Nintil aims to be the world's best blog as per my own criteria of what the best blog is, and one of my requirements is correctness, I want the things I say here to be accurate. It's hard to keep track of everything I've written, and I may overlook things. Sometimes it's just easier to see mistakes in other's work rather than on one's own.
That's why, in the spirit of Donald Knuth I'm launching the "Prove me wrong, earn money!" rewards program. It covers everything I've published since 2017-01-01 and onwards.
Depending on what kind of mistake it is, I will pay you more money. I won't pay for typos or grammatical mistakes. For simplicity the values here are not inflation indexed, but I'll revise them upwards over time.
The way it works is you send me the article I wrote that has a mistake, along with why you think it's mistaken (Ideally link to supporting evidence, or write an explanation). As I will only pay once I see that I'm wrong, the more I have to read to understand your argument, the more delay there will be
|Kind of error||Reward|
|Minor argument, doesn't affect conclusion||4$|
|Major point, invalidates a section's main point||20$|
|Fundamental error, entire article is wrong||200$|
|I cited a paper that was retracted/failed to replicate|
|- If the paper* was published before I published my piece||60$|
|- If the paper* was published after I published my piece||4$|
* By paper I mean here the piece of evidence that contradicts what I had said before. The idea with the second item is that even if something seemed true at some point, and then after publication new evidence comes in, I should have to update it.
The rewards compound: If the cause of a section being wrong is a paper that failed to replicate I was unaware of, that's a reward of 80$! If you find 10 papers that failed to replicate in one of my old articles, that's 600$. Of course, this only counts if I cite the paper as supporting evidence. If I cite something and say "This won't replicate" it doesn't count if it doesn't replicate. However, if it does replicate, that might be a wrong minor argument, so $4!
There is no financial downside for you, if you don't manage to convince me, you don't have to pay me or anything.
Once I receive the potential critique, I may either reply explaining briefly why I think I'm still right, concede that I'm wrong, or promise to write a longer piece that goes into the specifics of the disagreement.
I will keep track of all the money I owe, but for logistics reasons, I will only make payments on amounts of 10$ or larger. Once the critiques start flowing in, I will set up a page with the names of those who have submitted them, and the owed amount (You can opt-out of this and choose to remain anonymous).
You can also choose not to receive the money, and instead have it donated by me(by default, to Givewell)
This rewards programme will remain active until I have paid 200$, then I will finetune the payments if I have made more mistakes than I thought in the past :)
An ongoing list of mistakes and money I paid can be found at mistakes
- 2019/12/22: I 2x the rewards from today on
- 2022/05/24: Linked to Mistakes You can get in touch at artirj at gmail dot com