Philosophers of science like to remind us, as if this were the epitome of good common sense, that we should never confuse epistemological questions (what our representation of the world is) and ontological questions (what the world is really like) . Unfortunately, if we followed the philosophers’ advice we would not understand any scientific activity, since confusing those two supposedly separate domains is precisely what scientists spend much of their time doing.
Bruno Latour, Pandora’s Hope (1)



“An effort in vain” I recall thinking when I read the The God Delusion by Richard Dawkins,(2) the first time, about 15 years ago. Why refute the belief in the existence of god by employing scientific reasoning? The question belongs to a categorically different context, rather than being a scientific hypothesis. Of course god does not exist, it is obvious. Dawkins is a brilliant scientist and a student of Darwinism drawing from many different fields to set up his argument, from biological to cosmological reasoning not only to discredit the existence of god but also what follows, namely the case of intelligent design and creationism. Re-reading the book, I am now more impressed that his effort is an excellent argument to convince any rational thinking person that had a relatively decent scientific education, a kind of ‘materialism’ to deconstruct religious faith, frequently using the terminology of evolution, the birth of religion, altruism and morals. But still, I have doubts about the urgency of setting up an argument in these somewhat adversarial lines, as though the two ‘episteme’ are compatible. The shades of agnosticism that he lists in the book cover much of the educated population worldwide who are rather indifferent on the outcome. Every so often Dawkins lists the dark side of religious faith from wars to intolerance. One is tempted to think that the question of god’s existence and varied motives of the faithful and their deeds from good to outright evil and vile may be separated. Shouldn’t we differentiate between the socialized, organized and militant religion (as in American Evangelicals or political Islam) and that of pure inner faith and belief in a higher being? (as though there is such faith that does not touch the body social). Possibly, the God delusion and a kind of folklore, of interest to anthropology, can be separated. Or maybe not.
And so is it with Artificial Intelligence. Its existence is in no doubt, one does not have to attribute god-like qualities to it even though some call it ‘the deity’. But the hype that surrounds AI is akin to the initiation of monotheism, a ground breaking and world-changing phenomenon whose bible even comes with its own doomsday script and the armageddon. Indeed, the fuzzy term AI refers to techniques, tools and products bundled in a single package, creating a large bubble. It is also a kind of ‘machine intelligence’ set up by humans, but never a ‘consciousness’ or a ‘sentience’ . So, why fight it on human (and humane) grounds? Millions of years of evolution on earth sets the difference between the human (and biotic) response to a given stimulus and that of a large language model or Artificial General Intelligence. They are categorically different. Still, I think, it is imperative to refute the hype of AI as feverishly as Dawkins does for the delusion, on ontological, moral and economical (thus ecological) grounds.
I now think that I might have been better equipped, about 45 years ago at the end of high school education, to comprehend the scientific-rational basis of AI, how machines are fed, trained and enabled to retrieve, compile and present information in human language in vast neural networks using probabilistic methods, math, logical propositions, statistics and else. But by now, as priorities had evolved, I guess I have to settle with the vague (semi-accurate) understanding of the scientific terminology and algorithms, and be dependent on books by investigative journalists that look into the matter not so scientifically, or on popular science written by scientists for general audience. Among them is the short history of AI from Alan Turing to ChatGPT written by insiders (engineers/scientists) that is illuminating for a lay person. For now, the (practical) capabilities of the much touted AI revolution, for an average citizen like myself, appear to range from randomly useful (time saving) to trivial at best, from self driving vehicles to the mechanics of humanoid robots, and from medical diagnostics to intelligent home systems, to automation of manual tasks and to production of fiction writing, music and images. Some are annoying, like the automated call centers, and the most profitable applications are outright sinister and evil, as in surveillance systems and the self driving weapons of warfare. All these appear to be an insult and injustice to the physiology, intelligence, imagination and creativity of the collective of the living, a poor imitation supposedly of endless capabilities. Gaia ‘extends’ in time and space through thresholds, as in evolution, adaptation and mutation. We may not call this a progress, but with each transgression the collective’s field is expanded, and more possibilities open up the future. As in the science of chaos, when a threshold is overcome, the unpredictable and never-to-be-imagined comes to the surface. It is like the case in phase transitions. Unlike AI, this is never merely the sum total of constituent parts, or bound to the biggest data sets, or even the inevitable outcome of a cosmic intelligence. The thresholds of intelligence are like glitches, unpredictable like leaps, but never like a fork on the road to be decided as in a game of go, like retrieving a million similar situations from the past and computing a billion more future moves. For now, what is beyond human imagination regarding the Artificial Super Intelligence is the gigantic quantity of information at its disposal surpassing any combination and collaboration of humans, a colossus yet to produce a qualitative shift for the well-being of the collective.
Lately, it surfaced that book sellers from Europe and the USA are receiving huge orders from dubious sources. (3) It turns out that the books are used for ‘destructive scanning’, a term used for circumventing copyright laws to train large language models. As the books’ spines are sliced and the pages are turned into pulp after scanning, to the dismay of book lovers worldwide, the ‘Gargantua’ that is fed billions of dollars and all the information on earth is expected to spew all possible narratives in the future. When all data on the world wide web is consumed for the purpose, the ‘well curated facts’ and ‘captivating fictional narratives’ in books (printed before 2022, before infiltration of AI generated content) are the next thing. To treat the total library of texts, images, sounds and statistical data as ‘super memory’ is another insult to the memory of the collective, which, for millions of years accumulated mostly without a sign but a genetic code, a neuron in waiting, built in, in situ. We expect that the all-pervasive AI, this gluttonous entity is bound to implode sometime in the future, destined to choke on what it is endlessly fed, or become deranged. As I am neither the biggest fan of science fiction nor a believer in replicants taking over the world in some future time, I am not much worried. My favorite: 2001 A Space Odyssey, from 1968, I watched several times especially for the visuals (imagination) and the genius of the filmmaker and the accomplishment of his crew. The making of 2001 is well exposed in several documentaries: huge revolving rigs, together with revolving and steady cameras in the ‘zero-gravity’ set, a single light source (a star) in a dark-black studio as opposed to the extreme diffusion of light in the interiors, the articulated spacecraft, aligning of planets on glass sheets in a row, and the famous ‘stargate’ sequence (an extended acid trip) all achieved without the help of computing. Meanwhile, in the movie HAL (the computer) speaks in an emotional soft tone, and sees through a hi-fi speaker / lens / light modulator (Braun-design style,) more admirable than a chatbot and better than any AI slop, until it is unplugged and goes dark and mute.
August 2026

(1) Bruno Latour, Pandora’s Hope: Essays on the Reality of Science Studies, Harvard Un. Press, Cambridge, 1999. p.93
(2) Richard Dawkins, The God Delusion, Bantam Press, London, 2006
(3) https://www.theguardian.com/commentisfree/2026/aug/05/anthropic-ai-destroying-books

