Tom Zahavy
PulseAugur coverage of Tom Zahavy — every cluster mentioning Tom Zahavy across labs, papers, and developer communities, ranked by signal.
-
AI's scientific discovery potential debated: automation vs. embodiment
Two recent perspectives challenge the capabilities of current AI in scientific discovery. One view, exemplified by Jeff Dean's new venture Discovery Loop, posits that AI can automate the entire scientific discovery cycl…
-
Google DeepMind researcher: AI models stuck in statistical rut, lack creative leap
Tom Zahavy from Google DeepMind has stated that current language models are stuck in a statistical rut. He believes these models lack the creative leap necessary for genuine scientific breakthroughs. Zahavy suggests tha…
-
Google DeepMind researcher: LLMs lack innovation spark, world models may succeed
A Google DeepMind researcher argues that current language models are incapable of initiating scientific revolutions due to a lack of necessary cognitive mechanisms for true innovation. The researcher suggests that "worl…
-
LLMs like Kimi k3 lack creative reasoning, mirroring Einstein's era arguments
The author argues that current large language models (LLMs), even those with massive parameter counts like Kimi k3, are fundamentally limited in their ability to perform true creative leaps or resolve contradictions. Th…