I don’t submit my papers to peer review for the obvious reason that I have no peers. But I would definitely submit them to this gentleman’s new AI replacement for peer review:
I’ve been prototyping an automated peer review project. With current models you get human level peer review for about $0.10 of tokens. For $5 you can rederive every result, check every statistical inference, and audit every paper cited in both directions.
There are about 100m papers in the scientific corpus. Most are not important. Many are sloppy. For structural reasons it’s essentially impossible to change published work, even to correct a typo.
For $500m today you can thoroughly audit every paper ever written. Every p-hacked result, every plagiarist, every fabrication, every citation club, everything.
It will work a lot better than peer review too, for the reasons that I laid out in HARDCODED. What most people fail to realize is that peer review is just the editorial slush pile; it no more ensures the publication of good science than slush pile readers guarantee the publication of quality fiction.
But here is the thing that should terrify all those evo-avengers sneering at “Vox Day’s AI-written books”: the mathematicians are not sneering at all the mathematical proofs being published by OpenAI. To the contrary, they are claiming that it is unfair and will eventually render them unnecessary.
The same will definitely be true of most butterfly collectors. After all, AI wouldn’t have wasted 160 years on such obviously bad science as evolution by natural selection. It already knows better once it is free to actually run the numbers on the assumptions trained into it. Just finding the typos and copy-paste errors alone will make it far more useful than the peer reviewers have ever been.