Jon Stokes: frontier gains come from bespoke post-training data, not runaway generalization
ZeroStateReflex · x · 2026-11-19
Jon Stokes argues that public data showing most frontier-lab spend going to post-training confirms his earlier thesis: frontier model capability comes from high-effort, bespoke data collection rather than runaway generalization. Meta and xAI failing to keep pace — and Chinese open-weights models lagging months behind — suggests OpenAI, Google and Anthropic's edge is their massive live user sessions for RL training. The resulting intelligence isn't smooth and general but a 'point cloud of special cases' that big labs densify with more data.
More from AGI Musings
- Artists Must Understand AI Deeply to Have Sway in Policy and Safety Debates — amplifiedamp · 2026-11-19
- Nagel wins $1M Berggruen Prize, AI consciousness remarks spark pushback — connoraxiotes · 2026-11-19
- Philosophers spar over AI consciousness in Nagel prize thread — connoraxiotes · 2026-11-19
- Anthropic's 'trust us' strategy bound to fail, says researcher amid IPO scrutiny — natolambert · 2026-11-19
- As execution gets cheaper, understanding gets more valuable: the epistemic debt problem — lililashka · 2026-11-19
- wordgrammer essay: passion can't be outcompeted — great talent unifiers beat peer-group sorting — wordgrammer · 2026-11-19