@danluu We've already seen a wave of slop report spam come and go in bug bounty programs. IMO they were the first because of the financial incentive. In the near future I expect waves of PRs in FOSS which implement an actual missing feature or fix an issue. They won't be obviously wrong. They will make sense in the smaller scope but won't take the wider architecture into account. Just because average code quality is the best we can hope for when we feed the code available online into an LLM.