Compute vs data: X debate over what really drives the US-China AI gap
teortaxesTex · x · 2026-11-21
X users debate the root of the US-China AI gap. Leroy Li cites the view that 'the real difference comes down to resources — talent, model capabilities, and application gaps all stem from a compute gap,' arguing data isn't the core issue. teortaxesTex pushes back: Chinese labs are getting far less out of 30-50 trillion pretraining tokens than 'Inkling' got from <15 trillion, so raw compute alone can't explain capability differences.
More from AGI Musings
- AI writes 90% of code but plateaus at ~50% of online articles — bendee983 · 2026-11-21
- Jack Dorsey says Block is 'a lot flatter' with 3 core roles after AI-driven restructuring — shivsingh · 2026-11-21
- Superintelligence could solve cancer treatments theoretically before wet labs — Dr_Singularity · 2026-11-21
- Yang You's 'Large Partial Derivative Theory': pick research where your marginal contribution is biggest — YangYou1991 · 2026-11-21
- Equity vs credit: why AI bubble math looks different to VC and lenders — sebkrier · 2026-11-21
- Looking back: OpenAI reasoning model autonomously disproved Erdős's unit-distance conjecture — aran_nayebi · 2026-11-21