl Mitcham and
Jim Grote in Theology and Technology (volumes 1 and 2). My own experience, in witnessing
that even deeply committed and educated scientists and philosophers adore the
power of mathematics and logic (a power as suggested by Jan Brouwer, cf. my Computers and embodied mathematics and logic),
is that they are seduced by the feeling of power
(supposedly for good and for bad) over
nature and humans to the detriment of the Christian primacy of charity and love. It may be a case of the analytical psychology's conception of
Ego Inflation when believing the they
are high priests gifted with the capacity of understanding the
mathematical-logical language (synthetized in mathematical logic) in which God
himself is supposed to have described the created universe. Or, as I saw in a
letter written by a particularly gifted and educated physicist:
I particularly
remember an experience when I was taking a PhD course in quantum mechanics and
going through a particular proof using group theory to predict atomic states.
It was like a lightning strike and I actually started crying with emotion. It
was a powerful experience.
It is
interesting and symptomatic that this is a powerful experience, to the point of
convincing if not compelling humans to a religious conversion that could have
been a sign of the Biblical message that humans have a divine spark in
themselves because they are created in the "image of God".
Nevertheless, it can easily be a symptom of proud "Ego inflation"
mentioned above. It is therefore also the source of powerful speculations
raised around the names of Fibonacci and Mandelbrot, as
already considered in a couple of other essays, on Computers as Embodied Mathematics, Information on Christianism or Atheism, and Quantum Physics, Computers & Psychology. A
well-known Swedish professor of theoretical physics tells in an interview
at the Swedish Radio about a likewise life-changing (but not religiously
converting) powerful childhood experience in his witnessing the appearance in
the sky of the announced apparition of a comet. Other theoretical physicists
witness but are equally insensitive to religious conversions, such as e.g. the
famous Nobel prize laureate Peter
Higgs about whom Wikipedia reports:
Higgs was an atheist. He
described Richard Dawkins as having adopted a "fundamentalist" view of non-atheists. Higgs
expressed displeasure with the nickname the "God particle". Although
it has been reported that he believed the term "might offend people who
are religious", Higgs stated that this is not the case, lamenting the
letters he has received which claim the God particle was predicted in the Torah, the Qur'an and Buddhist
scriptures. In a 2013 interview with Decca Aitkenhead, Higgs was quoted as saying:
I'm not a believer.
Some people get confused between the science and the theology. They claim that
what happened at Cern proves
the existence of God. The church in Spain has also been guilty of using that
name as evidence for what they want to prove. [It] reinforces confused thinking
in the heads of people who are already thinking in a confused way. If they
believe that story about creation in seven days, are they being intelligent?
— The Guardian, 6 December 2013
But I
have not seen speculations at all about parallels with the relation to our
"powerful" experiences of feelings of passionate love as considered
in the essay on Reason and Gender. There are
people who infer from their own feelings of love, love for children spouses and
parents, that they have to ask wherefrom
comes this powerful feeling of love, as well wherefrom and why do tears
come when listening to certain kinds of music that ultimately can be religious, as I remind in my text on “Intuitions
in Music” in the paper Information
and Theology. There must be a “powerful” feeling of love in them, instilled
from outside, from above or the “inside” (Genesis
1:27). That is, love up to the point of being able to
sacrifice their own life in saving or losing the loved ones’.
If, as
in analytical psychology, it is a question of “archetypes” or worse,
“mechanisms” rooted in the human mind, then the answer will be that psychology
only affirms that they are there, not claiming to know wherefrom or from whom
they are placed there. Up to now I never heard people confessing to have
“started crying with emotion” or sobbing and getting tears in their eyes, upon
the intuition that their feeling of loving or being loved by their children,
but unhappily sometimes less by spouses divorcing spouses, is the same feeling
of godly love for us (not “intelligence”), making them to intuit how much we
can be loved, and to wonder where this feeling of love comes from. But many if
not most of them seem to be readier to prize mathematics for their own supposedly
godly mind that in fact navigates in the world as a proudly engineered Titanic
did in the ocean. Perhaps fantasy in science fiction has a presage of decadence
in imagining, as in the film Zardoz,
that the immortal “Eternals” leading a luxurious but aimless existence “are overseen and protected from death by
the Tabernacle, an artificial intelligence”. And
this is consistent with the hypothesis that ultimately it was mathematical
logic that raises disoriented powerful feelings.
A
support for this hypothesis comes from the simultaneous tendency to argument
only or mainly in logical terms (and/or with the help of ChatGPT!), without reference
to historical sources or “philosophical” thinkers. Whenever other sources are
mentioned, as here, the accusation is raised that it is an illegitimate “psychologizing”.
But see an example in my bold attempt to suggest with the accusation of
psychologizing, the enforcing of pure logic
can be experienced as a spiritual “rape”. Or the
accusation is raised that references beyond pure logical argumentation are an
abuse of “academic-philosophical-historical” speculations, while the question should
be seen “pragmatically”, whether it is convincing and “it works” in the
physical world, often combined with “it sells” that in turn is reduced to
coarse usability, utility and economic profit or “it is profitable”. One should
instead say: “As quantum physic works, until
a next nuclear third world war”. The appeal to pragmatism, however, is then
done without questioning the history and essence of philosophical pragmatism that
is downplayed to being only “philosophical-academic” interest.
And
talking about the forgotten love it comes to my mind what Oswald Spengler
writes in his amazing great work The
Decline of the West. Building, as he claims, upon mainly Goethe and
Nietzsche he writes in a book that I bought some more than 40 years ago but do
not claim to have been able to study and digest before reaching old age.
Nevertheless, in this present context I cannot avoid to remark what he writes
in vol. 2, pp. 218f.:
Religion is metaphysics and nothing else – Credo quia absurdum
– and this metaphysics is not the metaphysics of knowledge, argument, proof
(which is mere philosophy or learnedness) but lived and experienced metaphysics – that is, the unthinkable as a
certainty, the supernatural as a fact, life as existence in a world that is
non-actual, but true. Jesus never lived one moment in any other world but this.
[…]
“His” teachings,
as they had flowed from his mind and noble nature – his inner feeling of the
relation between God and man and of the high meaning of the times, and were
exhaustively comprised in and defined by the word “love” – fell into the
background, and their place was taken by the teaching of him.
And
today, when for many there is no much left of His teachings, but only “powerful experiences” possibly seasoned
with some theological erudition and reference-dropping, this can be taken as a
support of my view that mathematical powerful
experiences are an expression of the inflationary power in the Ego-Mind, and
not of human or godly love. It was the "super-intelligent"
power of mathematics and logic over
nature (and climate), and power over other human beings, the
"enemies". It was a “science” applied to wars with high-tech weapons
including "lethal autonomous weapons" (see here, here and here).
Application to wars, however, is more complex than the manufacturing of
products. AI will claim that it is, and will also be, applied to the general
use of weapons such more or less conventional bombs, as illustrated in the case
of the 2023 Israel-Hamas war in an article by Jonathan Cook (see
also here in
the necessarily highly controversial Unz Review, December 5, 2023). Here follows my
excerpt from the article, including a couple of readers’ comments:
The whistleblowers confirm that,
given new, generous parameters of who and what can be attacked, the artificial
intelligence system, called “Gospel”, is generating lists of targets so rapidly
the military cannot keep up.
…
In a report last Friday [December 1, 2023], the Guardian [in an article with the title ‘The
Gospel’: how Israel uses AI to select bombing targets in Gaza, with
repercussions as in Le Monde, Dec. 6, 2023] corroborated Israel’s reliance on the
Gospel computing system. The paper quoted a former White House official
familiar with the Pentagon’s development of autonomous offensive systems as
stating that Israel’s no-holds-barred AI war on Gaza was an “important moment”.
…
Israeli military is now using an artificial
intelligence system, Habsora or Gospel, to identify targets.
…
Speaking of the military’s new reliance on Gospel,
Aviv Kochavi, the former head of the Israeli military, told the Israeli Ynet website earlier this year [June 23, 2023] “In the past, we
would produce 50 targets in Gaza per year. Now, this machine produces 100
targets a single day, with 50 per cent of them being attacked.”
…
A former intelligence officer told that the Targets
Administrative Division that runs Gospel had been turned into a “mass
assassination factory”. Tens of thousands of people had been listed as “junior
Hamas operatives” and were therefore treated as targets. The officer added that
the “emphasis is on quantity and not on quality”.
…
COMMENTS [two
selected comments of the article]
The REAL value of “AI” is as lying sockpuppet to take
the blame:
•Health Insurance companies use “AI” to deny claims. (Don’t blame us)
•Military uses “AI” to kill civilians. (Don’t blame us)
• Hedge Funds use “AI” to insider trade. (Don’t blame us)
• Corporations use “AI” for sales targets during down quarters (Don’t blame
us).
This is the other face of Sam Altman’s empire!
The
Ego’s inflationary power offers also an important advantage of logic (and
consequently AI/AGI) in the politics of academic careers: it dispenses painful
references, not to mention charitable attention to contemporaneous and historical
work of the “neighbors” that we are supposed to respect, if not to love. A
logically structured text is the only way of claiming attention without a
single reference to, and (worse) dependence upon prior knowledge except
technical and analytical. That is: unless one questions the essence of
mathematized science, or of logic and mathematics in inquiry. So, in my
experience, logically oriented scholars, now with the support of heavy AI/AGI
technology, can do well without any academic “fatherhood”, let alone an ex-post legitimately criticized
fatherhood. The more so by adducing the rapid technological development
invalidates the import of earlier, even 5, or 10 years old insights. This can
be related to youngsters’ lack of respect for the knowledge and experience of
their parents, the more so of their grandparents, not to mention of generations
of people who with struggles and sacrifices have built our accumulated
knowledge and wellbeing. Their implicit fatherhood is the pseudo-religious Logos that is felt as implicit in the abused assumed
etymology of the word “Logic” as used today.
It is
also symptomatic that there is a spread rumor (see Internet browser on
<mathematics god language>) as if it were Galileo Galilei's quotation,
that he would have written that mathematics is the language of God. There is,
however, only one documented source, The Assayer, in which he really only claimed that it is the
language of science (natural science), without mentioning God. Further
theological aspects of the issue are found by searching the keyword <mathematics>
in my essay on Information and Theology.
The contempt for philosophy and not only theology,
based on ignorance and secularism is also what guides smart authors of smart
articles in prestigious journals such as The
Economist. In an ambitious article
noted in one main Swedish newspaper, Svenska Dagbladet (Nov 19 2023)
about the economic-political events in November 2023 in the OpenAI organization, it
writes, for instance, referring to the events that are detailed below, relating to the board and CEO of
OpenAI:
The events at OpenAI are a dramatic manifestation of a wider divide in
Silicon Valley. On one side are the “doomers”, who believe that, left
unchecked, AI poses an
existential risk to humanity and hence advocate stricter regulations. Opposing
them are “boomers”, who play down fears of an AI apocalypse
and stress its potential to turbocharge progress. The split reflects in part
philosophical differences. Many in the doomer camp are influenced by “effective
altruism”, a movement worried that AI might wipe
out humanity. Boomers espouse a worldview called “effective accelerationism”,
which counters that the development of AI should be speeded up.
So,
the difference between doomers, among
whom many readers would like to see myself included, and boomers, would reflect
“in part” philosophical differences. Both
doomers and boomers could be influenced by “effective
altruism” (that I have already considered in a section
of my text on Information and Theology). What
is not said is that both of them may be projecting their desperation, respectively
their hopes, in lesser or greater capabilities of “democracy”. What is not
spelled out is that the effective altruism, as represented by the outspoken atheist
“moral philosopher” Peter
Singer, is a failed substitute of religion. Boomers do not
seem to be influenced by anything. Once upon a time they might have been
denominated as being simply optimists, while the further discussion in The Economist indicates that
psychologically and politically they might be rather seen as entrepreneurs or
“individuals who create and/or invest in one or more businesses, bearing most
of the risks and enjoying most of the rewards” in this life, in view of absence
of a future one. It was the strong concern of investors that drove the dramatic
development of the relation between board and CEO of the OpenAI organization.
So much for philosophy and money, forgetting also the philosophy of technology,
besides economics and politics.
Ego Inflation or philosophy behind “Powerful
Experiences”
I
know of a particular case, beside one
documented on the Internet, in which an ego inflation may have contributed to
the total conversion of a thoughtful educated and intelligent man to
Christianity, a process that I characterize as a serendipitous, paradoxical
"right decision based on wrong premises". He found that since the
decision was right it doesn’t matter whether the premises were wrong. But it
discloses a defective understanding of the “system” and of the interpretation
of Christianity as illustrated in the essay on Conscience and Truth by
former cardinal (and later pope) Joseph Ratzinger. The same premises may lead
to other related wrong decisions, including in applications or implementations
of the original decision, such as in interpretations of Christianity when
advocating and justifying USA’s use of nuclear bombs on Hiroshima and Nagasaki
(justified for avoiding “worse” consequences) or the handling of the conflict
between Israel and Palestine, ultimately illustrated by the 2023
Israel-Hamas war, or as I show, between Russia and
Ukraine. The Israel-Hamas war offers a good example of
defective natural intelligence that obviously undermines the developments and
evaluation of AGI, as illustrated by a (here slightly edited) communication I
received from an avowedly Christian engineer and scientist who felt strong
sympathies for Israel:
[N]ow we touch
on a question that cannot be solved by philosophy. Regardless of all philosophies,
there is a reality that cannot be philosophized away. For example, in some
cases you have to make decisions, which regardless of which decision you make,
have negative consequences for people and society. And making no decision at
all can lead to even worse consequences
[...]
It is easy to
sit in your best armchair with a cup of Darjeeling tea and a slice of toast
with Cooper's Oxford Marmalade and comment on [others'] actions. And weigh the
different options against each other. When you are in her situation, it is not
so easy to play wise and good. Then one is forced to confess color.
[R]eal life
shows how we humans can be faced with terrible decision situations where
whatever we decide leads to negative consequences. Still, a decision must be made.
Not making a decision at all may lead to even worse consequences. Trying to
play well in situations like this usually doesn't work.
Sitting in one's
armchair and thinking beautiful philosophical thoughts that have nice Latin
names (that make oneself appear wise and good, etc.) can certainly be done. But
reality is sometimes more complicated than our theories about it (the kind of
philosophy I now call "mental masturbation"). Over the years, I have
become increasingly negative towards intellectual people, which led me to
define the term "stupid-intelligence".
It is
a similar coarse conception of “philosophy” and a-philosophical pragmatism or common
sense, or process of conversion or decision by unconscious
emotion in face of powerful experiences, that has "divinized"
violence and wars, including the nuclear bombs on Hiroshima and Nagasaki in
order to avoid supposed “worse consequences” (for me and/or others than the
bombed?). It is also a way to relativize a defective understanding of love, as
when stating,
for instance (transl. from Swedish) “Without truth, love risks being reduced to a
syrupy drug. And without love, the truth risks becoming an icy regulation”.
This statement, which per se may legitimately states the Christian dialectic
between charity and justice, is understandable in view of the present daily
tendency to use and abuse the word truth as mathematical formal truth, and love
as covering everything sexual and “nice”, while ignoring the problem of Reason and Gender. But,
besides the typical logicist (logical positivist) assumption that regulations
or logical rules only need to be complemented with separate vague “love”, it is
also a simultaneous denial of the Christian primacy
and essence of love (Matthew 22:37-40,
esp. Galatians 5.14)
coupled to a subjectivist implicit perception of the meaning of truth. It is a
“truth” that, opposite to the Catholic need of a Church, relies only upon a
Protestant personal reading and interpretation of the Bible, eschewing the
supposed logical positivist “psychologism” in
relating the whole question to depth psychology as in the former cardinal and
later pope Joseph Ratzinger in his essay Conscience and Truth.
It is
also this contemptuous attitude to philosophy that allows the ignorance of the
meaning of “powerful” as already painstakingly discussed in the context of “the sublime”. It
is also this kind of ignorance that allows for the accusation of “psychologizing”
when one does not know that psychology originated in philosophy, and further,
that historically philosophy appropriated the realm of theology, while the
latter is offset by cheap acknowledgment of being a pious Christian believer
who quotes the Bible. This when, in fact, one is a logical
positivist who not only legitimately distinguishes but also separates facts of science from values
and religion
that is soon reduced to politics, as I show in the essay on Information and Theology.
A
separation or facts of science from values is paradoxically also a disclaimer
of the engineers’ responsibility for the consequences of their work and
possible failures (so long as they are on the payroll and get paid by
somebody). This is so because the engineered products, including “tools” can be
said to have been always “right” but were used in the wrong way, equivalent to
that they should not have been used
(before being tested in the right way), or should not be have been given to
people who could use them wrongly (e.g. in the case of weapons). An example I
have personal experience of is in the context of repeated groundings of the
aircraft Boeing
737 MAX 9, the latest in January 2024 (historical
overview here). A
knowledgeable aircraft wizard assured me that the problems were not really
technical but rather managerial, being a consequence of the Boeing’s management
in later times having prioritized economic interests before technical ones. It
is easy to imagine how disclaimers are going to work with engineering products
like AI/AGI that are less “concrete” than aircraft.
The
separation we talk about, implies a contempt for philosophy that fits the
protest and rejection of being labeled at all; as, say, logical positivist or
logical empiricist. The answer, then is “I am not logical positivist, I am only
my name, (say) John Smith”, picking an opinion here and an idea there. This
then happens without realizing that this also has a labeling name, eclecticism, and
that by the (father’s) family name one is labeled as a member of a family,
which in turn is labeled by the name of the father, as Christians acknowledge
having a common Father that makes them brothers. Others unconsciously may
assume that their family is only their Mother (and half-brothers) who loves all
her children obtained (“eclectically”?)
from many fathers (many cultures having a special contemptuous and cursing denomination for
such behavior), according to the conceptions that I illustrate in my
introduction to Reason and Gender.
It is only gods, geniuses or “logicists” who without
any outer references, have an innate skill (often related to high IQ) to
connect logical Leibnizian networks of empirical data (logical empiricists),
who dare believe that they can create their own individual a-historical “-isms”
and create an own logically consistent “Leibnizian” system of thought and
unconscious feelings. They live without a family
and without a history, since the renounce to the evaluation of “-isms” is also
an implicit contemptuous renounce to the supposed “obsolete” thoughts of our
forefathers whose errors and achievements are the base of our ongoing life.
