>>13462997I don't think I am engaging in cope, I think this is a realistic take. Mind you, language models writing webbed boilerplate is a good thing. The techniques that allow for this can probably lead to tools that will augment thoughtful programmers working on many different things. Likewise, whatever happens with automation in mathematics in our lifetime will probably take the form of tools for researchers, rather than replacing researchers.
If you want to talk about possibilities beyond our lifetime, that is fine. I'm less likely to cast doubt the farther out you go in time, but I also have no idea what form that future will take, and I see no reason to believe that mathematics research, whatever it winds up looking like, will exclusively be carried out by AIs. To believe that, I would need to believe that humans will not change or adapt to these new technologies and would simply stop participating. This doesn't make sense to me, since we are still playing chess after getting bodied by machines, and mathematics research on the whole is way more nebulous a task (which means it has room for humanity to hide). This is to say nothing of the possibility that humans begin to augment their own intelligence with biological engineering, or the possibility that rational intellect is unique gift bestowed on us by Allah rather than something that can manifest in a golem.
