maybe the real proof was the friends we made along the way

If this ends up resolving NO it will be worldview-shaking enough that I won't really care about losing the mana lol

https://github.com/BuddenD/hodge-x33/blob/main/hodge-x33-draft.pdf

He posted a paper about Hodge, not the one he has second bet with Daniel Litt about, but related(?) and tagged Daniel Litt

damm i just realized this is all play money i put in 100 bucks thinking i could make 5 bucks easy and it's all gone 🤦 can you at least donate 100 bucks to charity on my behalf or something like that

@mods Hmm, good UI feedback.

@IsaacKing oh gosh. I just realized the existence of ubiquitous prediction markets now means the average consumer might just assume Manifold is a real money platform.

@bens gambling is illegal in my country and i'm not part of that bubble, never made an account on that kind of site, i just did a very uninformed decision lol.

@RATO Yeah sorry about that! Manifold is a bit of an outlier, more similar to Metaculus than a real-money market.

I'm noticing now that the mana-buying page does say "not redeemable for cash", so I don't think this is a UI flaw, just an unfortunate error.

Well RATO, I hope you like the community here, might find it fun to keep betting with your newfound funds. (And you never know, maybe Manifold will add back the ability to donate mana to charity. It used to exist but they removed it about a year ago.)

@IsaacKing I was aware of the existence of a "real" prediction thingy, and i thought this one was the real one, i looked up the one that actually uses real life money and i was immediately plastered with a warning popup saying it's not avaiable in my country, so this isn't ubiquitous in my context as it would be to an American.

@IsaacKing yeah it's unfortunate, i definetly didn't pay attention to the warning message

@IsaacKing IMHO the second sentence should be in bold.

Hasn't even published his full proof yet and people are building on his work already. That's how you know it's good.

@IsaacKing oh he’s… really not…

@IsaacKing Super awkward when I publicly call it out as an impressive hoax and then realise its a real attempt

Oops, I sold everything because I interpreted "Resolves N/A if the bet is called off. (One party not paying out and the other party magnanimously saying \"it's ok, I won't hold you to it\" does not count as the bet being called off.)" as meaning that in the magnanimous case it would resolve as NO, but based on the phrasing of the other market it sounds like Isaac intends to resolve that as YES. So I lost out on a little bit by selling everything, oh well.

@PeterSchmidtNielsen Wait, what? If the bet is called off this resolves N/A, not YES or NO.

@PeterSchmidtNielsen If Hutter wins and magnanimously does not hold the other party to it, it resolves YES, and if Budden wins and magnanimously does not hold the other party to it, it resolves NO was my interpretation

@IsaacKing I read:

"Resolves N/A if the bet is called off. (One party not paying out and the other party magnanimously saying \"it's ok, I won't hold you to it\" does not count as the bet being called off.)"

and I interpreted this as meaning that if "One party doesn't pay out and the other party magnanimously says \"it's ok, I won't hold you to it\"." happens then things do not N/A, and then magnanimous party has not "won" the bet, because they have not received the money. They have not lost it either, but they have failed to win it, and the market doesn't N/A, so in this case, the market would resolve as "NO", because the market was phrased in terms of one party winning (not the other party losing).

But it's clear now (and I think the text is clear enough, and I was just being silly) that in the above case you would consider the magnanimous party to have "won", even if they have not received any money, and even if the other party hasn't conceded. You intend to independently evaluate the terms of the bet yourself.

Excellent marketing campaign for his stealth startup lol

Assuming that this is an instance of AI psychosis, it raises an interesting ethical question: is it ethical to try to cash in on a bet you made with someone while they were having one such psychotic episode?

@NBAP If he had actually solved it during such an episode, do you think the other party should be released?

@IsaacKing Plausibly not. But I don't know if it's symmetrical. I think that people of sound mind can make binding commitments towards people who are not of sound mind, but I don't necessarily think the reverse is true.

@NBAP The thing is, he is not 'psychotic' in some clinical sense. 'AI psychosis' is a convenient shorthand for describing a certain new problem we see; but I think we would be making a mistake if, on a really largely lexical basis, we decided that people suffering from this failure mode were actually entitled to the same kind of protections as people who need to be committed to institutions.
There are after all a whole lot of ways and reasons to be wrong. And personally I feel like 'AI-catalyzed hubris' might be a more accurate description than 'psychosis'.

If this were going to cause him financial hardship I would see the argument, but this is a completely negligible amount to him, so I don't see any reason to release him from the wager. He's still of sound-enough mind to be keeping himself financially secure: I would have preferred to bet a larger amount, he's the one who set the 10k cap. And he's refusing bets from everyone else, and not betting on the real-money prediction markets. I think he's well aware on some level that him losing is a real possibility, he's not going to do anything too self-destructive.

I plan to donate my winnings to charity in any case.

@BorisBartlog I really like "AI-catalysed hubris", thanks

@IsaacKing due to infinity problem, the mathematical problem can't be solved. Big Bang in a river stream - easy

@BorisBartlog Psychosis is probably the wrong word but I think this analysis is ~100% wrongheaded. Hubris is a personality trait. What we're seeing is a huge number of usually smart people latching onto total fairy dust with a level of conviction that's completely uncharacteristic of them and which systematically leads them to make choices that, without their altered mental state, they would not endorse. These are almost exactly the circumstances under which mental health interventions and protections are justified.

I mention this not because I want to pin down Budden's diagnosis specifically, I think for any individual person less speculation is probably better. But this is going to keep happening, this will not be the last time, and we really need to get serious about this phenomenon and treat it as something worth understanding and solving or we're going to keep losing people.

Is anyone aware of any cases of people recovering from this state of mind? Budden is unusually smart among the victims and fully acknowledges that he didn't have a proof and is obviously aware of and seems to take seriously the idea that he could be cognitively compromised, but apparently remains convinced that he's days away from solving a problem I'm still not convinced he can state correctly. If he can't get out of it under those circumstances, I'm not sure anybody can under any circumstances. If we don't view this as a serious and legitimate mental health problem right now, I'm worried about how much damage it will do before we change our minds.