On the map

At the centre A passing test suite is not evidence that the software works, so an agent must also be made to start the thing and exercise it the way a person would. Holds SWSimon Willison Read this korrent →

• Similar wording In an agent-written codebase the agent-written tests are equally untrustworthy; manually using the product is the only reliable measure of whether it works. Holds MZMario Zechner
• Similar wording Agent-written code that runs and passes its tests is not enough: security, maintainability and being able to roll it back still need humans in the loop. Holds TFTony Fadell
• Similar wording A language-agnostic conformance suite is the most powerful thing you can hand a coding agent, because the whole instruction becomes: write code until these tests pass. Holds SWSimon Willison
• Similar wording Letting a coding agent live-edit the production server works fine in practice. Holds PLPieter Levels
• Similar wording The code today’s models and agents write is very hard to follow: you get better performance without knowing why, and no way to fix it when it breaks. Holds ESElizabeth Stone
• Similar wording Writing code with a coding agent and no tests at all is indefensible, because the old objection to testing — that it is extra work you then have to maintain — no longer applies. Holds SWSimon Willison
• Similar wording Software engineering agents will work before any other kind of agent, because code is the one domain where the machine can check its own answer. Holds DPDylan Patel
• Similar wording AI agents and AI coding will run on servers and from the cloud first, not on your laptop. Holds PLPieter Levels