veerd
2. If you want something with less physical grounding, you could use lambda calculus instead of Turing machines.
3. Kolmogorov Complexity and Shannon Entropy are compared with one another because they both are talking about the same thing: optimal compression. Kolmogorov Complexity talks about the compressibility of individual objects and Shannon Entropy talks about compressibility of streams of i.i.d. random variables.
If instead you want to consider the highest status/most famous people working on AI in general, then the list of signatories here is a pretty good summary. From my flawed perspective as a casual AI enthusiast, Yann LeCun and Jürgen Schmidhuber are the most glaring omissions (and both have publicly stated their lack of concern about AI x-risk).
Of course, the highest status people aren't necessarily the most relevant people. Unfortunately, it's more difficult for me to judge relevance than fame.
Both of them are criticizing their own life's work and the source of their prestige. That has to be emotionally painful. They aren't doing it for fun.
I totally understand not agreeing with AI x-risk concerns on an object level, but I find the casual dismissal bizarre.
I mean... that's not even debatable. Geoffrey Hinton and Yoshua Bengio aren't financially motivated to talk about AI x-risk and aren't ignorant idiots. In fact, they both have natural reasons to not talk about AI x-risk.
Scaffolded LLMs as natural language computers
Bioelectricity: A new perspective on cancer as a disease of geometry
I haven't seen anything like this elsewhere.