Ex-OpenAI: Most Users Are Very Bad at Describing What They Want
prajdabre · x · 2026-08-30
Former OpenAI employee Praj revealed that after reviewing PII-scrubbed Gemini user logs, the key takeaway is that most users struggle to articulate their requests clearly.
This observation aligns with the cited frustration that while coding abilities improve with new models, instruction following remains unreliable—highlighting a gap between model capabilities and user prompt engineering skills.
More from Models
- OpenAI stealth-testing mysterious internal model 'Aeon' in Codex desktop app and Dots — cedric_chee · 2026-10-16
- Argon edges out astra as daily driver, but Claude still wins heavy coding work — iruletheworldmo · 2026-10-16
- NVIDIA ships FP4-quantized Qwen3.8-Flash for local use, trending on Hugging Face — nvidia · 2026-10-16
- RoboVista benchmark: Gemini-4 Argon hits 81.9% on robotics understanding — berkeley_ai · 2026-10-16
- Dev reports decision model jev beats gpt-6-luna on quality in real training pipeline — genmon · 2026-10-16
- Google won't publish Argon 4 limits in AI Pro—users compare experiences — Consistent-Text-2662 · 2026-10-16