Tool-assisted proofs typically already get a description of the tool usage. Even if one wanted to start giving authorship credit to llms, they’re too non-atomic (did it have web search, which mcp?, etc.)
Tool-assisted proofs typically already get a description of the tool usage. Even if one wanted to start giving authorship credit to llms, they’re too non-atomic (did it have web search, which mcp?, etc.)
Not trying to delving into a deep philosophical question here but...
I asked chatgpt to write a poem about my mothers dog a while back. It spit out a poem that my mother likes and keeps around. When asked, I say chatgpt wrote it. If I asked chatgpt for a proof of the Goldbach Conjecture and it spit out a verifiable proof, I think I would go ahead and give chatgpt credit. Not that I think it is likely. It would be more of some ability (like a robot end effector is able to hold an egg) and monkeys at typewriters.
Maybe not Fields Medal worthy, but worthy of some credit.
Anything morality-related is going to be subjective, but I think there are just too many practical problems with LLMs as coauthors. For me, in a scientific context 'chatgpt' is too vague, and I think it would be logically inconsistent to have LLMs as coauthors and not other forms of Monte Carlo. I also think LLMs are just too mechanical to be ascribed 'people words' (in the same way I don't consider my automated coffee machine a barista).