Overview
- Recent reports compile accounts from software engineers who say they now spend most of their time reviewing AI‑written code rather than writing it themselves.
- Several developers describe concrete lapses in know‑how, including forgetting how to set up a common web framework task after relying on assistants for routine steps.
- A widely shared thread says AI has increased context switching, with teams juggling multiple chatbots and losing the deep focus needed for complex work.
- Senior engineers report reviewing more AI‑generated pull requests and less hands‑on coding, and some teams say quality has slipped after cutting manual QA in favor of tools.
- Companies are accelerating adoption with tactics like Meta’s internal leaderboards that track AI token use and quiet rollouts of tools such as Cursor, fueling fears of long‑term skill and design erosion even as most evidence remains anecdotal.