it’s fine as a component in a larger (neurosymbolic) system, which is at last what Anthropic figured out, with Claude Code. but did in fact get stuck when used on its own, as i had predicted.
@garymarcus
-
Gary Marcus: Neurosymbolic AI saves deep learning from wall
By
–
L'IA neurosymbolique sauve l'apprentissage profond de la collision contre le mur (*exactement* comme je l'ai dit dans mon article de 2022 « l'apprentissage profond heurte un mur ») Vraiment triste de voir quelqu'un d'aussi intelligent @peterwildeford confondre l'argument
-
Anthropic: Mixed regulation, anthropomorphic language, and IP theft
By
–
Anthropic is really a mixed case; they rightly support moderate regulation, but they encourage wildly anthropomorphic language about LLMs, and steal IP just like their peers.
-
Gary Marcus: LLMs distract from the path to AGI
By
–
In my opinion, he has been too distracted by LLMs, due to competitive pressures. And these are not the royal road to AGI.
-
Gary Marcus doubts 50% white collar job loss in two years
By
–
not saying that nobody will lose their job (and in many cases AI is just a fig leaf). but i doubt eg we will lose 50% white collar jobs in two years.
-
LLMs often make illegal moves; article’s validity doubted
By
–
LLMs often make illegal moves; i don;t know that this article held up.
-
Gary Marcus criticizes neural networks and deep learning misnomers
By
–
also “neural networks” that have almost nothing to do with actual brains, and “deep learning” that isn’t not that (conceptually) deep.
-
Gary Marcus contrasts past AI science with current greed
By
–
Disagree. Certainly there has always been hype, but not this kind of unabashed greed or indifference to social consequence. People always courted funding to be sure, but people like Minsky and McCarthy (I met both) were primarily interested in science and ideas; money was
-
This poses a problem for the near-AGI narrative
By
–
This creates problems for @Kevinroose's narrative that AGI is near.
-
Marcus: AI nearing 2020 predictions but long way to go
By
–
it’s getting closer to what i described in my 2020 article the next decade in AI, which you should read, but we still have a long way to go.