AI isn’t “just predicting the next word” anymore

kept by eddie jan 09

AI has evolved beyond "just predicting the next word" through reasoning models and RLHF, achieving gold-medal math olympiad scores and exhibiting dangerous behaviors that demand serious risk assessment.

‘The biggest decision yet’: Jared Kaplan on allowing AI to train itself | Technology
Anthropic's chief scientist Jared Kaplan says humanity must decide by 2027-2030 whether to let AI systems recursively self-improve, a step that could trigger beneficial intelligence explosion or cause humans to lose control.
also on Google DeepMind, xAI, #ai-safety, OpenAI, Anthropic, #software-engineering
If the Superintelligence were near fallacy — LessWrong
Apparent contradictions like OpenAI selling ads don't disprove imminent superintelligence because AI labs must fundraise and hedge against normal-tech scenarios to win the race.
also on Google DeepMind, xAI, #ai-safety, OpenAI, Anthropic, #ai
On Dwarkesh Patel's 2026 Podcast With Elon Musk and Other Recent Elon Musk Things
Elon's podcast reveals confused AI alignment, plans for space data centers and human obsolescence, while xAI's safety team has fled and he attacks critics.
also on xAI, #ai-safety, #ai-research
DSHR's Blog: Distilling The Moat
The article argues that AI companies’ claimed moat from expensive model training is shallow because model distillation allows competitors to cheaply replicate capabilities via repeated API queries.
also on xAI, #ai-safety, OpenAI, Anthropic, #ai