Critique of the Power of Drawing

A hidden connection exists between drawing and mathematics - it involves an abstract ability (even in the most realistic drawing - the abstraction is to forget what you're drawing and draw what you see: not to draw a woman but a shape and a stain). Therefore, they are perceived as masculine professions, like philosophy. But in fact, these are two opposite directions of abstraction: one analytical and the other synthetic. That's why the problem of computer vision is at the bottom of the artificial intelligence problem, while the problem of computerized proof is at the very top of it

The most important thing in drawing is what not to draw (Source)

It's unbelievable how bad we are at drawing. A camera is not better than us at seeing, it doesn't see anything - it's better than us at drawing. Even an expert artist constantly makes mistakes in sketching - in angles, for example - and relies on various external things and technical aids. In fact, we don't know at all how to recreate, even in an embarrassing way, what we've seen, if we're asked to do so and draw what we saw a moment ago or even imagine it. Even an artist constantly needs to look at what is seen and see new things there.

In other words, we see very little of what we look at. And the artist tries to recreate access to the low-level of vision that is blocked for us. Constantly overcoming perception. Accessing the hardware of the brain before the software, and trying to get closer to the data. But it's even more severe than that - an artist is constantly cheating, trying to use the high-level of perception to disguise their failures at the low-level. An artist answers the question of what there is to see in the world. What is worthy of seeing. Therefore, the critique of judgment is an extension of the critique of pure reason. The artist's work is anti-Kantian work, anti-perceptual, a practice of personal war against epistemology - and against categories of perception. It is a struggle of the brain to descend as far down as possible - to machine language and hardware.

But precisely the opposite direction, upwards - is the direction of the future. Art will not be the supreme test of artificial intelligence, but rather the most difficult conceptualization at the highest level, without any connection to the lower levels, or to the connection between the lower and higher levels - which is the pride of humans. A computer that is unable to see or has no connection to the sensory world - that is what will humiliate humans. The killer application of artificial intelligence will not be the Turing test, or consciousness, or something like that, but when it knows how to prove things in mathematics better than humans. And then when there are no mathematicians, science will quickly be done by computers, and humans will return to a medieval state where they don't understand reality, and its magic is explained to them by others and not deciphered by them. A computer mathematician better than a mathematics professor - that will be the end.

In other words, the danger is not that a computer will surpass humans in the power of judgment, but in pure reason. And then it will be a source of intelligence. In the end, all sciences, and all faculties at the university, know that there is one that is harder than all of them, and at a higher intellectual level than all of them, and that the closer they get to it, the more successful they are - mathematics. Even philosophy, the queen of humanities, is humbled by an encounter with mathematics and bows its head. Therefore, it is the peak of the human brain, and if the computer surpasses us - it will only need to conquer the peak.