They do not foresee that their own present work and conclusion will be
contemptuously rejected or outright ignored by others within few years.
Unconsciously they rely on other “-ism” such as the mentioned eclecticism and state individualism, where
the state, often confounded with society or one of its scientific subcultures, is
the family of those without family. They do not realize that in relying on
their own logical networks without references to earlier authorities it is like
they were gods editing their own “bible”, and forgetting that the Bible relies
on God (and historical accounts). And they forget that typical logicists who do
not understand the meaning and limitations of language from which logic is
extracted (and not the other way round) reject the Bible itself because of
logical contradictions. Or abuse the Bible by ignoring them, reading the Bible
as the devil does. These considerations lie behind the insight, highly relevant
for understanding the present essay in face of the issue of AI/AGI, that the
crushing amount of references and apparent complication of its argumentation are not necessary for a reader who is guided by the Bible
instead of treatises on mathematical logic. This fits a key statement of Blaise
Pascal in a relevant section of his famous Pensées (§ 251) which,
in order to be properly understood must be read and studied in its
context, readable on the net (preferably in original
French because of orthographic misspellings in translation):
Other religions, as
the pagan, are
more popular, for
they consist in externals.
But they are
not for educated
people A purely intellectual
religion would be
more suited to the
learned, but it
would be of
no use to the common
people. The Christian
religion alone is
adapted to all,
being composed of
externals and internals.
It raises the
common people to
the internal, and
humbles the proud
to the external;
it is not
perfect without the
two, for the
people must understand
the spirit of the
letter, and the
learned must submit
their spirit to
the letter.
Otherwise,
it requires advanced secular ingeniousness to conceive tests of AI-AGI products
beyond the example of my essay on the Ukraine-conflict. I mean tests
sufficiently pedagogical in order to put into evidence their possibly dangerous
limitations, exemplified by the AI creation of original jokes, overviewed in
e.g. an article
in Time magazine in
January 2022, being an issue that expands on the Internet as it can be
perceived by browsing with strings like <artificial intelligence AI humor
humour>. One such test of AI limitations was illustrated by Anders Q.
Björkman in the Swedish newspaper Svenska
Dagbladet, December 2-3 2023, “Jag
hetsar AI:n – och då händer något” [I provoke the AI – and then
something happens]: it is the request that the AI tells an own joke. It fails,
but the journalist refers to the Swedish National
Encyclopedia, as defining humor as arising from
"the
collision between (at least) two human rule systems. To the extent that the
collision turns out to be apparent or explicable from a third rule system,
there is a case of humor"
What
is interesting here is that humor is conceived as depending upon rule systems.
As such, if rule systems are synonyms for logical Leibnizian networks it also
illustrates the spontaneous psychological relief experienced with the insight
into the relativity and insufficiency or framing statements in only logical
terms. I read somewhere that the German philosopher and theologian Ernst Troeltsch
indicated that humor may have similar influence as religion in fostering
humility by means of a downplaying of the importance of the big Ego (that is
inflated by logic). And yet we are far from understanding humor if we consider
the difficulty in understanding a standard work on humor in the world’s
literature: Henri Bergson’s book Laughter. I
claim that the process of understanding, and evaluate exceeds the scope of this essay, but returns us to the broad philosophical (including the
psychological), and theological issues suggested in The Design of Inquiring Systems. A meaningful curiosity, however,
can be mentioned here is that Bergson emphasizes that it is easier to laugh collectively, which can be seen as
implied in humor’s downplaying of the individualizing Ego.
It is
the afore mentioned separation between facts of science from values of religion
(nowadays profane democratic “core values”) that the aforementioned Churchman
struggled to counter in his whole work, at the (academic survival’s) cost of
renaming God (in his The Design of
Inquiring Systems) with the acronym GOD
- Guarantor of Destiny, which in turn allowed an academically acceptable,
watered down further
(mis)use of the term, while “divinizing” a priestly role for, and
responsibility of the Designer. That
is, a divinized “Design” which in turn is a misunderstanding and abuse of the
Kantian role of aesthetics in his Critique of Judgment, as
addressed in my text on computerization of society, and in a commentary to Gunnela Ivanov’s doctoral
dissertation relating to the democratization of design.
David
Noble and “The religion of technology”
In
1998 the historian and “critic of
technology” David F. Nobel
published the book The Religion of Technology, a subject that from a (for me) controversial “leftist”
perspective recalls the more politically neutral idea of Richard Stivers’ book Technology as Magic: The triumph of the irrational. What is of immediate interest here is Noble’s
historical account of the reports and thoughts about AI up to about the year of
publication of his book, as it is found with the title “The immortal mind:
Artificial Intelligence” on pages143-171. In Noble’s account, the “powerful
experiences” surveyed in the above section are a are not a revelation but only
a triviality that accompanies the development of science and technology.
Nevertheless, he does not linger on the reason of the main involved phenomenon that
he expresses as follows (p. 145, my italics):
Descartes’
peculiar obsession [with geometry and arithmetic, closer to God; my note] became
the principal philosophical preoccupation for three centuries, as diverse
thinkers sought to comprehend the mechanisms of human understanding, the
categories of reason, the phenomenology of mind. Moreover, in the midnineteenth century, mathematics became not just a model for pure thinking but the means of
describing the process of thought itself.
But why it did just became? Thinking and feeling like a historian he was, Noble does
not dwell on a philosophical criticism of Descartes as related to theology, and
why it would have prevailed. In other words, Nobel’s argument allows him to
dispense the expansion of his thought in the direction of philosophy of
science, and the painstaking work of placing Descartes in a wider context as in
Churchman’ s DIS (pp.18, 22, 37, 61-62, 218, 263, 274, 276). It would have
spared him to develop or envelope further his thought in the direction of Marxist thought,
which he does not perceive and foresee that, in lack of theological
insight, would in turn reinforce faith in, and commitment of capital to
technoscience in business and industry. As when reading his concluding affirmative
quotations from other authors (p. 208):
The religion of
technology, in the end, “rests on extravagant hopes which are only meaningful
in the context of transcendent belief in a religious God, hopes for a total
salvation which technology cannot fulfill…By striving for the impossible, [we]
run the risk of destroying the good life that is possible.” Put simply, the
technological pursuit of salvation has become a threat to our survival…
[…]
Transcendence…means
escape from the earth-bound and the repetitive, climbing above the everyday. It
means putting men on the moon before feeding and housing the world’s poor…The
revolutionary step would be to bring men down to earth…But respite from our
transcendent faith in the religious machine requires that we “alter the
ideological basis of the whole system.”. Such an undertaking demands defiance
of the divine pretensions of the few in the interest of securing the mortal
necessities of the many, and presupposes that we disabuse ourselves of our
inherited other-worldly propensities in order to embrace anew our one and only
earthly existence.
Despite
my admiration for Nobel’s historical accounts, his conclusions are a temptation
for an understanding, despite of it all, of the motives for USA’s McCarthyism or Second Red Scare. It
can be seen as a reenactment of the Inquisition and
its problems, except for Democracy (with capital D) being seen as
a substitute for God, as explained by Tage Lindbom in
his book The Myth of Democracy. The same
question reappears in Nobel’s apparent “Marxist
cultural analysis” that would be concretized and lead to the Cultural
Marxism. What is
interesting for the argument here is not Nobel’s “quoted conclusions” above,
but his gathering of the historical arguments that he witnessed during the
formative years of AI development. Among others:
The reduction of
human thought to mathematical representation made imaginable the mechanical
simulation or replication of the human thought process. For once, the mysteries
of the immortal mind were rendered transparent and comprehensible… The thinking
person might then be joined by the thinking machine modeled upon the patterns
of human thought but independent of the thinking person. [p. 148.]
Confronted by
the limitations of mechanical analog computers while overseeing the operations
of MIT¨s Differential Analyzer, the most advanced computation machine of its
day, [Claude] Shannon suggested speeding up
and simplifying the system by substituting electromagnetic relays for machined
parts, using Boole’s binary arithmetic to describe the electrical network. By
using the Boolean system, invented to describe the laws of thought, to describe
the operation of electric circuits, Shannon laid the groundwork for the
electrical simulation of thought – the foundation of electronic computers. [p.
149.]
With his
minimalist definition of machine intelligence, [Alan]
Turing had deftly sidestepped philosophical discussions about
the actual meaning of mind and thought; his materialist approach dismissed at
the outset any discussion of the existence of an autonomous mind or a soul,
which had preoccupied Descartes and Boole. (Turing had by this time become an
avowed atheist). [p. 151.]
Nearly all of
the theoretical developments that made possible the design of computers and the
advance of Artificial Intelligence stemmed from military-related experience.
[p.152.]
[Allen]
Newell and [Herbert]
Simon wrote: “The vagueness that has plagued the theory of
higher mental processes and other parts of psychology disappears when the
phenomena are described as programs.” [p. 155.]
The generally
recognized father of what became A-Life [Artificial Life] was the mathematician
John von Neumann, the “main scientific
voice in the country’s nuclear weapon establishment.” Toward the early end of
his life, suffering from terminal cancer, von Neumann earnestly devoted himself
to weapons development, advocating the use of nuclear weapons and favoring a
preventive nuclear war. At the same time, he began to ponder the fundamental
logical similarities between life and machines…[p.165.]
Buttressed by
government funding and institutional support, A-Life advocates shared with
their Artificial Intelligence colleagues and arrogant impatience with
criticism. [p.169.]
In short, Noble purport to show how religious and dangerous fantasies
have invaded the imagination of many people who were the pioneers of computer
applications related to AI. In order to appreciate them properly without underestimating
their human dignity we may see them as particularly gifted in one specific aspect
of their personality, recalling the controversies and discussions of “Ashkenazi
Jewish Intelligence” (see e.g. here, here, here, here, and here). Regarding the AI colleagues
showing “arrogant impatience with criticism” I may confirm or debunk this
affirmation depending upon possible reactions to the present text of mine.
Arrogance is the psychological sign of what psychologists have indicated as
“Ego inflation” as mentioned above and below in my text, and more provokingly
in other texts such as on Logic
and Rape.
Autism, or “Mathematical computer-oriented Minds”?
The
following is a digression along the lines of the previous sections on cultural
lobotomy, the human psyche, driving forces behind technology, and ego-Inflation
behind powerful experiences.
It
deals with the kind of decisions based on logic and “facts” and unconscious
emotion that in the future can be taken more easily with an apparently
undebatable ChatGPT-support, and without having an idea of what systems
thinking is about. This has also divinized relativity theory, quantum physics (“it
works”), music and luminaries such as the “divinized genius” Albert
Einstein, and others such
as Richard
Wagner, all of them often with their own theological musings
(here, here, here and here).
There are many logical-mathematical
manipulators whose thought process is described in the study of the foundations
of mathematics by Jan Brouwer.
There are also
information and computer geniuses or more generally engineering
geniuses like the historical Vannevar
Bush (that I quote and comment elsewhere,
without relating him as I could to the mention of Claude Shannon above) who could
as well be characterized also particularly gifted humans such as great writers,
musicians, dancers, poets, artists, sorcerers, or now computer wizard or
"mental acrobats". What is opportunely disregarded is sometimes
meaningful aspects if not troubles of their personal lives. They can also have
read the Bible very carefully, being able to quote selected pieces of text that
either support their own interpretation of its message, or show perceived
contradictions which supposedly (logically) demonstrate its irrelevance (cf. “Even the Devil
Quotes Scripture”). They display capabilities that recall the “savant syndrome”. It
is worth noting that this latter denomination is a "politically
correct" shortening an earlier one: idiot
savant, which can be misinterpreted when ignoring that etymologically idiot, as in the world idiographic, derives
from the root idio, with
a dictionary meaning as "only one's own": personal, separate,
distinct, unique, also sometimes qualified, respectively in Dictionary
boxes on Google and in Wikipedia, that it really means (if one does
not deepen the meaning of “wisdom”):
"A person
who has an exceptional aptitude in one particular field, such as music or
mathematics, despite having significant impairment in other areas of
intellectual or social functioning; or a person who is extremely unworldly but
displays natural wisdom and insight."
“Savant syndrome
is a phenomenon where someone demonstrates exceptional aptitude in one domain,
such as art or mathematics, despite social or intellectual impairment.”
It is
interesting to note the relation to lobotomy in the case of the savant example
of Kim
Peek, who is reported to have had a “condition in which
the bundle of nerves that connects the two hemispheres of the brain is
missing”. I will propose another analogy to the “cultural lobotomy” and “savant
syndrome”: the likewise complex but very modern term autism. It
is qualified in Wikipedia as a neurodevelopmental
disorder, soon argumentatively adjusted to be seen as a particular
neurodevelopmental “order”, as in a broadcasted program
(March 12, 2024) at the Swedish public television SVT in a series on
“Personality” where both autism and ADHD were
seen as human diversities, or just different capabilities. I see this language manipulation as a postmodern
relativism, as when dysfunction or disability tend to be renamed as a
(functional- or ability-) diversity. All
this while the American Psychological Association, APA, still defines them as
for dysfunction (updated April
19, 2018) as “any impairment,
disturbance, or deficiency in behavior or operation”, and disability (updated November 15, 2023) as “a lasting physical or mental impairment that significantly
interferes with an individual’s ability to function in one or more central life
activities, such as self-care, ambulation, communication, social interaction,
sexual expression, or employment. For example, an individual with low vision or
blindness has a visual disability.”
My
point in this section will be to show how this sort of autism or diversity finds
a legitimate expression in a book by Paul
Hoffman, Archimedes’ revenge: The joys and perils of
mathematics, published
in 1988 during one of the hype periods of artificial intelligence, which was
not yet hyped with the addition of neologism as artificial general intelligence. The
title itself refers to the frustrating “Archimedes’
cattle problem”. The point I want to make may be also enhanced by
reading and meditating upon at least the Wikipedia
account of a later book by Hoffman, with the revealing title
The man who loved only numbers, published in 1998.
Keeping
to Hoffman’s first mentioned book, it is interesting to note that besides the
understandable lack of reference to G. Leibniz in his extensive 11-pages word
index, he has the following kind of disclaimers, which position his
understanding of mathematics and logic (pp. 2-4, 159) :
Many books have
been written about the philosophical underpinnings of mathematics, about the
extent to which it is the science of certainty, in that its conclusions are
logically unassailable. Many other works have rhapsodized at length about the
nature of infinity and the beauty of higher dimensions. Such philosophical and
poetic excursions have their place, but they are far from the concerns of most
working mathematicians. In this book I give a glimpse of some of the things
that mathematicians, pure and applied, actually do. […]
With fundamental
questions about number and shape still unsettled, it is no wonder that there is
much disagreement and confusion about that the computer -- a very complex
mathematical tool – can and cannot do. I have tried to stay clear of mushy
metaphysical issues about the nature of man and machine in favor or presenting
what little is known about the theoretical limits on computing. […]
Much ink has
been spilled in the philosophical literature on what the absence of judgments
means in terms of a machine’s ability to think, but the pursuit of such
intriguing speculations would take us too far afield.
This
does not prevent Hoffman from, deep in the book (p. 168), speculating about
using Boolean algebra in e.g. representing the state of two people’s being
friends and not being friends, irrespective of the meaning and measurement of
friendship. The problem is extended to his discussion of democracy when stating
that indeed, mathematics demonstrates the theoretical futility of creating a
perfectly democratic voting system. He refers (pp. 5, 223) to the “wrestling
with the mechanics of setting up a democratic nation”, and refers to the “Nobel
prize-winning work of the American economist Kenneth Arrow”
that shows that achieving the ideals of a perfect democracy is a mathematical
impossibility. Never mind about the problem of Arrow’s
conception (“impossibility theorem”) indicated by
e.g. West Churchman, and about the
myth of democracy, while the attention is focused on its mathematics.
And Hoffman spends pages (e.g. 237f.) referring to books applying game theory
to conflicts in the Old Testament between God and human beings, including
game-theoretic implications of omniscience, omnipotence, immortality, and
incomprehensibility. Not to mention the pondering (p. 144, 211) that “a machine
state is like a mental state” and the pondering about the difference between
electronic components in a computer and neurons in the human brain.
All
this shows how the hubris of artificial intelligence, even early long before
the success of today’s so-called general artificial intelligence, grew from the
apparent self-imposed limitations of reasonable people who were certainly not
mentally impaired, if yet possibly living and thriving in a sub-culture. The
problem then is not that they were autistic in simple, raw medical clinical
sense. In analytical psychology there is the conception of Psychological Types, which may be seen as vulgarized in the culturally and
historically uprooted Five-Factor Model of Personality- A better variant is popularized in the book Surrounded by Idiots, and others by the author and management consultant Thomas
Erikson with the DISC assessment
tools.
In these popularizations of types the question is the interplay between
variants of mathematical statistics and especially cluster analysis, in
mathematical approaches to personality psychology, in oblivion of its “philosophical
assumptions”. Otherwise,
the question in analytical psychology is of the interplay between thinking,
feeling, sensation and intuition in relation to consciousness and the
unconscious, and psychic development is the capacity to achieve a balance among
them in the so-called process of individuation. Such
conceptual constructions are not more absurd than the mathematical ones as in,
say, quantum
physics, if philosophical pragmatism is taken seriously. They
have a theological counterpart in the Bible’s famous Corinthians
12:14ff.
“Now to each one the manifestation of the
Spirit is given for the common good. To one there is given through the Spirit a
message of wisdom, to another a message of knowledge by means of the
same Spirit. […]”
The
cultural problem appears when a specific constellation of aptitudes are awarded
improper high value by society at large, such as by industry, finance and
academia in a secular technical-industrial culture developed after the scientific revolution. As in
Wikipedia’s quotation from William Whewell’s Philosophy of the Inductive Sciences: it
is the (my italics) “the
transition from an implicit trust in
the internal powers of man's mind to
a professed dependence upon external observation”. It is the external
observation of empirical experiments that are assumed to verify results of
mathematically conceived relations to earlier external observations. Society
and its politics begin to reward mainly thinking seen as equivalent to mathematical-logical
ability while mathematics and logic undergo changes in themselves and in their
functions, subservient to only external observation. And this can be seen as
inflating the self-confidence (“ego inflation”) of seduced mathematically and
logically gifted scientists who are overvalued by “democratically” selected
political leaders and managers who strive for profitability and economic-industrial-military
power in a society in a cultural crisis being driven by the Western world.
