> What part do you think is going to become obsolete? Thank you for the question. I guess what I'm saying is: Will LLMs (or whatever comes after them) be _so

> What part do you think is going to become obsolete? Thank you for the question. I guess what I'm saying is: Will LLMs (or whatever comes after them) be _so_ good and _so_ pervasive that we will simply be able to say, "Hey ChatGPT-9000, I'd like to see if the xyz conjecture is correct." And then ChatGPT-9000 just does the work without us contributing beyond asking a question. Or will the technology be limited/bound in some way such that we will still be able to use ChatGPT-9000 as a tool of our own intellectual augmentation and/or we could still contribute to research even without it. Hopefully, my comment clarifies my original post. Also, writing this stuff has helped me think about it more. I don't have any grand insight, but the more I write, the more I lean toward the outcome that these machines will allow us to augment our research.