1/ While most RL methods use shallow MLPs (~2–5 layers), we show that scaling up to 1000-layers for contrastive RL (CRL) can significantly boost performance, ranging from doubling performance to 50x on a diverse suite of robotic tasks. Webpage Paper Code: wang-kevin3290.github.io/sca…

1/ How should RL agents prepare to solve new tasks? While prior methods often learn a model that predicts the immediate next observation, we build a model that predicts many steps into the future, conditioning on different user intentions: chongyi-zheng.github.io/info….

Ilya Sutskever, in his speech at UToronto 2 days ago: "The day will come when AI will do all the things we can do." "The reason is the brain is a biological computer, so why can't the digital computer do the same things?" It's funny that we are debating if AI can "truly think" or give "the illusion of thinking", as if our biological brain is superior or fundamentally different from a digital brain.

Richard Sutton says that the current dominance of LLMs is a "momentary fixation" the real breakthroughs will come from scaling computation, not building AI systems based on how humans think they work. LLMs will not be the leading edge of AI for more than another decade, perhaps even half a decade.