Artificially intelligent artistic production?
Extending
the question of powerful experience we
come to the problem of how to frame the question of AI/AGI influence and use in
artistic production, where art stands for the production of powerful experiences.
The matter has already been noted in the cultural world and even mentioned
above, but here it will be inspired by an article on (my trans.) “Artificial
music is the future’s melody” published in the major morning
newspaper Svenska Dagbladet (30
December 2023), written by Björn Ulvaeus,
presented as a Swedish musician,
singer, songwriter, and producer best known as a member of the musical
group ABBA. The hype about this issue is
illustrated also by the fact that the following day, another author, Ebba
Bonde, wrote an article dealing with the AI-manipulation of pictures/images
having, however, the typical uncommitting and advertising title “The
use of AI is as much a danger as an opportunity”.
Ulvaeus
starts by acknowledging that we all are influenced by what we have listened to
in our lives, that he does not know where his inspiration comes from, and he
expresses a deep gratitude for the miraculous mystery of music’s mere
existence, without referring to his otherwise well-publicized atheism (that
prevents thanking God). He assures the readers, however, that “we do not need
to feel shame for having difficulty with the definition of what music is”. He
notes that sounds have a remarkable ability to influence our “feelings”, and to
penetrate in our “unconscious”. In this way he proceeds without feeling shame
for having difficulty with definitions, this time not only of music but of
feelings or the unconscious, which happen to have been main matters of
struggles for theologians, philosophers and psychologists for centuries. Giving
up these struggles, the ongoing drive for computerization of society, focused
here on AI, is combined with an aestheticizing post-modern abuse of Kantian
aesthetics that has been renamed DESIGN.
In this perspective Ulvaeus as musician and songwriter is a designer. It all follows unconsciously
the steps of the likewise abused post-Kantians Nietzsche and the academically
often referred, if not namedropped, Heidegger. I have written about this in the
essay on Computerization as abuse of formal science, with the alternative earlier title of “computerization
as design of logic acrobatics”. It deals now with the acrobatic use of AI, when
its users admire the circus acrobatics of computer scientists turned into
craftsmen.
Therefore,
the text proceeds swiftly to state that up to now songs that we love have come
from humans how have dug deep in their psyches, experiences, sorrows, triumphs,
fears, losses and endeavors. “But now has appeared a new type of composer,
eager to learn by means of “deep learning…with a neural network that loosely
resembles the human’s, but not yet.” Ulvaeus tells about being introduced to Google
Deep Minds “Music AI incubator”, which he felt was an overwhelming
experience that seems to have made him mainly if not only aware of the importance
of finding solutions for protection in “AI-training” or “prompting” from
infringements of artistic copyrights related to sound, style, genre, voice,
instrumentation, melody and lyrics.
But
the reader of his article senses some sort of insight in what follows:
Songwriting is
an intuitive and deeply human activity, and the big question is whether an AI
model can create something that reflects the humanity behind a song, and
whether that actually matters to the listener. Is there a depth to what a human
can create that an AI cannot replicate? If so, can we appreciate that depth? Do
we care? Or will our brains fill in the blanks and make us think we're hearing
something that isn't there? - Musical experience is subjective. And isn't the
only thing that matters what comes out of the speakers and the effect it has on
the listener? Whether it was created by AI or a human? Or?
[...]
The Abba story
made our "Voyage" avatar concert
in London possible. It is a human story of hard work, triumph, disappointment,
joy, love and sorrow. [...] It is so organic and full of contradictions and
crossroads where we could have taken a different path that I cannot help but
cling, perhaps vainly, to the hope that it could not have been artificially
created.
[...]
If AI, through
the lens of language, can learn everything a human can know about the innermost
nature of the world? Can the creative process itself be replicated by AI? The
process by which ideas born of curiosity and imagination are realized and push
the boundaries of the human art world.
[...]
Will AI help us
create great new work or make us obsolete? Personally, I think we will learn to
coexist, as we have always done with other technological inventions in the
world of music.
To
start from the end: “we shall learn to coexist?”. So, it is not AI that will
teach us to do it? As I claim that it will not write my essay on
the Russia-NATO-Ukraine conflict? Nor it will make labor unions superfluous by
teaching “us” what to do about conflicts of interest, greed, hate and
unemployment? And Ulvaeus refers to the assumedly prophetic Yuval Noah
Harari who means that AI has “hacked” the operating system
for the human civilization. So much for prophecy and for knowledge about
operating systems, as it transpires in Harari’s Can mindfulness save us from the menace
of artificial intelligence? An unthought prayer to an
imaginary God apparently cannot compete with “mindfulness”, since it is reported
that Harari is “derisive about the
great religions having anything to offer” and
“to the best of our scientific knowledge, all these sacred texts were
written by imaginative Homo sapiens. They are just stories invented by our
ancestors in order to legitimize social norms and political structures.”
That is, “to the best of his knowledge”, which also
fits what has been called Sweden’s official religion of atheism, also practiced
in the philosopher Åsa Wikforss’ analysis of so-called
fake news (while ignoring The Myth of Democracy), which are related to the consequences of AI/AGI.
Harari’s “mindfulness” promisingly should have meant that “we better understand
our minds before the algorithms make our minds up for us” but it seems to
either ignore or counter everything that the most mindful students of the mind
have come up, beginning with the analytical psychology that e.g. Jordan
Peterson adduces as I myself have
done in my writings. The problem, however, is defectively posted to begin
with, since it does not address the question of the meaning and function of art
as necessarily done in philosophy, as suggested in my essay on the influence of
Immanuel Kant in the conception of societal computerization
as an abuse of formal science. And regarding music Ulvaeus’
understanding misses the primary meaning of art as human communication,
communication between humans closely related to love, whose corruption becoming
visible in this discussion of AI in art is surveyed in my preface to Reason and Gender. Not to mention the missed history and meaning of music
in religion and theology as indicated in my essay on Information and Theology.
Concerning
“avatars” as in avatar concerts it is
possible to identify the same tendencies of dehumanization of human
relationships as outlined about 2D-Lovers
in “The
technological turn” of my Reason
and Gender.
For
the rest, I claim that the above approach to artificially intelligent artistic
production exemplifies the same dehumanization process that I noted above in
the context of the hyped Turing Test, about
which I express that the less one understands and feels what a human being is,
the more he will be equated to a machine, an avatar, an ant or whatever, as in
the Holocaust. The decline of culture and civilization coupled to an
undervaluation of philosophy and theology will necessarily become visible in
the discussion of more abstract issues such as art. Consequently one will not
be able to understand neither what Plato writes about the effects
and importance of music (Republic, 398b-403c,
but esp. 424b-c) nor music’s role and effect in the paganism, atheism, and future of our youth.
More
generally one will not be able to understand the meaning of difference between
an original work of art and a copy of it which stands at the heart of the
(infringement of) copyright. Original means that something is (problematically
established to be) not consciously and openly influenced by, or rather
generally recognized as similar to a prior human creation. In terms of
analytical psychology, however, two or more works of art can be influenced or
inspired by an archetype that belongs to the collective unconscious and as such
legitimately common to humans. This problematizes the concept itself of copyright
at a more basic level than the erudite and politically oriented famous work of
Walter Benjamin, The work of art in the age of mechanical reproduction. It puts in evidence the economic-political core of the
creation of copyright, while it can be questioned in spiritual terms as it was
in earlier times e.g. among musicians who freely “borrowed” and lent from each
other and from themselves, justifying it by seeing that their own imagination
and creativity was a gift from God. Or, as I heard once the world-famous tenor Plácido Domingo
respond to somebody in a broadcasted interview who praised his voice, by
uttering that it was not his own merit but a loan from God who could at any
time take it back.
I
myself come to think of my own indignation and feeling of revolt in that my
mentor West Churchman’s mentioned book The
Design of Inquiring Systems (1971) is not being reprinted and is not made
available on the Internet, while a direct request for it
had not even been answered (as per January 2024) by the owners
of its copyright. It is an eloquent example of the unethical disinterest for
disclosing truth, while also negating the supposedly “holy” democratic right of
freedom of expression, all under the silent mantle of narrow political-economic
(dis)interest.
Return to The Design of Inquiring Systems
Well,
the definition of wisdom is controversial. In way of summary: it all is about
chapters 2 and 5 of part 1 of DIS, while
ignoring or not understanding the whole of its 16 chapters in both parts 1 and
2 of the book. In particular, it is
worth mentioning that in AI and AGI the emphasis is on chapter 2, the building
of fact nets, and not on chapter 5 since the AI community is not specifically
designing and performing scientific experiments and creating or establishing
facts, but only selecting them, in oblivion of the intellectual challenge
illustrated in Steven
Shapin's famous A Social History of Truth. This is a phenomenon that I observed during many years
of teaching at universities: most readers understand the book up to chapter 4
or chapter 5 discussing empiricism, but later they seem to get gradually lost,
up to the point of confessing that they cannot see the necessity and meaning of
the second part of the book with the last 7 (where appear the references to
Carl Jung and analytical psychology) out of its 16 chapters. For many the final
result is that when discussing the book’s content, they testify that yes, they
have read it, and that they already “know” it, assuming that it does not
influence their pre-conceptions on the AI-issues, and their conclusions.
This
is not to “divinize” West Churchman who, in my opinion, got limited by his need
to survive academically by fitting in the USA academia
(also here),
and for reasons of the problematic relation between science, theology and
politics that I survey in my text on Information and Theology. I think that he basically did it by his repeated
relating to the academically divinized philosopher Immanuel Kant that I comment
upon in my
essay on the ongoing computerization of society. Kant also
deserves and important place in Churchman’s latest major book, The Systems Approach and its Enemies, in
chap. 4 on “Logic: A Theory of Reality, or Kant for Planners”. In my view, Kant
along his numerous followers such as Jürgen Habermas, diverted the development
of many scholars, including one among Churchman’s most knowledgeable and
faithful students, Werner
Ulrich who,
by the way, authored the best biographical documentation of his thesis advisor,
accounted for in the summary of
Churchman’s influence on me.
On the basis of
the above I wish for the time being to emphasize one basic conclusion. The
possible if not probable result of the use of AGI is that it will suffocate
debate by cementing the past and what
(and where) has been written or said up to now, both in terms of choice of, or
availability of data (measured and created, or chosen by, and available to, or
affordable for whom?), imposed way of thinking (as illustrated in my Logic
as Rape) and
of affordable technological base. An alternate simple formulation found in an
article in The
Economist (Nov
9th 2023) in
the context of Hollywood actors’ strike reflecting a fear of the technology is
The process of ingesting everything and then spitting
out an average may lead to a stylistic and conceptual regression to the mean,
says a literary agent, who sees similarities with the algorithms on social
media that help propagate the most pedestrian views.
AGI may work for contexts of basic natural science
isolated from human interaction, as is the case of established classical
physics such as in astronomy (but even then, not necessarily in quantum physics). It
will cement what has existed up to now around us or around those who supply or
have control over the environment that has produced the data for unknown uses,
and have control over the equipment. It will even include logic manipulations
(deductions and inductions) in e.g. large language models LLM (soon "very
large language models" VLLM?) of what is electronically recorded as having
been said, written or heard up to a certain moment about "something"
with no agreed-upon definitions, and finally selected by unknown GPT
"trainers" for known or indeterminate uses. It will keep lots of
people busy in a difficult rebutting of what has been easily said or written by
anybody with the help of AGI-GPT. It may be seen as a loudspeaker or broadcaster
of it all to all of us, or to selected audiences. All this coupled to the
difficulties if not impossibility of genuine debates as I try to survey in a
text on Information
and Debate. From
the teachings of the most serious research on the history and sociology of
science, one is neglecting every insight into the fact that most of the ongoing
scientific work is based on confidence in the truth of others that from now on
will not be known, as masterly illustrated by earlier mentioned Steven Shapin
in The Social History of Truth. It is
an image that revives the mind-blowing concerns raised already in 1967 by Russell Ackoff’s
article on Management
Misinformation Systems, which
also clarifies the pitfalls of offering ready-made "recipes" on how
to solve known problems using known methods:
Five assumptions commonly made by
designers of management information systems are identified. It is argued that
these are not justified in many (if not most) cases and hence lead to major
deficiencies in the resulting systems. These assumptions are: (1) the critical
deficiency under which most managers operate is the lack of relevant
information, (2) the manager needs the information he wants, (3) if a manager
has the information he needs his decision milking will improve, (4) better
communication between managers improves organizational performance, and (5) a
manager does not have to understand how his information system works, only how
to use it. To overcome these assumptions and the deficiencies which result from
them, a management information system should be imbedded in a management control
system.
DIS
deals just with what is or should be management, control and system. In terms
of DIS, AGI that in the best case may
work as a profitable "user manual", implies increased dependence upon
available data that unknowingly may be fake data, misinformation and disinformation, and upon computer technology, including dependence
upon those who own and can afford the data and technology. In other words: DIS
chapters 2 and 4 on Leibnizian
Inquiring Systems: Fact Nets,
chapter 3 and chapter 5 on Lockean Inquiring
Systems: Consensus, which
probably is the last chapter understood by most readers, and is the coupling to their blind reliance upon
the “myth of democracy“ (more on this myth below and in my essay
on the Russia-Ukraine conflict). This will be to the detriment of partly
chapter 6 on Kantian
Inquiring Systems: Representations which is the dogma of most
contemporaneous philosophers, chapter
8 on Hegelian
Inquiring Systems: Dialectic - the last chapter that is
usually understood by most labor unions, Marxists and leftists, and chapter 9 on Singerian
Inquiring Systems: Measurement and Progress – unconsciously accepted if yet not understood in
detail by most scientists. Plus the
whole part 2 with its 7 last chapters starting with chapter 10 on the three
basic models or imageries of inquiring systems. It implies emphasis on Democritean
Imagery or Mechanism, and Carneadean
Imagery: Probability (Statistics) to the detriment of Aristotelian
Imagery: Teleology or
goals, which is already
applied earlier in chapter 3 on the teleology of The Anatomy of Goal Seeking, and I would outright relate to Aquinian
theology with the consequent teleology and (analytical) psychology. The
rest of part 2 introduces the problem of implementation
recalling the question of The
Researcher and the Manager, plus the question of religion, theology, psychology and faith, the challenges to
teleology, and the dimensions of inquiry that are developed in the next
published book The
Systems Approach and its Enemies.
In still other words: it is the problem that I try to
approach in my earlier essays on the computerization of society and the human-computer interaction. One of the late suggestive reports is the case of an
insurance company that trained advanced text chatbots to initially reject the insured clients’ written
requests for compensation based on their particular insurance policy. It is
thought provoking to imagine what this way of thinking means for the medical safety of patients to be in the future served by commercial medical AGI chatbots. This
trend has already started and has been commented in Is
ChatGPT Smarter than a Primary Care Physician? without a visible understanding of which is the basic
problem, besides moral and juridical “responsibility”, beyond an initial
comparison between the performance by ChatGPT and human clinical practice. Or,
in a medical report on This
Drug works, but Wait till You Hear What’s in it, without an understanding of the difference between
statistics and a scientific theoretical “Why”, as explained by Churchman in the
book Prediction
and Optimal Decision (chap
6 on “Objective Probability” relative to statistical
population or group). I know of
statisticians who made up their living practically by writing physicians’
doctoral theses, since such theses consisted mainly of statistical work with
their available data. This is as medicine tends to become more and more result
of statistics on big data, and correlation a substitute of problematic
causation when the concept of cause itself is put in question as in advanced
quantum physics and philosophy of science. The whole is further complicated by
physicians themselves, e.g. in health care centers, giving in for smart
AI-sellers’ offers of “amplification” or computer “support” (that I comment in
the context of human-computer interaction). They will illude themselves that they retain a
responsibility which in fact can be no longer their own, being simply victims
of the good old Hawthorne effect, seduced as they get for being the focus of attention
and by the novelty of being research subjects of fashionable AI. All this while
the first symptoms of AI-abuse in health care are being reported (in Medscape, November 20, 2023) in Accused
of Using AI to Skirt Doctor’s Orders, Deny Claims.
It is obvious that in today’s scientific and cultural
milieu readers of these lines feel frustration in face of the perception of an
“enormous impossible” task of understanding all the above references as related
to the scientific and technical presently “dogmatic” bases of western
civilization. I claim that this stands at the heart of what I sense as being a
western cultural crisis. The crisis of understanding the bases of our own
western culture in analogy to the task of understanding Christianity as
compared to, say, Hinduism, Islam or Confucianism. Nevertheless, there are
people who do not shy away from such challenges. An example is the Swedish
intellectual Ellen Key who in launching in Sweden a sort of “enlightened feminism” did not shy away from embarking in a study of
Theosophy related to Buddhism and Anthroposophy. I also heard of students who
meet analog problems as the just mentioned reference to Theosophy and Buddhism
when envisioning critical hospital care by computer aided dying of patients
through the study of the relationship between Heidegger,
death, and technology in Japanese philosophy and engineering. More specifically
in a field known as kansei engineering, accounting for "emotions, affects, sensibility or “perception”
(whatever they happen to be or happen to interrelate) in line with traditional
Buddhist and Shinto philosophy, while in fact ignoring the whole body of
analytical psychology as I explain in the above mentioned criticism of Key’s
enlightened feminism.
