Voice mode breaking up, Codex erroring: reliability still far from AGI-ready
koltregaskes · x · 2026-09-16
Commentator koltregaskes argues reliability is a key "floor capability" that must be solved before AGI, and current products fall far short.
Citing his own experience: ChatGPT voice mode stutters, stops, and lags; Codex still throws errors, runs slowly, and even loses visible message exchanges; usage limits remain a major pain point, with no fallback tier once credits for voice, chat, and schedules run out.
He also notes Codex and ChatGPT have yet to be merged, and expects some issues to ease once the merge completes — but for now, the experience is still a struggle.
More from coding & agent
- Muse agent navigates Amazon support to recover $70 in credit on its own — armand_ruiz · 2026-09-16
- 20 Astra Commons Codex community events remain worldwide before Sept 30 — paw_lean · 2026-09-16
- Event-driven multi-agent architecture beat model routing on cost, 32-agent finance system shows — tradfyAi · 2026-09-16
- Where do you stop trusting an AI agent with write access to Salesforce, SAP and Jira? — ken_kauneki10 · 2026-09-16
- Hands-On LLMs Authors Talk the Shift From LLMs to Agents at PyData — MaartenGr · 2026-09-16
- One prompt on a local Qwen model and an RTX 3090 built a playable HTML zombie FPS — LankyGuitar6528 · 2026-09-16