As Everything Melts

Artificial intelligence is going mainstream. It's no secret by now, and while I have unfounded reservations on it's viability, I cannot deny that it's a big deal. Advancements in the various technologies are coming faster than people can really process. Some people will point out the uncanniness to comment on the computers' lack of humanity.

But lately, I've been thinking that humanity is actually the uncanny party to this affair. These machine learning models are trained on unfathomly large datasets, which are derived from average (at best) human output. So the computers are just aping us, and doing a damn good job of it. Humans, on the other hand, aren't.

Back in November, when OpenAI released chatgpt, I spent a couple of hours playing with it, and spent perhaps a dozen more seeing what others were doing, and generally taking in the commentary of an interesting point in history. In short, I wasn't blown away, but impressed by it's consistent satisfactory responses.

It was only when I stepped away from the world of promtpt and response when I began to question reality. Everything I encountered felt like it was generated by a computer. But I knew that it hadn't, that we were still at least a few months from the technology being embraced at more than a tinkerer's scale. Real live humans were responsible for all of these uncanny words and images.

It's at that point you realize that this has been the case for quite some time, that our Work has become so generic and undifferentiable that it could be spit out after prompting the computer with a sentence or two. It's really nothing on a case-by-case basis, but as soon as you start thinking about the implications of this at scale, your mind can go in a million different directions.

At scale, you start to wonder what does this all mean? I'm not steeped in any particular intellectual tradition, but I know this question and it's permutations are asked by philosophers and computer scientists and linguists and beyond.

But, to me, I think some of the most interesting takes come from semiotics, the study of signs and what we make of them. Last month, I read Umberto Eco's Travels in Hyperreality, a compilation of essays written in the 60s, 70s, and 80s about the rapidly changing media landscape. It still holds up today, and the collection prompted me to start work on three separate essays, which I plan to publish next month.

While I've just started to explore, semiotics seems to have some substance that might help us (or at least me) answer some questions about humanity's relationship with both technology and itself.

I'll leave you with a quote I read this morning, not about semiotics but about modernity in general, from Marshall Berman's All That is Solid Melts Into Air:

To be modern is to live a life of paradox and contradiction. It is to be overpowered by the immense bureaucratic organizations that have the power to control and often destroy communities, values, lives; and yet to be undeterred in our determination to face these forces, to fight to change their world and make it their own. It is to be both revolutionary and conservative: alive to new possibilities for experience and adventure, frightened by the nihiliststic depths to which so many modern adventures lead, longing to create and to hold on to something real even as everything melts.