I do not assume that e.g. feminism needs to be a much
more complex phenomenon than the ongoing computerization of society. But I have
already tried to explain (cf. above) why writing and discussing serious matters
is in general a hopeless task today. Regardless
the paradox of my writing down this just now, repeating then the error
committed by Churchman when writing and publishing the DIS to which I referred above. Because of all this I rely upon some
excerpts from some of my earlier writings, starting from my article on The Russia-NATO-Ukraine Information Crisis where
I quote what Plato, before or "as if" presaging the advent
of Christianity and of the Bible, wrote (as in my available translations) in
his famous Seventh Letter (341d, and 344c):
|
If I thought it possible to deal adequately with the
subject in a treatise or a lecture for the general public, what finer
achievement would there have been in my life than to write a work of great
benefit to mankind and to bring the nature of things to light for all men? I
do not, however, think the attempt to tell mankind of these matters a good
thing, except in the case of some few who are capable of discovering the
truth for themselves with a little guidance. In the case of the rest to do so
would excite in some an unjustified contempt in a thoroughly offensive
fashion, in others certain lofty and vain hopes, as if they had acquired some
awesome lore. […] |
|
For this reason no serious man will ever think of
writing about serious realities for the general public so as to make them a
prey to envy and perplexity. |
This
insight is summarized by the Latin locution Intelligenti pauca (“few words suffice for him who understands”), as well
as by the point in my essay Information and Debate on
the difficulty if not meaninglessness of a supposedly democratic argumentation
on deeper issues. In view of the Quality of Information. Its
meaning, however, goes deeper in the Bible as it can be seen in the Bible
Gateway by typing in the search field e.g. “whoever
has ears”, and “ears
eyes”, my own main choice being Matthew 13:1-23,
Lucas 8:4-15,
Mark 8:18, and
Acts 28:27. I
realize that my writing this text then is a paradox, and it all is an answer to
what is the conclusion of this paper for all those who do not want or cannot
read and understand the linked references. They might have been superfluous if
the population practiced some of the great religions, for us in the West
beginning with Christianity. It has been said that if people had followed at
least five or six out of the ten Commandments, the world’s situation would have
substantially improved, with or without AI-AGI.
The
reader who feels somehow impatient because of the length of this paper may now
jump over the following chapter 6 that contains occasionally commented (in bold font) further
explanations and examples of problems of AI in earlier papers of mine, going
directly to chapter 7 on The open letter:
"Pause Giant AI Experiments".
6. References
to ChatGPT in other essays
In
what follows below I try to offset the hopelessness of discussing matters and
references that are difficult to read and understand, the more difficult in
lack of motivation. I do this by means of (sometimes heavily edited) excerpts
from texts that I have written as per July 2023, which when necessary are
interpolated by my specific comments in bold text style. They will be completed
with my comments to answers obtained from ChatGPT to some specific questions of
mine. But let me start with a somewhat "lofty" section out of Information and Theology that matches
the probable theological meaning of the Plato's quotation above:
"Myths" of Artificial Intelligence. The explaining away of religion and Chesterton's "believing in
anything" also opens the way for believing in the future interaction with
other planetary worlds in outer space, and in artificial intelligence, AI,
computers with self-consciousness, whatever it is, will overpower the human
mind, or that AI-robots should have human rights, or that we will ultimately
create the superintelligent robotical paradise on Earth, or whatever wherever.
Or believing in discussions about, say - machine consciousness and machine ethics ending in so called technological singularity.
All this without the possibility of "debating" the matter
because of myriads of misunderstandings arising in part from faulty difficult
definitions and presuppositions, to begin with about the meaning of
intelligence. Buchanan's initial consideration says it all for those who
can understand: "In calling AI an experimental
science, we presuppose the ability to perform controlled experiments
involving intelligent behavior." I repeat, for emphasis: "perform controlled experiments involving intelligent behavior". I add: where
control is possible and advisable, as discussed by West Churchman in his book
that followed DIS, on The Systems Approach and its Enemies, especially
the chapter on "Logic: General", followed by "Logic. A theory of
reality, or Kant for planners." But who cares? "Please give me a few
lines' summary because I have not the time to read all this stuff"…
This phenomenon of misunderstandings and phantasies about AI is
revealed and, in a sense also "solved" in all its
complexity by what I regard as a fundamental work by Churchman by the time we elders met the first wave of hype-enthusiasm
about AI in the seventies. It was The Design of Inquiring Systems:
Basic Concepts of Systems and Organization that I myself tried to
expand and facilitate the reading of, by means of a Word and Issue Index, followed by a sort of contextual evaluation in The Systems Approach to Design and Inquiring Information Systems.
I think that with this kind of understanding it is not, anymore, a
question of whether AI in its many forms will be applied in modern society. It
is rather a question of forecasting the consequences and the possibilities of
counteracting the dangerous ones, becoming a problem that I considered in
my Trends in Philosophy of Technology,
and ultimately a theological problem that motivates the present text. The
difficulties will be enormous, not only because the academic devaluation of
theology, and even philosophy in technical and political context. Even when a
professor of computer science warns about overconfidence in AI in Swedish mass media (Dagens
Nyheter, October 7th 2018), he relies upon exhortations for the need
to be conscious about the system's limitations. The
warnings are based upon appeals to understand and to be conscious that we are still far from creating (an
undefined) intelligence at a human level with
the ability to feel and reason, evaluate, make moral
evaluations and explain what it is doing and planning
to do. All italicized concepts remain undefined, presupposing political,
philosophical and theological competence, understanding why we are "still" far
from "creating" artificial life and paradise on earth, understanding
the why of not to "believe in anything".
Even a most sophisticated Italian mathematician, Paolo Zellini, who
dedicated much of his life writing about the philosophy of mathematics
including computer science, concludes his work with a rather inconsequential
book, so far only in Italian language, that vaguely warns about the Dictatorship of calculus. His barely outspoken warnings are supported by reference to the
extremely explicit ones by the more popular sort of polymath Jaron
Lanier. Lanier's limitations appear most
clearly when he introduces also provocative thoughts on virtual-reality that
challenge earlier elaborate condemnation by
others as being deleterious gnosticism in computer science. Despite positive ambitious reviews, Zellini's neglect of theology, particularly of Christianity, leads him to
miss the most relevant historical aspects of the contribution by Jan Brouwer to
the understanding of the problems considered here.
It is symptomatic that when we humans no longer believe in God, we
happen to believe that we (Nietzschean superhumans) are so godlike as to be
able to create machines in the track of ChatGPT that will transcend human
intelligence and be substitutes for God.
My comment: This excerpt illustrates how
the hype of AI and AGI may nourish itself upon the secular abandonment of the
idea of GOD and related theology.
All this while the modern technological
mindset can be envisaged as promoting the deconstruction or destruction of the
(cf. Martin Buber) “you-thou” by reducing it to a divinized capitalized “It” as
in the hyped “dialogue” of humans with a ChatGPT using Large
Language Model tools – LLMs (see also here). Language models are basically logic, and further: mathematical logic
that forgets the observation by the Jan Brouwer mentioned above (treated in my essay here) in his study of the foundations of mathematics and logic, that logic
was to begin with extracted from language. It further opened to the
tautological consequence of the ambitions of mathematize logic into "mathematical
logic" in order to finally being used as
substitute for language conversation between humans that logic was originally
extracted from. Humans then gradually leave the scene as unemployed, and the
original language conversations and negotiations turn into authoritarian LLM
declarative sentences sent through computer networks by a few to the many,
plunging everything into the problematic "human-computer
interaction".
It is the kind of chat that
produces instances in Japan where individuals are reported to have fallen
in love or having developed so-called romantic feelings for (holograms of) contrasexual (cf. LGBT), virtual or fictional
characters, sometimes referred to as "2D lovers"
or "waifu culture" (commented in depth by The New York Times, July 26, 2009). It all to be related to the concepts of “3D
lovers” and Real
Doll. This
phenomenon gained some attention in the media, particularly with the emergence
of vocaloid characters like Hatsune
Miku in a perhaps more permissive
Japanese culture permeated by Shintoism. More information about the topic can be obtained in connection with
the Nijikon and Hentai phenomena as well as fictosexuality and pansexuality, which paradoxically indicate the deep gender-psychological and thereby
theological implications of human-computer
interaction.
Such detours of so-called human “sexuality” stand at the basis of the
undermining of the fundamental “I – You” relation whose “archetype” is the
gender relation between humans represented by man and woman. This undermining
implies the sliding of the You into an It and
a consequent confusion of pronouns that are illustrated by what happens lately
in the computer field of human-computer interaction or HCI with
the appearance of the device ChatGPT mentioned above. In fact, when
I inputted into the machine the question “Please
explain which are the basic databanks, whether including the whole Internet,
which are used in order to find and elaborate the answers in ChatGPT”, I
initiated my question with a “please” felt to be directed to those They (or
a He or She?!) who are (or were?!) “responsible” for its operation.
But I got the following answer from an I that could have been
a We but was an It, as follows (my emphasis):
As an AI language
model, my responses are generated based on a mixture of
licensed data, data created by human trainers, and publicly available data.
I don't have direct access to databases or retrieve information from
specific sources in real-time. I have been trained on a
diverse range of data, including books, websites, and other texts, to develop a
broad understanding of human language. My purpose is to assist
and provide information to the best of my abilities, regardless of the specific
sources of my training data.
This is, of course, the problem of responsibility in
human interaction that is today exasperated also in the introduction of the
“autonomous” technology represented by the ramblings of so-called artificial general intelligence AGI where
no longer is questioned what intelligence is or should be.
My comment: All this highlights the
question of “WHO” is ChatGPT or whatever future actor in the expanded field of
Artificial General Intelligence. It is closely related to the question of
"Whose responsibility", which in turn tends to be easily transformed
into an essentially juridical law-question (as Portuguese-speaking readers can
hear at an interview with Brazilian computer scientist. Silvio Meira, here and here), while the source of the problem is
found at the basis of what fact, logic, and intelligence is all about as
related to ethics. Today there are (too) many brilliant minds (as I show in my essay on computerization) who are tempted to discuss, in an
extremely logically structured way, matters that they should understand better,
recalling again Plato's quote above. In general, the problems created by AI and
AGI are translated into assumed future solutions designed by the process of
assumed ideal democracy. For the rest It is unclear to me what the last
sentence means with “regardless the specific sources of my training data. This
considering what was reported by Reuters (Oct 17, 2023) regarding China’s Ernie 4.0 generative
AI model (China’s ChatGPT), stating that the prior week, Beijing had published proposed security requirements for firms offering services powered by
the technology, including a blacklist of sources that cannot be used to train
AI models, beyond copyrighted material. I have seen nothing having been said
about that kind of things in neither ChatGPT, nor its contenders Google’s
chatt-AI Bard and
Gemini. Symptomatically, Wikipedia’s article on Gemini quotes a couple of professors who in
the context of comparing these AI-products “cautioned that it was difficult to
interpret benchmarks scores without insight into the training data used”. All
this while extensive censorship was exerted on the news about e.g. the Russia-Ukraine
conflict, as remarked in my paper on it. But Fei-Fei Li (author of The
worlds I see – curiosity, exploration and discovery at the dawn of AI) at the Stanford Institute for
Human-Centered Artificial Intelligence (Stanford HAI) is reported in a
newspaper (Svenska
Dagbladet,
December 9, 2023) to “emphasizes the importance of an ethical AI development”. I am sure
that many perceive such entrepreneurial wording as well-meant uncommitted
buzzwords, as the title of the newspaper article, “AI is controlled by humans -
for better or worse”, identical to similar stuff found on the net. The involved dazzling successful
personalities, however, as well Jensen (Jen-Hsun) Huang co-founder and president of Nvidia Corporation, certainly deserve to be listed among the computer geniuses I introduce in my above-mentioned paper on Computerization.
From: Computerization
as Design of Logic Acrobatics
And today the ultimate consequence of this thinking is the recent and by
now famous ChatGPT, which probably few computer scientists seem as yet to understand what
they are doing in terms of logical fact nets or Leibnizian Inquiring Systems. They are basically logical operations
performed on the contents of the assumed “facts”, including facts and about
functions of devices that are stored and retrievable from the total
Internet.
I
think that those who feel compelled by their conscience to unravel this tragic
mental confusion can only do this by relating computer science to mathematical
logic and empiricism, or logical
empiricism, along the guidelines outlined by West Churchman in
his The Design of Inquiring Systems repeatedly
mentioned here, for good reasons. In
particular it is a question of what follows from the conceptions of chap. 2 on Leibnizian Inquiring System: Fact nets
(logical networks), and chap. 5 on Lockean
Inquiring Systems: Consensus. That is, consensus within the community of
Pre-Trainers (the PT in GPT, people selected by somebody), which establishes
the sources and selection of facts to be or not to be networked and
Generatively Transformed. The former fact nets are intertwined linguistic
manipulations, and the latter consensus is the manipulated factual
sentences, declarative
knowledge that in turn will trigger physical devices when it
all happens to be politically trusted wherever found in the Internet by those
who can afford the search for their own entertainment, profit or purposes. And
the first purposes unfortunately, and
ultimately, will be the purposes of automated warfare, and “killer
robots” or lethal autonomous weapons (see also here)
that will be certainly called "defensive". That is the core of GPT
& Co, which is supposed to revolutionize and save or at least to win the
world, it all seasoned with more or less intellectually and politically naïve
warnings for “Existential
risk from artificial general intelligence”. Whatever
existential means in its relation to religious.
Not knowing or understanding what intelligence is or should be, the door
is open for limitless fantasies about the next if not ultimate hype of Artificial General Intelligence including
the construction of the above mentioned Theory of everything (and
its philosophical counterpart here). Therefore, many will start
hoping not only to be able to ask any question and get the answer, but also to ask the computer to do anything and having it done, such as trusting self-driving
vehicles without even having understood the
problems of auto-pilots or problems of scientific experimentation and theory-building, or
technocratic planning being replaced by a sort of ex-post pre-training as in the elaborations of GP - Generative
Pre-Training. Whose
training and responsibility? In terms of the conflict between Russia and Ukraine it is like asking ChatGPT or Bing what to do in order to solve and stop the conflict by applying a
thinking that I have provocatively and rhetorically called Logic as Rape in the form of a ChatGPT instead of an inquiring system. Cf. West Churchman’s
“Logic: A Theory of Reality, or Kant for Planners” in his earlier mentioned
book The Systems Approach and its Enemies (chap.
IV). In other words, a revival of the fantasies about the HAL 9000 fictional AI “character” or of the older idea of Frankenstein ‘s
monster.
All this without understanding what it all is about, as illustrated in
Jordan Peterson’’s short video on ChatGPT with the title of The dark side of AI. If one has the time
it is possible to extend the experience by comparing it with Chris Hrapsky’s
video on The limits of ChatGPT and discovering a dark side. It
is not a question of increasingly faulty intelligence as suggested in an
ambitious essay: The age of stupidity.
(It is written in Portuguese, but with one main reference in English about The negative Flynn effect). It is more than so, it is a gradually increasing cultural crisis,
affecting the population’s intellect, becoming more visible in the affluent
western world where it is not masked by material privations. Ultimately, we can
consider interaction in terms of the sort of “archetype” of interaction
mentioned in my essays on Reason and Gender and Computerization. A CNN news report on a related event, The man who married a hologram related
to the phenomenon of Nijikon may give a hint of what is to come, stating that “Researchers say such events are
indicative of broader technological trends and social phenomena”.
The late and perhaps ultimate consequence of short-circuiting the human
and social element together with the consequent debate in the increasingly
inclusive logical processes is the expanding phenomenon of polarization of
parties, misunderstandings, conflict and violence. They could follow from the
ChatGPT’s sociopsychological choking of debate by forcing strong secularized Lockean consensus that was mentioned
above. In its bland realistic and immediate form it may be disinformation and ransomeware and, in particular, Ryuk under the general labels of Cyberattacks and its counterpart Computer
security.
My comment: Still worse it may be if the
information is perceived as so overwhelming, the more so when it is much more
difficult to question and demonstrate that a declarative sentence is wrong
(cost of measurement and gathering of data, hidden definitions and
environmental systemic conditions) than to just formulate a supposed factual
reality upon the basis of available collected data. I illustrate something
related to this in my essay on Information
and Debate with the rhetorical questions of “Why Not?” (exemplified with divorce,
homosexuality, polygamy, etc.).
The “apocalypse” of debate is pre-announced in the development of
artificially intelligent “chat-bots” that are
envisaged to allow anyone, in a sick society, to discuss with them, as
announced by META-Facebook on August 5, 2022
regarding “BlenderBot 3: An AI Chatbot that Improves Through Conversation”.
What is most symptomatic is its forgoing to mention the possibility of
religion being a necessary basis of meaningful consensus: charity and respect
for the human being, where respect does not mean “tolerance” in the sense it
does not matter what somebody thinks so long it does not bother me or
"us". It means that we all want to help each other to reach a common
truth and good. One opportunity for getting convinced about this is to study
very carefully the most ingenious analysis I have seen of the breakdown of
debate vs. political correctness, or whatever you want to call it, in Ian
Buruma's bewildering article, related to the #MeToo issue, "Editing in an age of outrage", in Financial Times, March 29th 2019. (Swedish translation "Det ska fan vara publicist när Twitterstormen viner" in Dagens Nyheter, April 14th 2019.) Ultimately, in the future, in analogy to the problem of self-driving vehicles, people may have to discuss mindless texts that are generated by
artificially intelligent agents such as improved GPT-2:
"GPT-2
generates synthetic text samples in response to the model being primed with an
arbitrary input. The model is chameleon-like—it adapts to the style and content
of the conditioning text. This allows the user to generate realistic and
coherent continuations about a topic of their choosing..."
and will be drained of their intellectual energy as human drivers will
be drained of their skill and intelligence in order to avoid incidents with the
self-driving vehicles. [Test a simplified GPT-2
here, linked at from here.]
In
order to limit the volume of text in this article, this case study is presented
in its original form, as a separate article on Information
on Christianism and Atheism. In it I make the observation that discussions on religion vs. rationality can be seen as
a poor reenactment and analogy of the misunderstood and endless theological
debate on the Filioque, subject of great controversy between Eastern and Western Christianity that
serendipitously deals with the meaning of Spirit. where today the most analytically or rhetorically
gifted and persevering party seems to win the game. Here I can specify that my
choice of the adverb serendipitously referring to the discussion of the
meaning of Spirit can be relevant for the ongoing debate
on Artificial Intelligence and especially Artificial General Intelligence –
AI/AGI, if intelligence is related as it should to the discussion of the human
intellect. It seems to me absurd to imagine that a deep-going “world-wide”
discussion on the Filioque among educated intellectuals should
have been going on for centuries without having a bearing on other intellectual
questions and on how to think and behave.
