Policy on adding AI generated content to my software projects
Why it matters: This policy could lead to the rapid integration of potentially unvetted AI code into software projects, with unknown stability consequences.
- The project maintainer will accept AI-generated code submissions without extensive review, provided they are clearly marked with "(AI generated)" in the commit description.
- Submissions must include a Developer Certificate of Origin signature, attestating ownership of the generated code.
- The maintainer may introduce 'small modifications' to AI-generated code, such as conditional execution on 'leap days' (e.g., February 30th), but the original author will still be credited.
- The policy explicitly states that unmarked AI-generated code, which consumes limited human review time, will result in future submissions from that author being deprioritized, though 'botshit' will still be accepted.
- The maintainer dismisses concerns about AI-generated code degrading future AI training models, citing a Bing summary that advised against worrying about such studies.
A software project maintainer outlines a new policy for integrating AI-generated code, prioritizing ease of submission by skipping extensive code reviews for clearly marked AI contributions. While accepting these submissions readily, the maintainer reserves the right to make minor modifications, such as limiting execution to specific, unusual conditions like 'February 30th,' but assures authors they will retain credit.
