Post #1736273
2026-04-27 15:32 UTC
I'm not going to read this guy's paper, it sounds awful, but I disagree with some points. If humanity survives long enough, we will create AGI. Consciousness is just not that complicated. But it won't be a direct descendant of an LLM
https://www.404media.co/google-deepmind-paper-argues-llms-will-never-be-conscious/
Replies (1)
-
@ebooksyearn@thepit.social 2026-04-27 15:34
No matter what OpenAI and Anthropic tell you, they aren't close to breakthroughs because the LLM model is a dead end. I think we'll still see small improvements, and it won't be too long before your phone will run a high end model without much trouble. But LLMs are close to their ceiling