My comment:
What will happen with the lawful responsibilities in using self-driving
vehicles, or automated weapons, and the labor market upheavals with the
introduction of AI and AGI in industry, commerce and government, is or will
make its appearance in the lame warning gathered under the Wikipedia title of Existential Risks from Artificial General
Intelligence.
From: Computers as
Embodied Mathematics and Logic
The reports in The Economist and Wikipedia(should) say it all, reminding also
cases of analog analytical logical-mathematical giftedness such as of Adam
Neumann and the related business WeWork,
and Sam
Altman as related to general artificial intelligence (Open
AI, ChatGPT, etc. as per April
2023) and its financing, or sheer analytical financial genius of a Bernard
Madoff.
[plus the following that were included in the essay’s later versions]
My own annotations on my copy of Carl Jung's book Memories,
Dreams, Reflections recorded
by his associate Aniela
Jaffé, indicate that I first bought and read it
in April 1975 but happened to definitely relate it to this essay of mine in
July 2023. The book is interesting especially because it tells about the
author's psychological interpretation of his understanding and feelings in
meeting mathematics. Such feelings, when experienced by common educated people
can easily be dismissed for being caused by lack of giftedness or sheer
stupidity. The complexity of Jung's account requires literal quotations (cf.
pp. 43 ff., 340 ff):
[…]
I felt a downright fear of the mathematics class. The teacher pretended that
algebra was a perfectly natural affair, to be taken for granted, whereas I
didn't even know what numbers really were […] they were nothing that could be
imagined, mere quantities that resulted from counting. To my confusion these
quantities were now represented by letters, which signified sounds, so that it
became possible to hear them, so to speak. […] No one could tell me what
numbers were, and I was unable even to formulate the question. […] But the
things that exasperated me most of all was the proposition: If a=b and b=c,
then a=c. even though by
definition a meant something other than b, and, being different, could
therefore not be equated with b,
let alone with c. Whenever
it was a question of equivalence, then it was said that a=a, b=b, and so on.
This I could accept, whereas a=b seemed to me a downright lie or a fraud. I was
equally outraged when the teacher stated in the teeth of his own definition of
parallel lines that they met at infinity. […] My intellectual morality fought
against these whimsical inconsistencies, which have forever debarred me from
understanding mathematics. Right into old age I have had the incorrigible
feeling that if, like my schoolmates, I could have accepted without struggle
the proposition that a=b, or that sun=moon, dog=cat, then
mathematics might have fooled me endlessly – just how much I only began to realize at
the age of eighty-four. All my life it remained a puzzle to me why it was that
I never managed to get my bearings in mathematics when there was no doubt that
I could calculate properly. Least of all did I understand my own moral doubts concerning mathematics.
In a later section of the book, in the chapter "On life after death",
Jung introduced some exceedingly complex ideas about the relation of human
death to the realm of consciousness vs. the unconscious or "the
intermediate stage between unconscious and conscious cognition". It would
take us too far in this our context but the following extracted comment (p.
341) may give a taste of the relation to mathematics:
Ideas
of this sort are, of course, inaccurate, and give a strong picture, like a body
projected on a plane or, conversely, like the construction of a four-dimensional
model out of a three-dimensional body. They use the terms of a
three-dimensional world in order to represent themselves to us. Mathematics
goes to great pains to create expressions for relationships which pass
empirical comprehension. In much the same way, it is all-important for a
discipline to build up images of intangibles by logical principles and on the
basis of empirical data, that is, on the evidence of dreams. The method
employed is what I have called "the method of the necessary statement".
It represents the principle of amplification in the interpretation of dreams,
but can most easily be demonstrated by the statements implicit in simple whole
numbers.
My point with these quotations is to suggest that mathematics because of
its very nature introduces unconscious processes in the human mind while
surreptitiously merging them with ordinary conscious ones. It all recalls what
has been considered in this essay about both the aforementioned Brouwer and
Zellini, as well giving a hint of the ultimate mysteries of quantum physics
(more on this above and below) in the relation between psyche and matter. An
alternative equivalent conception presented in my Information and Theology is
the balance between the inner knowledge of human psyche vs. natural science and
knowledge of the outer world. This role of the unconscious and inner knowledge
of the human psyche related to (the aesthetics of) mathematics and logic also
explains the problems of addiction (not only of children, but also of
“childish” adults), including computer
addiction, the overhype of virtual reality, and the success of mathematical-logical minds in all sorts of modern
endeavors including mathematized technoscience or white
collar crime. The latter is most brilliantly
exemplified by the genius of Bernard Madoff and by the fact that “Internet AI personalities” such as Tom Bilyeu at Impact Theory and Emad
Mostaque: see here a logical conversation between them with the ambitiously alarming title
How Ai will disrupt
the entire world in 3 years - Prepare now while others panic. The latter have strong opinions about AI
there is a penchant for bold entrepreneurship and/or interest in “volatile”
activities, financial and others, as hedge fronds. And much entrepreneurship with empty talk
about AI may be conscious or unconscious attempt to commit “white collar
crime”. Who is capable to discuss the degree of consciousness vs.
unconsciousness in order to determine culpability and whether it will be considered
as financial
crime?
The technical mind has been colorfully exposed in studies of the
philosophy of science in the tradition of Jacques Ellul. Some of the deep approaches in the latter’s “anarchist” tradition
(examples here, here, and here), however, have sometimes missed the point when they lose the
connection to broad Christian theology, philosophy and its derivate (Carl
Jung’s analytical) psychology. This appears in the following (below) de-contextualized
excerpt
from one of the deep approaches to Cybergnosticism, (a
term further explained here). The text of the excerpt her
below is followed sometimes by my own notes in italics in square brackets in order
to not leave everything to the general “Comment” at the end:
[J]ust
like alcoholics and drug addicts, game-addicted children are now regularly
treated by psychologists and psychiatrists in order to get rid of their
addictive behaviour, [Yes, but how
effective in this treatment on what grounds, and why, since it may by then used
for treating all “addicts”, and how should they be defined and identified,
since they may be even scientists?]
[M]any
videogames take advantage of techniques similar to those used by the military
to harden people emotionally. [Yes, but
what kind of the treatment of these hardened people would be allowed and
recommended by the military leadership vs. ex-post by e.g. the “Veteran benefits fof
PTSD”
or equivalent organizations?]
As the driving
force of la Technique is, according to Ellul, the crave for absolute efficiency
in all human endeavours. [Ellul himself
is quoted in Wikipedia from his book “The Technological Society” as writing: “Not even the moral conversion of the technicians could make a
difference. At best, they would cease to be good technicians. In the end, technique
has only one principle, efficient ordering. What about the difficulty, to begin with, in defining “efficiency”
vs. “effectiveness”, the more so when it is not even recognized that the difference lies in
the conceptualization of the systemic boundaries of (whose?) craves.]
It
was Eric Voegelin’s intriguing and much-debated thesis, that there is a
deep-seated disorder in our civilisation rooted in a ‘gnostic’ sentiment of
alienation and discontent with reality perceived as evil, in the con-
sequential ‘gnostic’ turn away from this reality [Psychologically, or how else, is a “sentiment” to be defined, compared
with, say a thought, intuition or perception?]
Arguably,
we cannot get at the real motives and ideas behind the computer phenomenon in
general, and the cyberspace and virtual reality sub-phenomena in particular,
nor arrive at a proper understanding of their roots and future direction of
growth, unless we take into account these mighty metaphysical driving forces
and motivations […] [How about
metaphysical forces and motivations, if not within the frame of a religion’s
theology?]
It
is my thesis that the roots of cyberspace and cybergaming must be investigated
in a much wider context than is done in these and other similar works so as to
clarify and make comprehensible the motive background and worldviews of the key
personages of the field. [What about the
difference between (here) “roots” vs. “forces and motivations” in the previous
paragraphs].
They rest ultimately only upon the criticism by Gnosticism and overworked erudition and rationalism (paradoxically, an apparently
own “gnosticism”) of the one “political philosopher” Eric Voegelin. Wikpedia
summarizes, as it follows some of his critics’ opinions, which match my own
impression from my readings and comment of Voegelin in another context of Information and Theology:
[C]ritics have proposed that Voegelin's concept of
Gnosis lacks theoretical precision. Therefore, Voegelin's gnosis can, according to the criticis,
hardly serve as a scientific basis for an analysis of political movements.
Rather, the term "Gnosticism" as used by Voegelin is more of an
invective just as "when on the lowest level of propaganda, those who do
not conform with one's own opinion are smeared as communists.”
That is, “smeared as gnostics”, as I heard one of his followers smeared
Carl Jung, based on Voegelin’s readings of second hand texts on him (as
contrasted to e.g. Jordan Peterson’s practice as first hand reader of original texts). Until further
notice I assume that the Jung's attitude mentioned above is motivated by his
psychic inclination to watch upon the unique (psyche) instead of "playing
God" or suffering an “Ego inflation” by imagining oneself as understanding
mathematics seen as God’s language, and searching general "laws" of
nature where the human being himself tends to be “abstracted from” or is
regarded as only an object of an undefined or pantheistic "Nature". A
concretization of these thoughts is suggested by problems and the history of
debates on the world’s first societal introduction of personal
identification number - PIN in
Sweden (as
a number = “John Smith”), followed by national identification numbers all
over the world as I
explain in my essay on The
Meaning of Man-Machine Interaction. It is meaningful to remember the practice of
tattooing a number on the skin of prisoners in Nazi concentration camps,
followed today by the increasing use of PIN’s,
not the least for (who’s?) control of the population by means of face
recognition systems and biometric identifiers. All facilitated by alleging a struggle
against “terrorism”. And all this while never reflecting upon the meaning of or
relation between identification and individuation, and never
reflecting upon the phenomenon, in certain corrupt political regimes, of
governmental production of false “identification” documents, including vehicle registration plates. When faith
and allegiance is rested on an assumed Democracy and technology instead of on
God, then the unavoidable failures of democracy imply a dependence upon
anonymous human power whose policing of citizens offsets the loss of judgment
caused by the computerized technological logical bypass of the network of
multiple human judgments.
The extension of these problems reaches the import of the whole issue of
my text on the consequences of the computerization of society which follows the
conclusions of the present essay, and that I rhetorically named Computerization
as Design of Logic Acrobatics but somebody else could have called On the
Digitalization of Society. A concrete manifestation of the latest problems of computerization,
which at the same time illustrates the psychic derangements of mathematical
thought as suggested in Jung's account is the open letter Pause Giant AI Experiments: We call on all AI labs to immediately pause
for at least 6 months the training of AI systems more powerful than GPT-4. By August 2nd, 2023 when I accessed it, it had gathered more that 33000
signatures based on what I consider as a legitimate fear that is paradoxically
based on wrong premises. It reminds an analog "moratorium" that I
recall in my essay on Trends
in the Philosophy of Technology. They
are wrong "Leibnizian" premises as I explain in my paper on Artificial
General Intelligence and ChatGPT.
My comment: these name-related cases
illustrate the destructive power of minds that are powerful in terms of
(unbridled) rationalist analytical logical-mathematical capability, which reach
their psychic apex in the field of quantum physics, and most recently in the field of
computer science that embodies and rewards mathematics and logic. Jung's
difficulties illustrate a mind's struggle in accepting a way of mathematical
thinking that opens the road for abuse of human "Leibnizian" thought
when applied to matters that require a whole human mind. It is an abuse that
was not feared at the time of the polymath Leibniz himself (1646-1716) before the breakthrough of the
misunderstood, all-encompassing and mathematized natural science. The
mathematization of natural science was later extended to all science,
forgetting the purpose and meaning of divorcing thought from associated feeling
and intuition, and partly even from sensation by means of mathematical notations as historically conceived in
the work of Florian Cajori. In common life it, as knowledge of “God’s language”, can lie behind
tendencies toward Narcissistic
personality disorder, especially the empirically verified subtype High-Functioning/Exhibitionistic.
From: The
Russia-NATO-Ukraine Information Crisis
“[G]eneral
artificial intelligence” (GAI), in the form of the latest hype of ChatGPT is seen as a super-human (again,
“godly”) intelligence to which human democracy will politically have to submit
itself despite of cheap warnings for the need of (paradoxically) “democratic
control” in face of “existential risks of artificial intelligence”.
Or the assumed brilliant future of ChatGPT (Chat Generative Pre-trained
Transformer) that “was tricked to endorse the 2022 Russian
invasion of Ukraine”.
I myself wanted
to complete such possible criticism by using the latest (as per June 18, 2023)
hyped computerized tool ChatGPT – OpenAI (see also here, version at the
same date).
On
June 17, 2023 I put the following two questions to the available ChatGPT, and
got the two subsequent answers (below) from the site https://chat.openai.com. They put in
evidence the disregard of the core issues in my text and its references. Elsewhere I
mention why ChatGPT itself could not write my text. Its creators, unless they
only wanted to market a profitable product, should understand that neither AI
nor AGI
could have written this text since they instead should have been trying to
apply the earlier mentioned Design of
Inquiring Systems.
My
two questions and respective answers were:
(1) Which
are the causes of the conflict between Russia, NATO and Ukraine?
(2) How
to solve the conflict and achieve peace between Russia, NATO and Ukraine?
(1) Which are the causes of the conflict between
Russia, NATO and Ukraine?
ChatGPT:
The conflict between Russia, NATO, and
Ukraine has its roots in a combination of historical, political, and security
factors. While it is challenging to summarize all the causes comprehensively,
here are some key factors that have contributed to the conflict:
Historical context: The historical relationship between Russia and
Ukraine has been complex and intertwined. Ukraine was part of the Soviet Union
until its dissolution in 1991, and there are historical and cultural ties
between the two countries. The legacy of this shared history has shaped their
interactions and perceptions.
Ukrainian independence and alignment: After gaining independence, Ukraine pursued closer
ties with the West, including aspirations for European integration and possible
NATO membership. This shift in alignment has been viewed by Russia as a threat
to its sphere of influence, particularly in the context of its historical ties
to Ukraine.
Crimea annexation: In 2014, Russia annexed Crimea, a region that was
previously part of Ukraine. The move was widely condemned by the international
community, as it violated Ukraine's sovereignty and territorial integrity. This
event significantly escalated tensions between Russia and Ukraine, as well as
between Russia and NATO.
Eastern Ukraine conflict: Following the annexation of Crimea, pro-Russian
separatist movements emerged in eastern Ukraine, leading to an armed conflict
between Ukrainian government forces and separatist groups. The conflict has
resulted in thousands of deaths and a humanitarian crisis in the affected
regions.
Geopolitical competition: The conflict can also be seen within the broader
context of geopolitical competition between Russia and NATO. NATO's expansion
eastward, closer to Russia's borders, has raised concerns in Moscow about
encroachment and a perceived threat to its security. Russia has expressed
opposition to NATO's presence in countries it considers part of its sphere of
influence.
Ethnic and linguistic divisions: Ukraine is a diverse country with different ethnic
and linguistic groups. The divisions between ethnic Ukrainians and ethnic
Russians, particularly in eastern Ukraine, have played a role in exacerbating
tensions and fueling the conflict.
Energy dependence: The issue of energy dependence has added another
layer of complexity to the conflict. Ukraine serves as a transit route for
Russian gas supplies to Europe, and disputes over energy pricing and supply
have at times strained relations between the two countries.
It's
important to note that the causes of the conflict are multifaceted and
intertwined, and different perspectives exist on each side. The complexity of
the conflict underscores the need for a comprehensive and nuanced approach to
finding a peaceful resolution.
(2) How to solve
the conflict and achieve peace between Russia, NATO and Ukraine?
ChatGPT:
The conflict between Russia, NATO, and Ukraine is a
complex geopolitical issue with no simple solution. However, I can provide you
with some general principles and approaches that can contribute to the
resolution of conflicts and the achievement of peace. Please note that these
suggestions are not specific to this particular situation and may not address
all the intricacies of the conflict at hand. Here are some steps that could be
taken:
Diplomacy and dialogue:
Encouraging diplomatic efforts and open dialogue between all parties involved
is crucial. Diplomatic negotiations can help foster understanding, address
concerns, and find common ground.
Mediation and international involvement: Engaging neutral mediators or international organizations, such as the
United Nations or the Organization for Security and Cooperation in Europe
(OSCE), can facilitate negotiations and provide a platform for discussions.
Confidence-building measures: Implementing confidence-building measures can help create an atmosphere
of trust and reduce tensions. These can include the withdrawal of troops from
certain areas, establishing demilitarized zones, and implementing arms control
agreements.
Respect for international law and agreements: Upholding international law and respecting existing agreements, such as
the Minsk Protocol and the Budapest Memorandum, is essential. All parties
should be committed to honoring their obligations and finding peaceful
solutions within the framework of international law.
Addressing underlying grievances: Understanding and addressing the underlying grievances of all parties
involved is important for long-term peace. This may involve addressing
historical, cultural, economic, and security concerns through negotiation and
compromise.
Economic and humanitarian assistance: Providing economic and humanitarian aid to the affected regions can
help alleviate the suffering of the local population and create conditions for
stability. This can include support for reconstruction, development projects,
and assistance in meeting basic needs.
Public diplomacy and communication: Engaging in public diplomacy efforts can help shape public opinion and
build support for peaceful resolutions. Promoting accurate information,
fostering understanding, and countering misinformation are essential components
of this approach.
Long-term conflict resolution: Developing a comprehensive and sustainable plan for long-term conflict
resolution is crucial. This may involve addressing broader regional security
concerns, promoting economic integration, and fostering people-to-people
exchanges.
It is important
to recognize that the conflict is complex and deeply rooted, and the path to
peace will require sustained efforts from all parties involved. The suggestions
provided here serve as general guidelines and should be adapted to the specific
dynamics and challenges of the Russia-NATO-Ukraine conflict.
My
comment: this use and response of Chat-GPT illustrates some of the reasons for
why the ChatGPT could not have written the kind of text in my essay of the
Russia-NATO-Ukraine conflict. Even less so when it is not designed with the
intention of relating to the "Hegelian Inquiring Systems" mentioned
above. That is: it is not intended to use as much as possible the same data or
"facts" in order to also support a counter-story with different if
not opposite conclusions to those of the initial main story.
