I have recently been reading some of Montaigne’s essays again. I think that there is a lot in his work that is relevant to present discussions about AI. For example, he frequently warns against anthropomorphism. Could engaging with his essays and learning more about his reservations in this respect offer us an interesting perspective on our own practice of doing so in the case of AI? I think so, and I might explore this and other topics at a later point. Here I’ll be interested in something a bit more fluffy.

The version of Montaigne’s works that I have come with a great introduction by M.A. Screech.

Screech convincingly shows how Montaigne’s writings is not just an self-love and obsession but a contribution to human self-understanding through the of one individual man, himself. Screen writes

Montaigne strove to follow the Delphic injunction “Know Thyself”. He sought to discover the personal, individual, permanent strand in the transient, variegated flux of his experience and sensations, which alone gave continuity to his personality.

And continues:

This was not a merely subjective indulgence. By studying his own individual form (his soul within his body) he aspired to know Man - not just one odd individual example of humankind

One might be tempted to think that this is a preset-day interpretation of Montaigne’s work that lends it an, but as anyone that has spend a minute reading the essays will know, Montaigne’s work is littered with references to classical antiquity and Montaigne knew his Plato and Aristotle as well as anyone.

Here’s Screech again:

Aristotle taught Montaigne that individual persons belong to a genus and a species; so each man and woman individually possesses ‘generic’ and ‘specific’ qualities; and each of them has a specific human soul (or ‘form’); it could vary in quality but not in nature. So any man or woman who remained human could at least partially understand any other, since all possessed a like soul. No virtue or vice known to any individual human who remains sane should be totally incomprehensible to any other. Even the virtue of Socrates can be momentarily glimpsed and indeed momentarily shared in, by a lesser member of this species. So too could the cruelties of Tamberlane be understood by better men. All individual human beings (as the scholastic philosophers put it) bore in themselves the entire ‘form’ of the human race. To study one man is in a sense to study them all. Not that all are identical but all are interrelated by species.

With a biological constitution and

Here’s what I’ve been wondering: Could an AI learn something like the human form in this sense? Sure, it would be able to learn a whole host of things about humans from reading our writings, consuming other cultural artifacts and through its interactions with us via chat or some other medium, but I there is a sense in which AIs will never be able to know what it is like to be human.

Caius is a man, men are mortal, ergo Caius is mortal

Humans differ from non-human animals in many respects and to a greater or lesser extent in these respects, but humans remain uniquely aware of their own mortality. When we are most ourselves relative to other species, I think it is this quality that makes us stand out the most.

I think this is the ultimate condition of the possibility of empathy. Real empathy is only possible because without this understanding we couldn’t make sense of

We could understand how something could be meaningful to someone

Absurd

Without a

Montainge mentions vices, but vices are only intelligble against a horizon of imminent death and a hope for salvation. If there is no hope for salvation, n

That to Philosophize is to Learn How to Die

Knausgaard

Megaloman livsloegn