How to deal with AI?

AI, computers and softwares in general, are demanding that we further loosen the relationship between language and thought. 

In fact computers and softwares are enabled by this loosening of this relationship, following the ground-breaking work of Alan Turing in 1936, when he designed an intellectual computer on paper, and then using this imaginary but working computer, he found the answer to his question: Are all numbers computable? He proved, from the computer he intellectually designed, that the answer is no: There are numbers that we can prove do exist, and at the same time we can also prove that they can not be calculated. One might suspect this question was in fact an excuse for designing this computer. 

Anyway, Turing was among the first to isolate in full details the thought process as a separate object, detached from any material or experiential reality, other than itself, and describing it with such accuracy that it can be materialised, represented with materials. What we have called later "the Turing machine" is the first time a material representation represented not an experiential concept like drawings or words, not an act of thought like a number, not a process like algebra, but the very act of representation itself.

Having isolated this act and represented it in a fully clear and complete way, almost like we isolate chemical elements or viruses (and after having built ad-hoc calculators for British military intelligence to successfully decipher the German submarines communications), Turing proceeded, in the last decade of his life, to build the first physical computer, capable of physically operating and not just on paper.

This isolation of the process of representation in a material object, demands that we enter in a new relationship with our thoughts. We must, if we want to stay free and maybe even alive, apprehend our thought in a way that is more detached from language than it was before computers and AI.

Printing and mechanical reproduction of images has made a similar demand, and before that, drawing and writing.

Each time, the language itself has become an object of language. 

Practically speaking, for architects today, AI and softwares demand that we loosen further the relationship between our thoughts and a given particular language. We must freely move from a software to another, from scale drawing to hand drawing, from calculation to drawing and back, and even from a material to another, and from a shape to another.

If we don't do that effort we run the risk indeed, of becoming trapped in the machine, as is the unavoidable goal of everyone working with AI and softwares: To make them ever more totalitarian -as Lucien Sfez brilliantly described with this word: Tautism, acronym of totalitarian, tautology, and autism, because this constant move towards totalitarism is the only way to make softwares more efficient, and that's not a political choice, it's an unavoidable necessity of the very concept of computers or "Turing machines".

As it turns out, AI which seeks to weaponise our identities, to lock us into our languages, is creating a situation where we must do a conscious and deliberate effort in the opposite direction, and when we meet someone, find ways to understand that person beyond her identity, find a path to that person beyond her and our language. This is a way to de-weaponise our languages and our identities, like a narrow, meandering and dangerous, but very effective, path towards practical love, in the middle of the forest of our misunderstanding and hate, on which AI feeds itself.

Because beyond our languages and our identities, we do not have much else, and by harnessing them, AI holds in its power our core. Destroying our languages and identities would be destroying ourselves. So we have to love them instead, following this narrow path to the other person, this path that AI, like these numbers that Turing has shown exist but cannot be calculated, is unable to understand.

Hats off, Alan Turing! Your first paper was the best.

https://www.cs.virginia.edu/~robins/Turing_Paper_1936.pdf