What Is ChatGPT Doing … and Why Does It Work?—Stephen Wolfram Writings

Wolfram is an expert at many things, including neural nets and the foundational functions of AI. This writeup is very long (I skimmed it and plan to give it a full read later), but the clarity of the models, the math, and how these Generative AI's and LLMs work is fabulous. I love his analysis of our, meaning us humans, view of the things that ChatGPT is doing (bolding my addition).

But there’s something potentially confusing about all of this. In the past there were plenty of tasks—including writing essays—that we’ve assumed were somehow “fundamentally too hard” for computers. And now that we see them done by the likes of ChatGPT we tend to suddenly think that computers must have become vastly more powerful—in particular surpassing things they were already basically able to do (like progressively computing the behavior of computational systems like cellular automata).

But this isn’t the right conclusion to draw. Computationally irreducible processes are still computationally irreducible, and are still fundamentally hard for computers—even if computers can readily compute their individual steps. And instead what we should conclude is that tasks—like writing essays—that we humans could do, but we didn’t think computers could do, are actually in some sense computationally easier than we thought.

In other words, the reason a neural net can be successful in writing an essay is because writing an essay turns out to be a “computationally shallower” problem than we thought. And in a sense this takes us closer to “having a theory” of how we humans manage to do things like writing essays, or in general deal with language.

If you find this topic interesting this is a super read, and it helps clearly show the boundaries of this technology. I also love the idea that there are problems that "computationally shallower" than we really think. I wonder how our own view of our unique human-ness plays into that perception.