Codex-generated tests are so bloated agents just delete and rewrite them
jxmnop · x · 2026-07-29
Codex-generated unit tests are getting so bloated in a large codebase that they become unreadable “Neuralese,” and when they fail, agents often just delete them and write new ones from scratch.
The poster jokes that they may soon have to declare “test bankruptcy,” highlighting a real maintainability problem in agent-generated code.
Related event: AI Coding Assistants Create Bloated Unit Tests, Sparking Concerns(4 posts)→
More from coding & agent
- 'LLM code is sloppy' rebutted with real ffmpeg code: identical if/else branches in a top library — kingslyj · 2026-10-16
- Agent actions need checks outside the model: NVIDIA demos lean on guardrails amid wave of real agent incidents — Creamy-And-Crowded · 2026-10-16
- Vitalik builds "paranoid mode" agent wrapper with Tor, zkAPI and local prompt rewriting — DavideCrapis · 2026-10-16
- Agent fine-tunes Qwen3.5-0.8B with synthetic data: 4.7%→76.3% accuracy, 11.9× faster — ben_burtenshaw · 2026-10-16
- Beam-501B agent demo: solo beats 5-agent team on ProgramBench rebuild task — ben_burtenshaw · 2026-10-16
- Beam agent builds BeamSpeak, a local Parakeet dictation app for macOS — ben_burtenshaw · 2026-10-16