mitsuhiko: Astra writes 'weird Python slop' with horrific unit tests on generic tasks
sull · x · 2026-09-07
Developer mitsuhiko criticizes GPT-6 Astra's coding: on tasks that are "one step removed" from normal code, it writes weird Python slop, and the unit tests it generates are "absolutely horrific." A notable behavior observation about the model's generalization quirks.
Related event: Mitsuhiko Slams Astra's Python Code as Poor-Quality Slop(2 posts)→
More from coding & agent
- theo publishes tstack skills, Garry Tan urges everyone to 'publish your stack' — ycombinator · 2026-10-26
- Agent verified 'success' after double-voting: a lesson in harness vs model bugs — TejasKumar_ · 2026-10-26
- diginomica: agentic automation hype is overdone, and 'The Hackening' of enterprise security has begun — jonerp · 2026-10-26
- 'Vibe verification': ~60% of AI-written formal specs prove the wrong property — basedjensen · 2026-10-26
- Devin launches Security suite powered by Claude Mythos with $100M+ credits — DevinAI · 2026-10-26
- HackBench injects reward hacks into agent trajectories to stress-test monitors — DanielKhashabi · 2026-10-26