"Best model ever, yet I have to babysit it": dev bemoans agentic reliability gap
yacineMTB · x · 2026-09-06
Well-known AI builder yacineMTB called a model "the best ever," then immediately asked what the point is if it still requires constant babysitting — highlighting the gap between frontier model capability and the autonomy/reliability users actually need for hands-off agent work.
Related event: Yacine: Even the Best AI Model Still Needs Constant Babysitting(2 posts)→
More from Models
- GPT-6 Astra users hit frequent 'model at capacity' errors — bytebot · 2026-09-06
- Astra User Says Weekly Token Limits Too Tight: 24-Hour Run Still Unfinished — Wooden_Drag9473 · 2026-09-06
- Researcher says OpenAI's Astra censors his newsletter, fueling safety-sandbagging concerns — sethlazar · 2026-09-06
- Deleting most Skill rules made GPT 6 better, sparking 'Skill is dead' debate — xiaohu · 2026-09-06
- Unverified: GPT-6 Astra matches GLM-5.3 max reasoning with zero reasoning tokens — BLUECOW009 · 2026-09-06
- Dev demo: OpenAI's Astra agent detects and fixes failed states via voice command — BLUECOW009 · 2026-09-06