From a first specific question put to ChatGPT
My
question:
Please
explain which are the basic databanks, whether including the whole Internet,
which is used in order to find and elaborate the answers in GPT.
ChatGPT
As an AI language model, my responses are generated
based on a mixture of licensed data, data created by human trainers, and
publicly available data. I don't have direct access to databases or retrieve
information from specific sources in real-time. I have been trained on a
diverse range of data, including books, websites, and other texts, to develop a
broad understanding of human language. My purpose is to assist and provide
information to the best of my abilities, regardless of the specific sources of
my training data.
It's important
to note that while I strive to provide accurate and up-to-date information, my
responses may not always reflect the most current news events or developments.
News is constantly evolving, and new information can emerge that may change the
context or understanding of a given topic.
My comment:
This says it all. The “human trainers” are particular,
unknown people with their own (or the producing company’s) unknown ethical and
political convictions based on unknown facts. What they are implementing is a
sort of inductive process as described in Churchman’s Theory of Experimental
Inference but renamed “Machine learning”.
Its basic assumption, which underlines my observation of “cementation” is “that whatever worked in the past (i.e.
strategies, algorithms, and inferences) will most likely continue to work in
the future.” Unsaid: working for whom. See also Wikipedia’s list of “See also” with reference to,
among others, “Automated machine learning” and “Big Data”. See also “Large Language models” and “Artificial Neural Network” including its
extensive “See also”.
From a second specific question put to ChatGPT
My
question:
What
can the NATO do for achieving peace in the Russia-Ukraine conflict?
ChatGPT
NATO (North Atlantic Treaty Organization) plays a
significant role in addressing conflicts and promoting peace and stability
among its member states. While I can provide some general suggestions, please
note that the situation in the Russia-Ukraine conflict is constantly evolving,
and specific actions may depend on the current circumstances. Here are some
measures NATO can take to contribute to achieving peace in the Russia-Ukraine
conflict:
1.
Collective Defense: NATO's core principle
is collective defense, stating that an armed attack against one member is
considered an attack against all. NATO can continue to reaffirm its commitment
to the security of its member states, including Ukraine, and maintain a strong deterrent
posture to discourage any further aggression.
2.
Political Diplomacy: NATO can engage in
diplomatic efforts to encourage dialogue and negotiation between Russia and
Ukraine. This could involve providing a platform for discussions, offering
mediation services, or supporting international organizations and diplomatic
initiatives focused on finding a peaceful resolution.
3.
Support for Ukraine: NATO can provide
political, economic, and humanitarian support to Ukraine. This assistance can
include military aid, capacity-building programs, training and advising
Ukrainian armed forces, enhancing cybersecurity capabilities, and facilitating
defense reforms to strengthen Ukraine's ability to defend itself.
4.
Monitoring and Reporting: NATO can
continue monitoring the situation on the ground through intelligence gathering,
satellite imagery, and other means. By providing accurate and timely
information about the conflict, NATO can contribute to international
understanding and help counter disinformation campaigns.
5.
Economic Sanctions: NATO member
countries, in coordination with other international partners, can impose
economic sanctions on Russia to exert pressure and encourage a peaceful
resolution. These sanctions can target specific sectors of the Russian economy
or individuals involved in destabilizing actions.
6.
Confidence-Building Measures:
NATO can promote confidence-building measures between Russia and Ukraine, such
as establishing channels of communication, implementing arms control
agreements, and facilitating exchanges and cooperation in non-military areas
like culture, education, and business.
7.
Cooperative Security Initiatives:
NATO can explore opportunities for cooperative security initiatives with
Russia, aimed at enhancing trust and reducing tensions. These initiatives can
include joint military exercises, information sharing on military activities,
and dialogue on shared security concerns.
It's important
to note that NATO's actions are often carried out in coordination with other
international organizations, such as the United Nations and the Organization
for Security and Co-operation in Europe (OSCE). Resolving the Russia-Ukraine
conflict requires a comprehensive and multifaceted approach involving
diplomatic efforts, political will, and engagement from all relevant parties.
My comment: This example, compared to the Chat-GPT
answers given to the earlier question (above) “What can the NATO do
for achieving peace in the Russia-Ukraine conflict? shows how the ChatGPT
machinery is biased in the sense that items 3, 4 and 5 imply direct support of
Ukraine. It disregards the main source of the conflict in that NATO being a
creation under the political and military influence of the USA would not in its
charter guarantee the exclusion or non-membership of countries neighboring
Russia or its “sphere of influence”. All this under the assumption of national
sovereignty and other considerations in my essay on The
Russia-NATO-Ukraine Information Crisis, which for these same reasons could never have been produced by
ChatGPT. This is related to neglecting, to begin with, the philosophy of the
earlier mentioned Hegelian inquiring system (IS) and abusing the philosophy of
Leibnizian and naïve Lockean IS.
This illustrates what was mentioned above, that the
possible if not probable result of the use of AGI is that it will cement the
past and suffocate future questioning and debate, also caused by the waves on
unemployment combined with automation that restrict the range of opinions. This
happens already today in that terminological neologisms and abuse of
prestigious and “mind-blowing pseudo-philosophical” words and acronyms of all
sorts are being used and abused in reference to AI and AGI in the press and in social media. See
Swedish TV 18 Oct and 1 Nov. 2022, The Conference 2023: “Förstärkt Mänsklighet” [Enhanced Humanity], covering e.g. “artistic practices involving
emerging technologies” and “organize the world’s information to
make it universally accessible and useful”.
The idea is illustrated in the journal The
Economist’s article (Nov 9th 2023) on “Now
AI can write, sing and act, is it still possible to be a star?” with the under-title “The deal that ended the Hollywood actors’
strike reflects a fear of the technology”, which indeed is fear of infringement
of artistic copyright (cf. the Swedish event
Tosbot commented in Svenska Dagbladet 1-2 December
2023; on Tosbot in English here).
Not only politically correct journalists who do not
even need to be educated in political science, foreign relations and diplomacy,
but even school children will be able, without understanding it, to use ChatGPT
in order publish a logically structured and credible essay with facts about the
salvific power of AI, e.g. in "How to solve the conflict and achieve peace
between Russia, NATO and Ukraine?" But it will be much more difficult if
not impossible to search for or finance research of dissenters who on their own
will be able to afford and formulate a counter-story like the above-mentioned
The Russia-NATO-Ukraine Information Crisis or a text like the present one.
Furthermore: there are many of those problems surveyed
in the above mentioned outline on Existential risks from artificial general intelligence, starting
with the upheavals of the labor market. A few of them had been already noticed
in research during and prior to earlier "waves of optimism",
exasperating the social games of reduction of ethical, intellectual, and
theological dimensions to politics of liberalism vs. Marxism. They were
outlined in the leftist approach of a timely doctoral dissertation at my
university on Work-oriented Design of Computer-Artifacts (1988) as I comment it in my essay on Information and Theology, where it can be seen as an example of reduction of
religion and theology to science and politics, or directly from theology to politics.
7. Pause Giant AI Experiments: An Open Letter
This
document (cf. “See
also” in Wikipedia), published on March 22, 2023 by a
community related to the above mentioned "Existential
risks from artificial general intelligence" was
accessed on August 3, 2023 on the net with the
subtitle "We call on all AI labs to immediately pause for at least 6
months the training of AI systems more powerful than GPT-4". My earlier references above to
ChatGPT were to the version for the general public, the free version GPT-2. The
text of the “Open Letter” in my adapted lay-out, is the following – in a
smaller font size - in turn followed at the end by my own comments:
-----
AI systems with
human-competitive intelligence can pose profound risks to society and humanity,
as shown by extensive research[1] and acknowledged by top AI
labs.[2] As stated in the widely-endorsed Asilomar AI Principles, Advanced AI
could represent a profound change in the history of life on Earth, and should
be planned for and managed with commensurate care and resources.
Unfortunately, this level of planning and management is not happening, even
though recent months have seen AI labs locked in an out-of-control race to
develop and deploy ever more powerful digital minds that no one – not even
their creators – can understand, predict, or reliably control.
Contemporary AI
systems are now becoming human-competitive at general tasks,[3] and
we must ask ourselves: Should we let machines flood our
information channels with propaganda and untruth? Should we
automate away all the jobs, including the fulfilling ones? Should we
develop nonhuman minds that might eventually outnumber, outsmart, obsolete
and replace us? Should we risk loss of control of our
civilization? Such decisions must not be delegated to unelected tech
leaders. Powerful AI systems should be developed only once we are
confident that their effects will be positive and their risks will be
manageable. This confidence must be well justified and increase with
the magnitude of a system's potential effects. OpenAI's recent statement regarding artificial general
intelligence, states that "At some
point, it may be important to get independent review before starting to train
future systems, and for the most advanced efforts to agree to limit the rate of
growth of compute used for creating new models”. We agree: that some point
is NOW.
Therefore, we call
on all AI labs to immediately pause for at least 6 months the training of AI
systems more powerful than GPT-4. This pause should be public and
verifiable, and include all key actors. If such a pause cannot be enacted
quickly, governments should step in and institute a moratorium.
AI labs and
independent experts should use this pause to jointly develop and implement a
set of shared safety protocols for advanced AI design and development that are
rigorously audited and overseen by independent outside experts. These protocols
should ensure that systems adhering to them are safe beyond a reasonable doubt.[4] This
does not mean a pause on AI development in general, merely a
stepping back from the dangerous race to ever-larger unpredictable black-box
models with emergent capabilities.
AI research and
development should be refocused on making today's powerful, state-of-the-art
systems more accurate, safe, interpretable, transparent, robust, aligned,
trustworthy, and loyal.
In parallel, AI
developers must work with policymakers to dramatically accelerate development
of robust AI governance systems. These should at a minimum include: new and
capable regulatory authorities dedicated to AI; oversight and tracking of
highly capable AI systems and large pools of computational capability;
provenance and watermarking systems to help distinguish real from synthetic and
to track model leaks; a robust auditing and certification ecosystem; liability
for AI-caused harm; robust public funding for technical AI safety research; and
well-resourced institutions for coping with the dramatic economic and political
disruptions (especially to democracy) that AI will cause.
Humanity can
enjoy a flourishing future with AI. Having succeeded in creating powerful AI
systems, we can now enjoy an "AI summer" in which we reap the
rewards, engineer these systems for the clear benefit of all, and give society
a chance to adapt. Society has hit pause on other technologies with potentially
catastrophic effects on society.[5] We can do so
here. Let's enjoy a long AI summer, not rush unprepared into a fall.
----
Notes
and references
[1]
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021,
March). On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?🦜. In Proceedings of the 2021 ACM conference on
fairness, accountability, and transparency (pp. 610-623).
Bostrom,
N. (2016). Superintelligence. Oxford University Press.
Bucknall,
B. S., & Dori-Hacohen, S. (2022, July). Current
and near-term AI as a potential existential risk factor. In Proceedings of the 2022 AAAI/ACM Conference
on AI, Ethics, and Society (pp.
119-129).
Carlsmith,
J. (2022). Is
Power-Seeking AI an Existential Risk?. arXiv preprint arXiv:2206.13353.
Christian,
B. (2020). The Alignment Problem: Machine Learning and human values. Norton
& Company.
Cohen,
M. et al. (2022). Advanced Artificial Agents Intervene in the Provision
of Reward. AI Magazine, 43(3)
(pp. 282-293).
Eloundou,
T., et al. (2023). GPTs
are GPTs: An Early Look at the Labor Market Impact Potential of Large Language
Models.
Hendrycks, D., & Mazeika, M.
(2022). X-risk Analysis for AI Research. arXiv
preprint arXiv:2206.05862.
Ngo,
R. (2022). The
alignment problem from a deep learning perspective. arXiv preprint arXiv:2209.00626.
Russell,
S. (2019). Human Compatible: Artificial Intelligence and the Problem of
Control. Viking.
Tegmark,
M. (2017). Life 3.0: Being Human in the Age of Artificial Intelligence. Knopf.
Weidinger,
L. et al (2021). Ethical
and social risks of harm from language models. arXiv preprint arXiv:2112.04359.
[2]
Ordonez, V. et al. (2023, March 16). OpenAI CEO Sam Altman says AI will reshape society,
acknowledges risks: 'A little bit scared of this'. ABC News.
Perrigo,
B. (2023, January 12). DeepMind CEO Demis Hassabis Urges Caution on AI. Time.
[3]
Bubeck, S. et al. (2023). Sparks
of Artificial General Intelligence: Early experiments with GPT-4. arXiv:2303.12712.
OpenAI
(2023). GPT-4 Technical Report.
arXiv:2303.08774.
[4]Ample
legal precedent exists – for example, the widely adopted OECD AI Principles require
that AI systems "function appropriately and do not pose unreasonable
safety risk".
[5]
Examples include human cloning, human germline modification, gain-of-function
research, and eugenics.
----
We
have prepared some FAQs in response to questions and discussion in the media
and elsewhere. You can find them here.
In
addition to this open letter, we have published a set of policy recommendations
which can be found here:
We
have prepared some FAQs in response to questions and discussion in the media
and elsewhere. You can find them here.
In
addition to this open letter, we have published a set of policy recommendations
("Policymaking in the Pause") which can be found here, and has the following
title and contents:
Policymaking in the Pause:
What can policymakers do now to combat risks from advanced AI systems?
Contents:
Introduction
Policy recommendations:
Mandate robust
third-party auditing and certification for specific AI systems
Regulate organizations’
access to computational power
Establish capable AI agencies
at national level
Establish liability for
AI-caused harm
Introduce measures to
prevent and track AI model leaks
Expand technical AI safety
research funding
Develop standards for identifying and managing AI- generated content and
recommendations
Conclusion
8. Comment to the Open Letter (Proposal for moratorium)
If it
is not disrespectful I would like to start by asking the idiomatic "What
are we to make out of this?" My spontaneous reaction, if not response, is
to think about a book that left in me a deep impression when I read it more
than forty years ago: William E. Akin's Technocracy and the American Dream: The Technocrat
Movement, 1900-1941 (1977). I perceive it as a painstakingly
researched passionate life-work of an apparently forgotten "Associate
Professor of History in the Loyola Campus of Concordia
University, Montreal".
Reflecting
upon my impression I drew the conclusion that in my present situation I cannot
afford to do what I should: write an analog of the book focused on the ongoing
AI and AGI-hype. In doing so I would be writing more than a doctoral
dissertation on these matters, including a connection to the history of bureaucracy seen,
as AI, a system of logical rules working on collected data, and related to
technocracy as indicated in Wikipedia’s section on “See also”.
On
the other hand, as I write in my general
disclaimer in the link that initiates the list of “Contents” at
the beginning of present essay, such endeavor should not be necessary because
"it all" is already written in Akin’s above-mentioned book, the
problem being what Plato writes in my quotation above, and what is repeated in
later occasions in such books as, say, by
Richard Stivers. Let's see some excerpts from Akin regarding
"technological unemployment" that a doctoral dissertation at my
university department translated into a Marxist analysis of the role of labor
unions under the title computer Work-Oriented Design of Computer Artifacts (published as book here and commented
by me in the context of Information
and Theology). The political reason for the rise and establishment of labor
unions is today forgotten in the facile wishful thinking that technological
unemployment caused by AI-AGI will be avoided by adherence to “ethical
standards” or offset by “universal basic
income”.
The
key issue is that the Open Letter requesting of moratorium surreptitiously
subsumes a political democratic process while assuming that its implicit
technocratic view, expressed and supported by computer scientists, can be
reconciled with a misunderstood or ignored mythological
democracy. It is a democracy that according to technocratic
"policy recommendations" mandates auditing and certifications, regulates
access to computational power, establishes
agencies and liabilities, introduces
measures to prevent AI-misuse, and expands
funding for AI safety research.
Related
to mythological democracy’s lame “tautological voluntary and non-binding
reliance on experts”, it is then interesting to remark what The Economist (Nov 23rd, 2023) writes
with reference to the above-mentioned
“boomers”, in the context of Sam
Altman’s return marking a new phase for Open AI:
[Boomers] will worry politicians, who are scrambling
to show that they take the risks seriously. In July President Joe Biden’s
administration nudged seven leading model-makers, including Google, Meta,
Microsoft and OpenAI, to make “voluntary commitments” to
have their AI products inspected by experts before releasing them to
the public. On November 1st the British government got a similar group to sign
another non-binding agreement that allowed regulators to test their AIs for trustworthiness and harmful capabilities, such as
endangering national security.
And
now over to Akin on technological unemployment (pp. 156-164.):
The charge of
technological unemployment was the most relevant economic issue. Far more
serious, lengthy, and passionate discussion took place over this question than
any other. It was the most difficult to resolve and perhaps of most lasting
significance. […]
Somewhat
surprisingly, the spokesmen for business were among those unable to present a
solid front when the issue first arose. Businessmen have never wished to admit
the existence of technological unemployment. Aside from the serious economic
problems involved, to do so raised equally important questions regarding the social
and moral values of capitalism, which both entrepreneurs and corporations
preferred to leave unmasked. It cut through and could potentially destroy,
their most cherished notions: their easy identification of technological change
and progress, their assumption that change was compatible with stability, their
belief in the social and moral value of work, the idea of the self-made man and
the theme of individualism, and the necessity of laissez-faire. […]
The vehemence
with which leading scientists and engineers repudiated the concept of
technological unemployment reflected their fear that it constituted a frontal
attack on their professional and social roles. Their disquiet was even greater
than that of businessmen. To them the benevolence of science and technology was
indisputable. […]
The assurances
of scientists and engineers, that technology would continue to create new
products and industries, "each demanding an army of workers," also
lent support to the optimistic common-sense view despite the fact that it
required the same kind of blind faith. […]
Most businessmen
[…] followed a mechanistic theory which held that technology reduced costs,
thereby creating greater product demand, which increased production, which in
turn necessitated higher employment. […]
To avoid serious
repercussions, the economy's well-being ultimately depended on constantly
increasing the level of consumption. […] The logic of those who argued the case
dictated that over the long run increasing purchasing power was linked to economic
growth. To insure minimum dislocations, economic growth must match
technological improvement. […]
As important as
the question of technological unemployment was, the broad implications of
technology for society was of even greater significance for some. One of the
most pressing issues raised was the effect of technology on social
organization. Was there a cultural lag that had to be lessened or bridged? […]
Did it require a technical elite to engineer society as well as the machine?
