Study: General Coding Agents Beat Hand-Designed Data Agents by 37 Points With 4x Fewer Turns
CShorten30 · x · 2026-09-08
A new paper from lianapatel and colleagues tracks the evolution of data agent capabilities over the past two years and finds general-purpose coding agents beat hand-designed data agents by up to 37 points, using 4x fewer turns.
The authors frame it as "the bitter lesson is coming for data agents too": as better models rapidly eat the stack, the key question for system researchers is what problems remain worth solving to enable the next frontier of data agents. A full thread and paper link accompany the finding.
Related event: Study: general coding agents beat specialized data agents by 37 points(4 posts)→
More from coding & agent
- Cloudflare Durable Objects validated as top architecture for AI agent WebSocket servers — tristanbob · 2026-10-27
- GitHub Next's Chopin prototype explores real-time multiplayer agentic planning — max__drake · 2026-10-27
- TMLR paper: enforced thinking doesn't reliably make LLM agents disclose key info to users — SharonYixuanLi · 2026-10-27
- LLM-written Rust library bbtree verified: speed claims hold, memory savings reverse — lmoroney · 2026-10-27
- Resin lets an agent build standalone LLM inference engines; vLLM and SGLang optimize DeepSeek kernels differently — AccBalanced · 2026-10-27
- How I let my cloud Grok Bot code on my Mac mini via Tailscale, no public ports — talkaboutdesign · 2026-10-27