I agree. What would stop the AI from explaining the proof in a way human mathematicians can understand, and then giving them a roadmap for spreading it through

I agree. What would stop the AI from explaining the proof in a way human mathematicians can understand, and then giving them a roadmap for spreading it through the community via textbooks, conferences, and so on?