[…]
One of the
shortcomings of the proponents of technocracy was their failure to reconcile
the technocratic view with democracy.
In
other words: the Open Letter is a replay or reenactment of the
"philosophy" of the technocratic movement with its problematically
related social
responsibility (c.f. also here and
on Walter Rautenstrauch
here), the more so when computerization in general and AI in particular seldom
if ever can be analyzed for profitability in terms of cost-benefit
analysis, but are valued in terms of saving costs of human
labor. And it is more than social responsibility, it is a matter of human love
in the sense of Thou shalt love thy neighbour as thyself (Matt. 22:39) It is not only a question of "unemployment".
A modern technocratic movement in its computer-artificial-intelligence dress,
does need neither the opinion nor a dialog and even less an "impossible
debate" with an untrained, not technically gifted and a
supposedly unintelligent workforce that meets the impossibilities
of a "human-computer interaction" that is, a
substitute of human language. It needs only the ultimate political legitimation
of a more or less mythological Democracy, It is a Democracy that today does not
display references to biblical quotations in order to expect that new computer
technology will create enough new influential jobs and will need neither the
Bible nor labor unions in order to allow a Universal basic
income to an increasing world-population. And it is a
Democracy that is used as a dumping place for personal responsibility, as I
write in another
theological context, and is done even by children in the context of climate global
warming where the “researcher” is the scientist and the
“manager” is the politic body of democracy:
Politics,
for instance, often is also explained away by scientists and engineers who
assume that all would be alright if only politicians followed the scientists'
recommendations, as denounced in the famous paper by Churchman
& Schainblatt The researcher and the manager. A dialectic of implementation and its Commentaries.
More
than so, the Open Letter ignores not only the paradox of computerization not
saving manpower and time, not allowing more time for culture and relaxation, as
surveyed Staffan Burenstam Linder’s The Harried Leisure Class, and in my text on the Meaning of Human-Computer interaction. It also mainly
ignores one main motivating force and commitments of the military-industrial
complex as in the latest conflict between Russia and Ukraine, and
the sociology as well as the political science lying behind the clash between
duly secularized socialism and liberal capitalism. A socialism with roots in
the absence of labor unions for countering the advent of atheist industrialization.
Today it is like enabling to ask ChatGPT-2 (or the later paid GPT-4) the
questions quoted above in this essay about the causes and the solution of the
Russia-Ukraine conflict, while one main force of the ongoing USA research may
be to complement the logical-mathematical rape of mind and nature by a quantum physics for
nuclear weapons, with the human-computer interacting AI rape of the human mind
for design and implementation of self-driving unmanned
weapons as analogs to drone warfare (as here, in
Ukraine).
This
Open Letter or proposal for moratorium will obviously not stand alone. It was,
for instance, followed by a related AI Safety Summit in London 1-2
November 2023, leading to the so called Bletchley Declaration (published by
the government of the UK, officially represented by the Department of Science,
Innovation & Technology, Foreign Commonwealth & Development Office, and
The Prime Minister’s Office, 10 Downing Street). It contained, among others,
the following typical policy thoughts, to be expected to be recurrent in a
plethora of coming AI conferences around the world (my italics):
Particular
safety risks arise at the ‘frontier’ of AI, understood as being those
highly capable general-purpose AI models, including foundation
models, that could perform a wide variety of tasks - as well as relevant
specific narrow AI that could exhibit capabilities that cause
harm - which match or exceed the capabilities present in today’s most
advanced models. Substantial risks may arise from potential intentional misuse
or unintended issues of control relating to alignment with human intent. These issues are in part because those
capabilities are not fully understood and are therefore hard to predict.
We are especially concerned by such risks in domains such as
cybersecurity and biotechnology, as well as where
frontier AI systems may amplify risks such as disinformation.
There is potential for serious, even catastrophic, harm, either
deliberate or unintentional, stemming from the most significant capabilities
of these AI models. Given the rapid and uncertain rate of change
of AI, and in the context of the acceleration of investment in
technology, we affirm that deepening our understanding of these potential
risks and of actions to address them is especially urgent.
Yes,
indeed: “These issues are in part because those capabilities are not fully
understood and are therefore hard to predict”. “Capabilities” or
especially risks not fully understood, or not
at all understood? This was the reason for the upheaval in the episode of
the world-wide diffused news in November 2023 of the sudden
sacking and reinstatement of genius-wizard Sam Altman from
the position of CEO of OpenAI, the creator of (excessively?) successful of
ChatGPT. The news agencies repeated that the explicit reason given for his
sacking had been “unclear-vague” shortcomings in his communications with the
board. They did not dare to advance the hypothesis that the OpenAI investors,
employees and prospective clients were welcoming Altman’s optimism in
downplaying the “existential risks” of AI (see above) in contrast to the OpenAI
board’s eventual mounting political-ethical concerns. This may be an illustration of the coming ethical and political
struggles in the development and applications of AGI, which may be hopeless
if one considers the risk that today political Democracy is a myth. It is a myth hidden behind non-binding assurances that
vague, undefined and problematic “productivity-efficiency-effectiveness” (see
the “theoretical-conceptual framework” in hodgepodge here)
will be “responsibly” managed with expertise, following democratically decided
governmental rules after a rich dialogue with all affected parties, avoiding
the production of fake information or information implying “existential risks”,
controlling for protecting from access by extraneous irresponsible influence
from non-authorized personnel, it all guaranteed by security-safety measures,
etc. etc.
I
repeat now from the beginning of the previous paragraph: “These issues
are in part because those capabilities are not fully understood and are
therefore hard to predict”. “Capabilities” or especially risks not fully understood, or not at all understood:
For
instance, the technocratically avoided issue of unemployment of people that are
simply sacked and replaced by AI-AGI. Even in so-called intellectual work,
mediocre professionals may be replaced by AGI/ChatGPT, but they may also
improve their productivity with its help, while the best top-professionals may
not improve their performance because they, to begin with, were not reasoning only
“logically-empirically”, as slow computers without even taking stand in the
debates about logical
empiricism, with the consequence that they will not continue to
be rated as “best”.
Paradoxically
this defective understanding is hoped to be achieved by a future long series of
conferences. Cf. the above-mentioned AI
Safety Summit (never mind what the concept of safety or security is or
should be, as suggested above)
that is only one conference in the first waves of the “conference industry” in
the increased hype of AI in years 2022-2023, immediately followed by e.g. a
conferences in London on November 9, 2023 Generative AI Summit, on
November 27-28 AI World Congress 2023, and on November 30-December 1, 2023 on Delivering AI and Big Data for a Smarter
Future (also AI & Big Data Expo). In general, risks are not understood,
and AI/AGI are the latest reminder of that ongoing societal computerization,
which may be soon followed by news on applications
of mind-blowing computational neuroscience, brain-computer interface and the like, employing, among others, neurophysiologists,
with skills exemplified by the president of the
Allen institute for Brain Sciences, and others who
are introduced as being e.g. a “neurophysiologist and computational neuroscientist”
and work on the “neural basis of consciousness”. That is: whatever that means
for even a highly educated democratically minded citizen who will exert his
democratic duties in future elections that will direct the desirable future of
national scientific efforts. It all can be seen as a gigantic irreversible
global experiment on humanity, as illustrated also in my already mentioned
paper on the
meaning of human-computer interaction. Let me complete
this section with replicating a quotation from the epilogue of
my paper on computerization and logic which I hope may convey an image of the
increasing scope of techno-science and of computer-logical dreaminess, from the
atomic bombs on Hiroshima and Nagasaki, up to today’s super-human, divine
intelligence. It can be seen either as a hymn to a
fantastic promising limitless technological progress (towards…) or as a
document of recurring naïve technocratic hype.
As an illustrative tribute to the genius and naivety of the celebrated
great engineer Vannevar Bush, I will terminate with a quotation of his
most famous prophetic article “As we may think” in The Atlantic Monthly issue
of July 1945, that I recommend to the readers for getting a time perspective on
the drive for computerization:
It
is readily possible to construct a machine which will manipulate premises in
accordance with formal logic, simply by the clever use of relay circuits. Put a
set of premises into such a device and turn the crank, and it will readily pass
out conclusion after conclusion, all in accordance with logical law, and with
no more slips than would be expected of a keyboard adding machine.
9.
Case study: “Creating safe AGI that benefits all of humanity”
On
December 24, 2023 I accessed the Open AI’s
policies and terms of use (effective December 23, 2023, previous versions here) for
the European Economic Area (EEA), Switzerland, or UK, that were effective on
December (for others living outside the EEA, Switzerland or UK, see other terms of
use). From the contents I made the following core
selection as material for the case study in the form of subsequent comments
along the above text in this essay, keeping most of the original layout except
of the font type:
Who We Are
OpenAI
is an AI research and deployment company. Our mission is to ensure that
artificial general intelligence benefits all of humanity. For more information
about OpenAI, please visit https://openai.com/about. Our Services are provided to you by:
.
OpenAI Ireland Ltd, a company
incorporated in the Republic of Ireland with its registered office at 1st
Floor, The Liffey Trust Centre, 117-126 Sheriff Street Upper, Dublin 1, D01
YC43, Ireland and company number 737350, if you are resident in the EEA or
Switzerland.
.
OpenAI, L.L.C., a Delaware
company with its registered office at 3180 18th Street, San Francisco,
California 94110, United States and company number 7063675, if you are resident
in the UK.
Additional Service-Specific Terms
Depending
on the specific Service or features you use, additional Service-specific terms
and policies may apply to your use of our Services. The key ones to be aware
of, and which form part of these Terms, are described below:
.
Usage Policies: these policies explain how you may use our Services
and Content.
.
Service Terms: these terms apply when you use certain Services or
features;
.
Sharing & Publication Policy: this
policy sets out rules for when you share Content;
Using Our Services
What
You Can Do. Subject to your compliance with these Terms, you may
access and use our Services. In using our Services, you must comply with all
applicable laws as well as the Service-specific terms and policies listed
above.
What
You Cannot Do. You may not use our Services for any illegal,
harmful, or abusive activity. For
example, you are prohibited from:
.
Using our Services in a way that infringes,
misappropriates or violates anyone’s rights.
.
Modifying, copying, leasing, selling or distributing
any of our Services.
.
Attempting to or assisting anyone to reverse engineer,
decompile or discover the source code or underlying components of our Services,
including our models, algorithms, or systems (except to the extent this
restriction is prohibited by applicable law).
.
Automatically or programmatically extracting data or
Output (defined below).
.
Representing that Output was human-generated when it
was not.
.
Interfering with or disrupting our Services, including
circumventing any rate limits or restrictions or bypassing any protective
measures or safety mitigations we put on our Services.
.
Using Output to develop models that compete with
OpenAI.
Content
Your
Content. You may provide input to the Services (“Input”),
and receive output from the Services based on the Input (“Output”).
Input and Output are collectively “Content”. You are responsible for
Content, including ensuring that it does not violate any applicable law or
these Terms. You represent and warrant that you have all rights, licences, and
permissions needed to provide Input to our Services.
Ownership
of Content. As between you and OpenAI, and to the extent permitted
by applicable law, you (a) retain your ownership rights in Input and (b) own
the Output. We hereby assign to you all our right, title, and interest, if any,
in and to Output.
Similarity
of Content. Due to the nature of our Services and artificial
intelligence generally, Output may not be unique and other users may receive
similar output from our Services. Our assignment above does not extend to other
users’ output or any Third Party Output.
Our
Use of Content. We can use your Content worldwide to provide,
maintain, develop, and improve our Services, comply with applicable law,
enforce our terms and policies and keep our Services safe.
Opt
Out. If you do not want us to use your Content to train our
models, you have the option to opt out by updating your account settings.
Further information can be found in this Help Center article.
Please note that in some cases this may limit the ability of our Services to
better address your specific use case.
Accuracy.
Artificial intelligence and machine learning are rapidly evolving fields of
study. We are constantly working to improve our Services to make them more
accurate, reliable, safe, and beneficial. Given the probabilistic nature of
machine learning, use of our Services may in some situations result in Output
that does not accurately reflect real people, places, or facts.
When
you use our Services you understand and agree:
.
Output may not always be accurate. You should not rely
on Output from our Services as a sole source of truth or factual information,
or as a substitute for professional advice.
.
You must evaluate Output for accuracy and
appropriateness for your use case, including using human review as appropriate,
before using or sharing Output from the Services.
.
You must not use any Output relating to a person for
any purpose that could have a legal or material impact on that person, such as
making credit, educational, employment, housing, insurance, legal, medical, or
other important decisions about them.
.
Our Services may provide incomplete, incorrect, or
offensive Output that does not represent OpenAI’s views. If Output references
any third party products or services, it doesn’t mean the third party endorses
or is affiliated with OpenAI.
My comments:
I see the policies
and terms of use mainly as an expression of goodwill, and mainly as a complex
juridical disclaimer based on two main concepts of service, and user. I
myself wonder whether OpenAI, without any
notice has deleted my account (as per January 2024) and started refusing my IP Address (I
can no longer neither log-in nor sign up from my regular IP address) because I
have used its services for commenting them criticizing AI as I showed above in
my present text, as they perhaps were judged to infringe the last paragraph
above on “When you use our Services you understand and agree”.
These
policies and terms of use can and must be seen and related to production-consumption, or means-goal, or producer-consuming client, or scientist
producer of a tool, or politician user of the tool. I have
already dwelled on the problems of this conception focused on “tool” in my
paper on The
meaning of human-computer interaction, and for the purpose of space and time I must refer the
reader to this source in its section
dealing with tool.
The
core of the problem, however, can be seen in the definition of producer,
product and user in terms of the explanation of the meaning of so-called
morphological or structural, functional and teleological classes as introduced
in the earlier mentioned DIS. It
includes the overview of complications, for instance, that one cause produces
one effect, but another cause or several other causes can also produce the very
same effect. But it is later realized that some of these productions can only
happen under earlier unknown external conditions that are under the control of
others who happen to be interested and motivated by the desire of seeing or
having the result, or motivated to sometimes prevent it, because of being motivated
to counter the happening of such result. And so on. In the middle of all this,
mathematics and its abuse has a special simplifying function as I consider it
in the section of my essay on the
famous rejected parts of Jan Brouwer’s
dissertation about the foundations of mathematics, where I write the following:
In presenting The
rejected parts of Brouwer's dissertation, [Walter
Peter] Van Stigt refers to Brouwer's interpretation of
causality as essentially mathematical: the ability to link events in the mind,
to see sequences and repetition of sequences in time, to link sensations as the
immediate source of awareness of time and discreteness. It is the source of
man's power to predict the future and interfere in the course of events. This
"intellectual or mathematical" way of looking at the world is not
only a one-sided concentration and interpretation of reality: by
ignoring and willfully removing aspects which deviate from the expected course
of events, man supplements and creates more regularity than exists in nature,
he makes the world linear or "one-sided". The regularity observed
in nature is due to the nature of the measuring instruments and physical
science has value only as weapon, not concerning life. It is clearly inferior
and has nothing to do with religion or wisdom. More in detail,
in Brouwer's own words:
"Man has the faculty, accompanying all
his interactions with nature, of objectifying the world, of seeing in the world
causal systems in time. The primordial phenomenon is simply the intuition of
time in which repetition of "thing in time and again thing" is
possible, but in which (and this is a phenomenon outside mathematics) a
sensation can fall apart in component qualities, so that a single moment can be
lived through a sequence of qualitatively different things. One can, however,
restrict oneself to the mere sensation of theses sequences as such, independent
of the various degrees to which objects are perceived in the world outside are
to be feared or desired. (The attention is reduced to an intellectual
observation.) The human tactics of "acting purposively" then consists
in replacing the end by the means (a later occurrence in the intellectually
observed sequence by an earlier occurrence) when the human instinct feels that
chance favours the means."
What
happens, then, is that the reader may find the explanation tedious and will
ask, as I have many times asked, to summarize the whole thing I a few words.
So, what happens is what I already have tried to explain in the context of Information and Debate: People feel that they have no time, no motivation or no
capabilities to read such texts. If they hear that the explanation in summary is partly also found in
Russell Ackoff’s and Fred Emery’s chapter on Structure, Function and Purpose in their book On Purposeful Systems (pp. 13-32), they may wish to get a summary of the
summary, in words that allow them to grasp the whole in a few hours, or
minutes.
But
the summary, which appears in DIS, (p.
59) is not easily understood and accepted when it states:
Thus in the
broader viewpoint one cannot distinguish between science and its politics; it
make no sense to the designer to say that science is a body of knowledge and
politics is people, and therefore the two must be different. For the designer
it is impossible to optimize the system of acquiring basic knowledge without
considering the political problems that such a system generates. The boundaries
of “basic research” expand into the area of national policy making, and the
client becomes larger than the scientific community.
And,
in fact, returning to text starting some lines above the last paragraph of the
selection out of “terms of service”:
.
1. Given the
probabilistic nature of machine learning, use of our Services may in some
situations result in Output that does not accurately reflect real people,
places, or facts.
.
2. Output may not
always be accurate. You should not rely on Output from our Services as a sole
source of truth or factual information, or as a substitute for professional
advice.
.
3. You must evaluate
Output for accuracy and appropriateness for your use case, including using
human review as appropriate, before using or sharing Output from the Services.
.
4. You must not use any
Output relating to a person for any purpose that could have a legal or material
impact on that person, such as making credit, educational, employment, housing,
insurance, legal, medical, or other important decisions about them.
.
5. Our Services may
provide incomplete, incorrect, or offensive Output that does not represent
OpenAI’s views. If Output references any third party products or services, it
doesn’t mean the third party endorses or is affiliated with OpenAI.
