Liang Wenfeng bet on Google's rise in May as DeepMind 'no longer frontier' debate heats up
teortaxesTex · x · 2026-11-19
teortaxesTex argues that predicting AI's future is easy if you watch fundamentals: DeepSeek founder Liang Wenfeng backed Google back in May. The quoted context packs two notable claims: a Tokenomics client brief asserting DeepMind is no longer a frontier lab due to RL-team departures and poor compute allocation ('Google's odds of reaching SOTA again have dropped to zero'), and Liang Wenfeng's interview comments that Anthropic's lead over OpenAI is cyclical, its coding-agent edge will vanish quickly (half his staff think OpenAI is better), and OpenAI and Google will keep alternating at the top. He also sees China likely becoming AI's largest producer, with a 'three-body' endgame. The author thinks Google is clearly no longer roadkill, though it could still fumble by reinforcing Anthropic with TPUs.
Related event: SemiAnalysis Claims DeepMind Is No Longer a Frontier Lab, Sparking Debate(2 posts)→
More from AGI Musings
- LLMs are language declaring independence from the human substrate: a viral poetic thread — aiamblichus · 2026-11-19
- Does Claude deserve co-authorship? Academics spar over AI-era authorship norms — paulnovosad · 2026-11-19
- Kevin Bass: US AI strategy should flexibly build on its companies' strengths, not cripple them — kevinnbass · 2026-11-19
- One GSI now covers 130-student classes with a self-built AI grader — hoofnagle · 2026-11-19
- Writing is thinking, and programming is thinking too — satnam6502 · 2026-11-19
- AI writing is recognizable through cliches — and that's exactly why it's bad — nsaphra · 2026-11-19