Frontier agents complete under 8% of tasks in 26-person simulated org, study finds
vasuman · x · 2026-11-21
- vasuman argues that what you see on Twitter (individual engineers loving Claude Code) is very different from what moves the needle at large companies.
- Quoted @jdpruettt, who is redesigning SDLCs for orgs with 700+ engineers, notes these companies rolled out Claude and engineers are more productive individually, yet see none of the speed startups feel — because "the software part is easy(er); you've got to reorganize people."
- Key evidence: the SWE-CoWork study found frontier models completed less than 8% of assignments when requirements were spread across a 26-person org of AI-simulated colleagues.
More from coding & agent
- Reverse-engineering Apple Watch: a single Claude Code prompt builds a physics watch face in 30 minutes — mtizard · 2026-11-21
- Agent RAG tweak: skip lossy summaries, extract answers, fetch 10 pages — 75% on validation — colinmcnamara · 2026-11-21
- us-legal-mcp: an MCP server exposing comprehensive US legislation to AI agents — modelcontextprotocol · 2026-11-21
- Maindex launches Expert Memory Service, a full-control memory API for AI agents — modelcontextprotocol · 2026-11-21
- GPT-6.2 Ultrafast in Codex: Up to 4.5x Faster Tokens, Dev Says It Restores Flow — dkundel · 2026-11-21
- AI-built oil painting engine: 2,300 commits in 8 days as Claude Code runs a 'studio' of sessions — FotoRe_store · 2026-11-21