I first realized how far things would change when a friend told me what was then an unusual story. In the middle of 2023, his daughter—no Virginia Woolf even by my friend’s account—had come down to announce that she had written a novel. What she meant by this was that she had got ChatGPT to draft the chapters, turned it into a PDF, and published it on Amazon.
Is this the future of writing and intelligence? When a senior academic explained to me that his role in the world of AI was to make sure that the information was correct—to be the ultimate overpaid factchecker—it became clear that the game was lost. If you pay any heed to the zeitgeist, you will be tempted to conclude that human ingenuity is already reduced to pathetic scraps: knowing which questions best to ask of the machine—henceforth considered the ultimate arbiter of morality and historicity.
The belief that AI will provide a solution to all humanity’s ills is surely evidence that the world’s entire media and political class is suffering from Stockholm Syndrome. The technologies of bombs, planes, and the internet haven’t improved our prospects of surviving the 21st century as a species—but AI will sort us out! Perhaps that will be by rendering life so dismal and pointless that human beings will simply stop having babies: without intelligent use of our faculties and awareness, human existence is indeed colorless.
The technologies of bombs, planes, and the internet haven’t improved our prospects of surviving the 21st century as a species—but AI will sort us out!
Those of us who have had to assess student essays produced by AI can see where this is going: “It’s disheartening to see students rely on AI for essays, undermining their creativity, critical thinking, and personal growth while compromising the integrity of education”—a sentence written by ChatGPT, of course. Future text appears certain to be filled with banal clichés that appear decisive but are empty of the volition and moral intent which characterizes human judgment.
Advocates of the new technology contend that nothing essential has been lost. This is just a continuity of the history of technology, and useless Luddites like me have always opposed progress. There is, it turns out, no accident in the fact that progressives have been among the most ardent advocates of this new turn, heralding digital utopianism—which is merely the praxis of technological renewal by a sexier name. Yet the historical reality seems to be that every new form of technological mediation of the world has increased the violence of our relationships with the planet and with one another—as our separation and isolation intensify.
Certainly, the loud noise around AI merely repeats the pattern of centuries. Financialized capitalism proceeds through investing in new technologies which will themselves be superseded by further future opportunities for “investment”, rendering the previous technology irrelevant. It’s the canals-trains-roads-aeroplanes-Zoom-AI model—but it’s also an entirely deterministic view, which strips human beings of agency and moral choice, rendering us mere playthings in the hands of technological manipulators.
With the irrepressible rise of this new form of mediation between humans and the world, what is vanishing before our eyes is the way of being inhabited through the old mediation of reading. Any teacher of young people knows that the ability to read systematically, intensively, and with concentrated intention is already degraded. Of course, there are still a lot of young people desperate to get off screens, and keen to discover things for themselves. But all the same, the direction of travel is clear: AI can summarize articles in 150 words, so there’s no need to prepare seriously for class (although a student did admit to me that she was surprised to find that one summary entirely misrepresented what had been written). Struggling to think out an essay plan (i.e., to form your own opinion)? You can ask AI to set your direction for you—no need to worry if the algorithms may reproduce “the answer” preferred by the richest corporations on earth! How does that fit with “electronic activism”?
Some may say that old skills do indeed become redundant, but that is just human history—and that to privilege reading is an elitist and a very Western view of knowledge. But this would be to misunderstand the history of scholarship. Across the world, this practice of reading has been the foundation of forms of knowledge for two millennia: in the bark-fibre manuscripts of the Malay Archipelago, the manuscripts of Timbuktu and stelae of Aksum, the libraries of Persian medical treatises found in medieval China, and the Mesoamerican codices. These productions of human artifice may no longer offer investment opportunities and the mirage of accumulation—and that may be why the intentionality behind their production can haunt us still, precisely because this is part of what is beautiful about them.
Some may say that to privilege reading is an elitist and a very Western view of knowledge. But this would be to misunderstand the history of scholarship.
Of course, humans will find beauty in the new forms, as they always have. Nevertheless, in this struggle to redefine the representation of the world, AI extracts value from text and renders it into inert matter to be parsed mechanistically. As Achille Mbembe notes in his new book Brutalism, the boundary between humans and machines has already become blurred. Once knowledge itself is rendered mechanistic and free of human subjectivity, the machine will render human subjects still more alike to itself—devoid of intention, and with no moral purpose. Yet the entire framework merely recreates a supreme moral arbiter, with a belief in the dictates of a mechanistic divinity replacing the preceding belief in the dictates of an actual divinity.
When did this shift really become irreversible? Here the history of the Covid pandemic is worth revisiting. For this was in some ways a crisis itself generated by AI, with bots joining in on social media to promote outcomes that were entirely contrary to the interests of the human. If there are any human historians left in three or five centuries time, they may look back on those painful years as the time where the irreversible shift came.
More likely, AI historicity of that future time will puzzle over this arcane and elitist practice of “reading” and “writing”, never fully understanding it—presenting it as something that progress and technology overcame, leaving humans the space to do the really useful work of adding value to the technological algorithm.
Deus is no longer ex machina, but is the machine itself. This transition may exemplify the human condition, but at least we should mark it as it passes.
