As LMs mature, the work reduces to two things: data and infra
saurabh_shah2 · x · 2026-09-16
A short industry observation: now that LMs largely work, the remaining work is only two things — data and infra — implying architectural differentiation has narrowed.
More from Infra
- RX 7900 XTX beats R9700 by 30% on GPT-OSS-20B local inference tests — glenbeer · 2026-09-16
- SemiAnalysis: Vera Rubin NVL144 hits ~7x tokens per MW vs Blackwell, over 2x profit per GW — sudoraohacker · 2026-09-16
- Musk explains why Terafab must exist: Taiwan chip risk plus capacity ceiling — elonmusk · 2026-09-16
- DeepMind's guide to sharded matrix multiplication: the math behind training LLMs on 10k TPUs — zacharynado · 2026-09-16
- YuE2 music sampling only hits ~7 tokens/s on RX 9070 despite mostly idle VRAM — Prestigious-Kick7291 · 2026-09-16
- StartLux Interview: A 27B Local Model Rivals DeepSeek Giants via AutoResearch and Local RSI — 机器之心 · 2026-09-16