It's one thing for a maintainer to use AI, and another to review AI code sent your way. The latter has been proven to be unsustainable, low value, and ultimately a challenge for open source. Everyone acknowledges this problem exists and some corrective action is required to mitigate it.