AI for science and math gets a trust problem

OpenAI published a repo of AI progress in mathematics, and the top comment said they "had to be publicly shamed" into engaging the mathematical community. Another called it unfortunate that gatekeeping got so bad they felt the need to ask permission to share math. Elsewhere, the Nobel Prize in Chemistry went to Henri Kagan and Kenso Soai for chirality work. That sparked a second thread on a report that mirror life may soon be possible, where the top question was what we would usefully get from doing it.

The pattern is that the claims are getting bigger, and the community's first question is about process and permission, not results. Who reviewed it? Who gets credit? What is the point? People are excited, but they want to see the work shared in the open before they believe the headline.