It's Hard to Feel the AGI

kept by eddie dec 01, 25

Leading AI researchers argue that current LLM scaling will hit a ceiling, requiring new insights before AGI emerges, pushing timelines back 5-20 years.

AI infrastructure in the "Era of experience"
As open-source LLMs commoditize due to fierce Chinese competition, companies can build defensible moats through custom RL models trained on proprietary environments using LoRA and GRPO.
also on Rich Sutton, #ai-research
Sam Altman may control our future – can he be trusted?
A New Yorker investigation reveals that OpenAI board members secretly documented a pattern of lying by Sam Altman, leading to his brief ouster in 2023 due to concerns he could not be trusted with superintelligent AI.
also on Ilya Sutskever, OpenAI
Technological Involution
Technological progress has stagnated as society lost its ability to imagine novel frontiers and founders now chase institutionalized narratives rather than deep personal conviction in technology.
also on Ilya Sutskever
Judgment isn't uniquely human - by Steven Adler
AI already possesses judgment in domains like investment banking, ethical dilemmas, and art, despite persistent claims that such skills are uniquely human.
also on Yann LeCun, #ai-research, OpenAI, #ai
AI for the Real World: A conversation with Yann LeCun (via @AnneliesGamble)
Yann LeCun argues that current large language models are a commercially viable local maximum, not the path toward human-level machine intelligence.
also on Yann LeCun, #ai-research, #ai