Pushing as much of the costs of generated AI onto the contributor is 100% the right way to do this. There has been a substantial shift towards generated code. Which might be temporary depending on your belief system but we are in the world we are in right now.
Contributors using LLMs should have a burden of proof higher than that of traditional contributors. This includes tests, recordings, proofs, discussion, etc.
Pushing as much of the costs of generated AI onto the contributor is 100% the right way to do this. There has been a substantial shift towards generated code. Which might be temporary depending on your belief system but we are in the world we are in right now.
Contributors using LLMs should have a burden of proof higher than that of traditional contributors. This includes tests, recordings, proofs, discussion, etc.