Anders Sandberg compared the standard to human coauthorship challenges
If you generate something with Claude or ChatGPT and you think it's good enough to publish, go ahead! Just credit the LLM as the author and yourself as the editor. Or at least credit the LLM as a ghostwriter. Just be transparent
No matter whether or how you use AI to produce writing, I think you should be able to pass this test. If you can't, I don't think you should publish under your own name
No Digg Deeper questions have been answered for this story yet.
@jasoncrawford I think about this a lot, having just finished a coauthored book. How much can I take credit for (or more importantly, explain) the finicky jurisprudence in the early chapters where I miss all the nuance? How do my coauthors deal with the dense math footnotes I added?
No matter whether or how you use AI to produce writing, I think you should be able to pass this test. If you can't, I don't think you should publish under your own name
@TrevMcKendrick You don't care who wrote an article? I do
In the future I expect I will want to know which AI wrote an article! Probably I will have favorite AI writers that I subscribe to. But I'll still care who or what wrote a piece
This just doesn't feel like the end state to me.... it feels like the old world asking the new world to keep a foot in.
So much will be created by AI that eventually we won't think about how anything was made, we'll just care whether it's good.
Similar to reading any article today... I don't care (or know) how many editors it had, how hard or easy it was to write, how long it took to produce, etc. I care only about the end result.
I care in that if it's someone whose writing I like I'm more likely to give the article a chance and keep reading if I'm not enjoying it. It's a brand thing. But whether I consider the article to be "good" has nothing to do with who the author is.
This would be true for "AI authors" too of course!
Why do you want to know how a piece was written or "who" wrote it?
@jasoncrawford We of course have a little declaration about who did mainly do what, and I think it is fairly obvious in many places. We have learned a lot from each other, but also lean a lot on each other. Proper coauthoring is a synthesis that individual authors could not do alone.