Running coding agents without approval prompts is a normalization-of-deviance risk: the fact that it has not caused a disaster yet is itself the problem.

Added to korrents 29 Aug 2026 · How quotes work · Something wrong? Tell us

Where this collides

• Cannot both be true

  OpenClaw is no more dangerous than Claude Code with permissions skipped or Codex in YOLO mode, which every engineer he knows already runs.

  Steinberger defends OpenClaw by pointing out that every engineer he knows already runs an agent with permissions skipped, because that is the only way to get things to work. Willison names that same fact as the danger: he runs in YOLO mode too, it has not burned him yet, and that is the problem. One treats widespread practice as evidence of safety and the other treats it as normalization of deviance.

Shared subject first, then similar wording. They may agree or disagree.

• AI coding agents advance open source's founding goal by letting far more people modify software, so barring AI-assisted contributions protects a guild rather than the mission. DHDavid Heinemeier Hansson1 Jun 2026coding agents
• AI coding agents have made redesigning existing systems cheap enough that developers actually redesign instead of living with a flawed design. KDKent C. Dodds16 Mar 2026coding agents
• AI coding agents pulling unknown code onto corporate networks will be exploited, and it will take the shape of a supply-chain attack. BSBruce Schneier4 Sept 2026coding agents
• AI coding agents work best with the same general-purpose tools built for humans, such as git and file utilities, rather than agent-specific interfaces. HRHarper Reed30 Sept 2025coding agents
• As a small design team, you should not stand in the way of engineers designing with coding agents but find ways to support them. NFNev Flynn15 Sept 2026coding agents

On the map

Open the map on its own page