Stanford paper: general coding agents beat hand-built data agents by up to 37 points
CShorten30 · x · 2026-09-08
A new arXiv paper by Liana Patel, Ion Stoica, Matei Zaharia and colleagues tracks two years of frontier models on data-agent benchmarks: general coding agents now beat carefully hand-designed data agents by up to 37 points with 4× fewer turns. Arguing the Bitter Lesson is subsuming data-system engineering layers, the authors identify enduring research in 'persistent semantic context' — curated context about the data environment served as a first-class abstraction — and outline open problems in context data structures, storage, compression, and semantic consistency protocols.
Related event: Study Finds General Coding Agents Beat Specialized Data Agents by 37 Points(3 posts)→
More from coding & agent
- Astra falls short of Fable in hands-on test: no one-shot complex features — bindureddy · 2026-09-08
- Engineer Says He Automated Himself Out of One of the World's Most Technical Jobs — MickeySteamboat · 2026-09-08
- No-experience dev ships cozy game in 7 days with an all-AI pipeline costing ~$120/month — Gambo7592 · 2026-09-08
- Indie Maker's Skillry Sees Users Adopting Web Skills to Optimize Their Websites — yihui_indie · 2026-09-08
- dembrandt: extract any website's design system into W3C tokens in one command — tom_doerr · 2026-09-08
- Sentdex: GLM 5.3 Flash-level capability would've been overkill a year ago — Sentdex · 2026-09-08