I personally minimize use of it because we have to test it for peroxidases and who wants paperwork? But honestly, good to know that it's worth some effort to find substitutes, and also that if I see anyone engineering a microbe to make it I might ask some questions.

Science is good, a bad epistemological framework except for all the others. So the question is not "how can AI fix scientific publishing?/how will AI ruin scientific publishing further" but "how can you understand the problems of science despite algorithmic framing of problems?"

Nah. Putin wanted us to know.