LLM performance plateaus confirmed by David Silver and Richard Sutton

June 30, 2025

🗓️ Posted on: 30th June 2025

📖 Research nugget

Top DeepMind veterans David Silver and Richard Sutton argue that large-language-model progress is stalling because we’ve exhausted high-quality text data, as highlighted in their “Welcome to the Era of Experience” paper.

They argue that the to get to superintelligence AI agents must learn from experience.

Note: How all the big tech companies have stopped using AGI and switched to the more achievable term superintelligence.