That
is an attempt to express a juridical
disclaimer, smart lawyers editing a text that relieves the producer from
risks, if sued. For me it is analogy to a manufacturer of weapons who advises
the buyer/client about the use of the weapons. There are national laws that
regulate such an industry and commerce but there is also a world of arms trafficking and criminal possession of weapons
related to international politics and conflicts. In our case it is a disclaimer
that seeks an assurance that weapons will not be given to, and used by children
but only used by “us” or by democratic and friendly people in friendly
countries, and only for self-defense, as only by Ukraine in the conflict with
Russia, disregarding all the complications that are suggested in my essay on the conflict. And
regarding the above repeated text of numbered “terms of service” above:
1.
The output may not accurately reflect real people, places, or facts. [Who
should and can do what about that it may not be truth?].
2.
They say that you should [or must?] not rely on output from our services as a sole source of truth [but claims that it
is truth?].
3.
You must evaluate. [But how? – this may
mean understanding and applying the ignored DIS].
4.
You must not use any output that could have a legal or material impact. [We,
who, “must not use”, but how to know whether it “could” have impact?].
5.
Our services may be offensive. [But it probably intends to claim that “we” are
not offensive].
And
who are “we” and who are “you”? Among the “we”, I can claim that I myself
understand and follow “musts” but who is responsible for the whole of “us” and
of “you”, and what does it imply to be “responsible”, for what, and to whom? To
the state, in state
individualism, which is often equated to “society”? All guaranteed
by the “myth
of democracy”?
10. Conclusion: The
meaning of hype
All
this up to the point of raising the question of what “hype” does mean in our
context. Today it is not general popular knowledge to know about and relate to
the historical issue of Gustave Le Bon’s The Crowd: A Study of the Popular Mind, as the beginning of crowd psychology, and at the edge of Groupthink, with
its relation to the wiki-concept and to my own doctoral thesis on Quality-control of Information.
Regarding
criticism and influence of the book The
Crowd, Wikipedia mentions feelings of power and security allow the
individual not only to act as part of the mass, but also to feel safety in numbers. Today the safety in numbers may be in
numbers of fellows on the net who share the misunderstanding or belief, the
number of nodes and hits in the Internet, or the volume of big data.
For
all its relations to Freudian psychology I think that the message is better
understood in the light of analytical psychology and Carl Jung’s reflections in
his Civilization in Transition (Collected
Works, vol, 10). I dare to mention my own association to the
particular chapter in the book (# 589-824) where Jung (mentioned by addicts to
the New Age
movement) analyzes “Flying Saucers: A Modern Myth of Things Seen in the Skies”.
In terms of philosophy of science, this as well as the case of the technical
hubris of the Titanic-case, may
very well be fruitful, if yet apparently farfetched analogies, recalling George
Lakoff and Mark Johnson’s Metaphors We Live By. A
thoughtful book by the professor of sociology Richard Stivers says it all: Technology as Magic: The Triumph of the Irrational, including the article on “The Case of the
University”. It is a serious alternative to Arthur C. Clarke’s
perhaps better known phrase “Any
sufficiently advanced technology is indistinguishable from magic”
(see also here).
Never mind that the complexities of the philosophy of technology hide the basic
role of the psychic abuse
of formal thinking. A reminder of the tragedy behind it all is the
tragic historic analogy is the Holocaust, based on a whole culturally advanced
nation’s deep popular, historic, and scientific “knowledge” of the minds,
behavior, and history of mainly Jews as an ethnic group. So many references and
so much, in part misleading, reading would not be necessary if, as Blaise Pascal
remarks in his Pensées, (1955,
§251) humans would “fall in love
for”, or accept and follow Christianity that unites psychological exteriority
with interiority, or follow at least some of the few great world religions.
I
think that the situation is very serious and the more so because the
consequences, grave as they can be, will not be perceived, except in the long
run, in the same way as cultures may degenerate and conflictual relations
between people and nations develop until exploding in murders and wars,
historically even two world wars. More recently we have the examples of the Russia-Ukraine conflict and Israel-Hamas
war. We have also cultural “civil wars” like feminism
(cf. our boomers vs. doomers mentioned earlier) where the gregariousness of
nationalism or sex-gender or analytical leaning attracts individuals who have
not a fully developed psychic identity (individuation) and therefore easily
achieve “meaning and identity” by melting with a collective. Even with no
visible conflict, the consequence can be an oppressive passivation of
especially old, weak, ill or handicapped people who will not be able to further
their own opinions and interests, as in my illustration of difficulties and
limitations of the already mentioned human-computer
interaction.
In
the meantime, the hype is invading our daily environment. I picked up a
concrete simple example in Sweden but I should have preferred to wait until
another text appeared that I saw much later in the Swedish design magazine Form, No. 2, 2024, p. 71, written by
Björn Nordin, with the title “Människan och AI” [Humans and AI]. I propose Swedish
readers to also consult it after reading my analysis below because it illustrates
what is coming to drain or drawn us intellectually. By then I had already,
painfully, worked out my comments that follow, to the first mentioned text that
is an advertising brochure enclosed in and distributed together with the
newspaper Svenska Dagbladet in
mid-December 2023, with the title (my trans.) AI or HI. I did not see an explanation of what HI is, probably human intelligence. It was produced by
the Swedish branch of the Universum Employer
Branding Agency. In two pages (pp.4-5) it offers a series of
affirmations about AI. I will begin by displaying a short summary of the text
(my trans.), followed by the same text interspersed with my own commentaries
[with my emphasis in italics and in
square brackets] based on arguments presented earlier in this essay. The
brochure begins by telling that:
Already in year
1589 Queen Elizabeth I rejected a patent application for a knitting machine for
fear that it would put the population out of work, so concerns about the impact
of new technology on jobs are neither new nor particularly surprising. Already
today many types of basic counselling and customer service are now carried out
without human contact. The CEO of a firm, Futurion, calls AI "the
super colleague who never sleeps", who knows most languages, has a number
of different degrees”, and at the same time notes that many of the skills we
were able to study in the past, AI is now able to handle perfectly well on its
own: only when it comes to managing human creativity and critical thinking, however,
it is not doing so well. The CEO also finds that the demanded skills of the
future will be very much about social and
creative skills such as dealing with people, the ability to put
together broad and to come up with original
and unusual solutions. New roles and responsibilities will emerge. We will
therefore have to learn and re-learn, probably
several times over. It is too early to say what these new roles will look
like, but there is already a growing demand for technical specialists in machine learning, prompt
designers and AI developers.
Generative AI
such as ChatGPT cannot replace the creativity
and imagination of writers, designers and other creative professionals. It
requires a level of originality, emotion and expression that machines cannot
replicate, at least not yet. What do we do with the time left over? We don't
know yet. But if we can avoid spending evenings reading meeting documents,
getting help to tidy up the statistics for the presentation and getting an
editable first draft of a document in a couple
of minutes, we should reasonably have a lot of time to spare. The
transition from an eight to six hour working day is possible with increased use
of AI, however, it is important to note that it is not only a technological
issue but also a socio-economic and cultural one. Future changes in working habits are likely to be
influenced by a combination of technological advances, policies, and changes in
societal attitudes towards work and
leisure. Whatever the outcome, we can say that AI is here to stay, and if you
don't want to be left behind, you might as well jump on the bandwagon.
WITH
MY COMMENTS:
Already
in year 1589 Queen Elizabeth-I [but, please, mind the
difference between the worlds of 1589 and 2023, and why not mention also the
better studied luddites among whom I myself may be
classified] rejected a patent application for a knitting machine
for fear that it would put the population out of work, so concerns about the
impact of new technology on jobs are neither new nor particularly surprising.
Already today many types of basic counselling and customer service are now
carried out without human contact. [Already commented
earlier in my earlier text]. The CEO of a firm, Futurion,
calls AI "the super colleague who never sleeps", who knows most
languages, has a number of different degrees”, and at the same time notes that
many of the skills we were able to study in the past, AI is now able to handle
perfectly well on its own: only when it comes to managing human creativity and
critical thinking [how to recognize that?],
however, it is not doing so well. The CEO also finds that the demanded skills
of the future will be very much about social
and creative skills such as dealing with people [not in chat bots!],
the ability to put together broad solutions [synonym
to the earlier-mentioned DIS “systems
thinking?] and to come up with original and unusual solutions [how
to recognize and classify them?]. New roles and responsibilities
will emerge. We will therefore have to learn and re-learn, probably several times over [equivalent
to outweigh and repair short-term truths and solutions?]. It
is too early to say what these new roles will look like, but there is already a
growing demand for technical specialists
in machine learning, prompt designers and
AI developers. [guess: what else?].
Generative
AI such as ChatGPT cannot replace the creativity
and imagination [what are they and their relation?] of
writers, designers and other creative professionals [but
cf. the earlier mentioned infringement of
artistic copyright]. It requires a level of originality, emotion and
expression [what are they and their relation?] that machines cannot replicate, at least not
yet [remark this not
yet]. What to do we do with the time left over? [Left over by whom to whom? More automation in supermarkets leading to
less employees, transfer of tasks to customers, same length of queues]. We
don't know yet. But if we can avoid spending evenings reading meeting documents
[why?], getting help to tidy up the statistics for the
presentation and getting an editable first draft of a document in a couple of minutes, we should reasonably
have a lot of time to spare [again who are “we”,
employers or fewer employees?] The transition from an eight to six hour working day
is possible [with the same amount of manpower?] with increased use of AI, however, it is
important to note that it is not only a technological issue but also a
socio-economic and cultural one [and religious about
greed, and political, cf. labor unions?]. Future changes
in working habits are likely to be
influenced by a combination of technological advances, policies, and changes in
societal attitudes towards work and
leisure [typical ChatGPT-text, and what attitudes?]. Whatever the outcome, we can say that AI is here to
stay, and if you [cf. “we” vs. “you”]
don't want to be left behind, you might as well jump on the [whose?] bandwagon.
11. Conclusion: Beyond the hype
ENOUGH
WITH (TOO MANY) COMMENTS. What are to make out of this nice streamlined text,
streamlined except for my own emphasis and commentaries? They must have been
partly and cheaply edited automatically by ChatGPT, such as many other coming
published texts that can be easily read and accepted but would require an
attentive reading and critical examination. It is like reading the ChatGPT
answered to my questions about the causes of the conflict between Russia and
Ukraine, and compare them with what I write about in my
paper on the issue. In contrast, we can expect the whole society to be
drowned in paying attention to an immense amount of such ChatGPT/AGI
advertising, The advertising will include the apparently more ambitious texts
such as the above-mentioned intelligently artificial artistic production. The stuff will be almost daily worldwide produced and
broadcasted also in documentaries on AI, in the coming explosive future of the
AI-industry.
In
this perspective the obvious real productive capabilities of AI/AGI may cause a
collective schizophrenia, or cultural crisis or lobotomy of the “crowd”. It
will keep few in the crowd busy in trying to respond to “automatic”
ChatGPT-produced texts, including continuous AI-advertising, produced by lots
of people, businesses, employees who have vested interests in continuing to
foster the AI/AGI-hype that guarantees their own employment. Problems that
would have required a broad social systemic understanding of the causes of the
problem will instead lead to the discussion of whether and what particular,
continuously arising AI/AGI “tools” developed by a few businesses may help to
solve the problems. A high rate of unemployment caused by AI/AGI will at the
same time disclose the high fraction of present job opportunities that do not
require humanity for their production, but mainly or only routine computer-like
routine calculation, mechanical movements, and consumption of demanded but unnecessary products (including future
AI/AGI) that keep the economy going. Promises for the above-mentioned “universal basic
income” and the financing of wars, irrespective global warming,
will ignore economy, social psychology and politics, the problem of human greed
and of the difficulty to practice the
goodness of Christianity and major religions, as evidenced by wars and
by the rise of labor unions related to communism and socialism.
The
Faustian bargain reminds that a good portion of science and technology could
have been seen as legitimate in facilitating everyday life at unknown costs,
but for whom? Credible long-term cost-benefit analysis will not be possible, as
little as it will be possible to check the accuracy of statistical predictions
of the future effects of new AI applications. When facilitation and increased
productivity will lead to more unemployment, to the production of nuclear
weapons and increased production of more lethal weapons for ongoing wars, and
cause continuous worldwide pollution, this will raise claims of dangers and of
climate warming, leading desperate souls to feel and write analogies to the Industrial Society
and Its Future (see below). The counterparts will claim to strive
for and rely on future solutions by means of “more of the same”, godly
super-human intelligence that makes many humans obsolete, confirming that it
was Faustian hubris. It recalls the famous title of Oswald Spengler’s famous
book The Decline of the West,
which may also be seen as a secular interpretation of, or one more steps in the
understanding of the Book of Revelation.
Atheists
and those who cannot surmount the difficulties of reading and interpreting the Book of Revelation can instead consider
the possibility of having in their minds an Icarus complex, despite
vague doubts that have been expressed as to the therapeutic value
of such a diagnosis. It displays, however, similarities with other better known
alternative concepts, including the already
mentioned autism or other diagnoses as narcissistic personality disorder. The description of
the Icarus complex does refer to the prominent American psychologist Henry
Murray who in turn has had his life story related to the famous case of the American mathematician Ted Kaczynski, who is further described as
“a mathematician who went on to be known as the
'Unabomber'. Between 1978 and 1995, Kaczynski murdered three individuals and
injured 23 others in a nationwide mail bombing campaign
against people he believed to be advancing modern technology and the
destruction of the natural
environment”.
More sophisticated than simple “luddites” and even activists in the crusade against global
warming, he must have felt an overwhelming crushing discomfort in the contact
with the technical-industrial society at the time, a feeling similar to the one
felt by many contemporaneous citizens who today live in complex febrile
megacity milieus but long for a simple life in contact with nature. He authored
his own description and interpretation of the Apocalypse in the thoughtful book
Industrial Society
and Its Future (Wikipedia report here, printed book here). It is a manifesto and social
critique opposing
industrialization, rejecting leftism, and advocating for a nature-centered
form of anarchism that
today would have a deal in common with the climate and environmental
movements. I treat those commonalities with emphasis on climate in another
essay on Climate
and apocalyptic global warming.
My
point here, however, is directed to those who have enough mathematical-logical
leaning for having been lured into an Icarus-complex, pathological narcissism,
or denial of possibilities to avoid the Apocalypse. Even if they do not have
the reported high mathematical genius and intelligence of Ted Kaczynski
with an IQ at 167, better they choose a career in the AI-industry, like
the above-mentioned boomers who are
opposed to doomers. Doing so, becoming boomers, they may avoid
Kaczynski’s tragedy that fits doomers, his being a good example of what even a
pure mathematical-logical genius can lead to. It can lead to an additional
“logical” tragedy of crime and war such as the one I recently described for Ukraine,
which can be seen as a logically motivated war between boomers and doomers. But
one may guess that the best solution is to upgrade the desperation of doomers
and exhilaration of boomers to a better understanding of how to avoid computerization
as abuse of formal science, allied in the driving West to an active Christian
attitude. It would prevent the vain hope that “democracy” will allow us the
“freedom” of setting up more laws, more police and an extended judicial system
for forced control of the future use and abuse of AI/AGI. It would also prevent
us from seeing a coming AI/AGI as the saving Messiah that leads us towards
paradise, alternatively as expecting the Apocalypse as an unavoidable tragedy.
For
the rest, it is symptomatic that I never found in the Swedish (and barely in
international) media an in-depth interview or discussion of the matters related
to this text of mine. That is: until February 2024 when in the Al Jazeera
English (Europe) news channel, in “The AI Series”, Nobel
peace prize laureate Maria
Ressa interviewed and discussed with Urvashi
Aneja (also here) in Goa, technology
policy researcher and founding director of Digital Futures Lab, the subject How currents AI developments impact the Global South
(also here, plus additional text here)
introduced with the following text, where “the Global South” refers to the view
from Goa, in Southeast Asia:
While many of
today’s headline-grabbing artificial intelligence (AI) tools are designed in
Silicon Valley, much of the labour that fuels the boom is based in the Global
South, raising questions over who stands to gain from the technology and at
what cost.
Nobel Peace
Prize laureate Maria Ressa and the director of the Digital Futures Lab, Urvashi Aneja, explore the impact AI is
already having on communities in the Global South – from labour conditions to
democracy and the environment – and why countries need to move beyond “catching
up with the North” when deciding what role AI can and should play in their
societies.
In another
text, the very same Urvashi Aneja relates to the same
thought I had expressed above
about the “cementing” of human thinking by people in industry, business,
finance and academia, living in sub-cultures or societies in cultural crisis:
”Artificial intelligence is a status quo technology, as it reproduces the
future based on the past. […] The current model of AI development concentrates
power in the hands of a few. What we need is a new model that puts public
interest at the centre.”
I
mean, it is dramatically symptomatic that a finally welcomed such a serious
discussion of AI/AGI should come from the “Global South”. It can also be seen
as a part of the “conclusion” except for that a new model that puts public interest at the center must be seen primarily as the “old” forgotten model, an ethical and religious rather than
political question. As I suggest in my essay on Information and Theology.
To
Fellow Engineers
Over
lunch at a Swedish embassy, the hostess asks an old engineer if he knows the
difference between a lady, a diplomat and an engineer:
- Do
you know the difference between a lady, a diplomat and an engineer?
The
old man calmly said yes and explained:
- The
lady, when she says NO, means MAYBE; when she says MAYBE, she means YES; and
when she says YES she is not a lady.
- The
diplomat, when he says YES, means MAYBE; when he says MAYBE, he means NO; when
he says NO, he is not a diplomat.
- The
engineer, when he says YES, means YES; when he says NO, he means NO; and when
he says MAYBE... he's not an engineer.
That's
it!
-------------
[Forwarded by a Brazilian colleague, from unknown source, translated
from the Portuguese]
Spoiler suggestion: The engineer is the only one who is supposed to deal
mainly with inanimate nature. And is supposed here to ignore statistical
probability and measurement theory, which introduce the MAYBE. The only
obligatory party in the engineer’s dialogue is the one who, YES, pays the
monthly salary. And it has been already recalled above that the German philosopher
and theologian Ernst Troeltsch observed that humor may have similar influence on
atheists as religion does, in fostering humility by downplaying the importance
of the big Ego that is inflated by logic of YES and NO.