kapynResearch

AI agents have no sense of time and are not aware of it

AI coding agents overestimate task duration and self-assess too optimistically, a new study finds. Claude Code and Codex both systematically overestimate how long tasks will take, with Codex off by up to ten times. They rate their own work about 20 percentage points too high, complicating oversight for long autonomous tasks.

The Decoder·Aug 30, 2026

Opening Kapyn…