“Meno” is a funny one by Plato, where Socrates asks the boy, a slave of Meno’s, a series of highly specific geometric questions, to receive, as Socrates typically would do, “Yes” or “Certainly” as an answer to each. “Do you observe, Meno, that I am not teaching the boy anything, but only asking him questions; and now he fancies that he knows how long a line is necessary in order to produce a figure of eight square feet; does he not?”, Socrates asks then, prompt-engineering Meno himself to agree with him as well.
Later, Socrates is a bit more earnest. He argues that if the boy did not always possess this knowledge "in his soul," he’d never be able to awaken that knowledge through his prompts. This notion potentially places the slave-boy on the same level as the learned men, implying that knowledge – and virtue – can be taught to anyone without being mediated by the “orhte doxa”, “the customs by which people live”. This idea is deeply unsettling for Anytus, who leaves the dialogue in a fury, to write an open letter that will later lead to Socrates’ indictment. To him, Socrates, who attempts to teach virtue in a somewhat abstract manner without deferring to doxa, poses a threat and is corrupting the already fragile social order.
Lacan sees here a pivotal moment in which a new episteme is proposed: one not of “customs”, but of the inner consistency of thought. All this Socratic probing and pondering, trying to lead people to contradictions, is not done simply to annoy them, but to explore the coherence that nobody really cared about before. “The notion of knowledge as tied to certain requirements of coherence has arisen, knowledge which is the prerequisite for any future progress of science”. This “coherence”, which we now regard as a self-evident virtue, is actually a rare and elusive quality that may not be useful or even feasible for many human endeavors. However, when taken seriously, it serves not only as a prerequisite for science but also as a guiding principle for its advancement. It’s a bet: in the face of chaos and weirdness of the incoherent world, hanging to that ideal is a dangerous path.
Now we’re talking to AI like Socrates to the slave-boy. We put words in its mouth, we get sensible responses. No matter how lowly it seems to us, we can’t, and shouldn’t, dismiss the impression that it has its own inner capacity for truth. I feel that a certain historical arc has reached its conclusion. Something was lost back then – this innocent ignorance of doxa – it was repressed to form a scientific ideal. It now returns as this great machine for customary and cheap public opinions, trained on the biggest compendium of doxa we could imagine. As we started to take science for granted, from a weird and dangerous bet on the sensibility of the world, it ended up a dogmatic social ritual. The nature of the AI's episteme, the new form of knowledge it can generate, remains to be understood.