Will Kenshin9000 solve the ARC-AGI challenge?
Basic
26
Ṁ4832
Jan 1
4%
chance
If anyone that reads this is wondering when my next major post is, I am currently solving ARC-AGI, which I expect to solve with an 85-90% score in the next 14 days or so

Resolves YES if Kenshin9000 releases code or other artifact which anyone else is able to use to solve the ARC-AGI challenge with at least a 70% score by end of 2024.

Resolves NO if the score is less than 70% or if nothing is released.

Resolution based on the leaderboard and kenshin9000 being credited to a qualifying entry.

Get
Ṁ1,000
and
S1.00
Sort by:

It would be very amusing if he did.

bought Ṁ5 NO at 7%

FYI: Kenshin9000 and Mira are two scammers. Don’t trade in their markets

sold Ṁ56 NO

Made a variation on this market that I think removes some possible weasely ways in which it could resolve YES, and asks if Kenshin9000’s claim will end up being true more literally. Willing to bet with anyone at 15% (I’m buying NO)

If Kenshin writes 10 pages of incomprehensible tweets that somebody turns into qualifying code, that counts here but not in @Bayesian 's market.

(but it does have to be "used" - so "Thanks to Kenshin for inspiration" or if I can't figure out how it was used, won't count even if his name is mentioned)

To win prize money, you will be required to publish reproducible code/methods into public domain.

This requirement should mean it won't be undecided whether he did anything useful. We can look at the code, his posts, and hopefully relate the two.

Right. a "Thanks to Kenshin for inspiration" thing submitted alongside the ARC-AGI solution seems more plausible to me as a meme than a testament of an actual important and concrete contribution, so I want to avoid stuff like that. I get the benefit of including the possibility where Kenshin sends out some hard-to-parse but highly significant bit of brilliance. It just seems so implausible to me that the ways this market can end up resolving YES under my model is that criterion being somehow misapplied, which makes me not wanna put too much mana in. With my resolution I'm not similarly concerned. Not meant as an offence to your market lol, and didn't mean it as an agreement that you're a scammer obv

another way I would be willing to bet more (barring interest rates) would be if it was some kind of manifold mods panel that decided if the Kenshin9000 contribution was significant, at more than say 30% of the insights needed to solve ARC-AGI, or something subjective like that, but agreed upon by people that wouldn't wanna resolve for the meme.

I'm not going to do a mod panel and I preemptively reject their authority here.

ARC-AGI is a real challenge run by real AI people at Google.

Nobody's going to spend $10,000 and weeks of effort to cheat you out of $2 in Manifold points.

So if there's a genuine high score, and Kenshin gets credited, I want it to count and not reject it on technicalities like "somebody else wrote a shim to match the contest requirements, and submitted for grading, so it wasn't Kenshin".

If you see in the news "ARC AGI solved by Kenshin", it's generally going to count regardless of details.

If there's a real solve, your market I think would anger people.

Fair enough! that's a good point, hmmm.
> "somebody else wrote a shim to match the contest requirements, and submitted for grading, so it wasn't Kenshin".
Is that a common thing to happen for contests like this one?

bought Ṁ40 NO

I'll bet against Mira being a scammer. Make the market

bought Ṁ200 NO

"Being a scammer" is not the point.

Mira would like to resolve markets in favor of Kenshin, like the chess ones.

Turns out no matter how generous the terms, it never works.

reposted

The famed ARC AGI challenge will fall to the notorious Kenshin9000 in the next 2 weeks!

Will it?

As a trader, it's your responsibility to answer that question.