Will AI wipe out humanity before the year 2029?
75
500
1.4K
2025
2%
chance

If humanity is still around on Jan 1, 2029, then this market resolves to NO. Otherwise resolves to YES.

Get Ṁ200 play money
Sort by:

⚠UNRESOLVED & Re-Opened

📢Resolution Made Early Without Proof

📝I f@TeddyWeverka would like to N/A the market, that is their choice.

📝If Creator has proof, please post it.

Resolving early since the answer is obvious. Should this resolution be wrong, I will rise from my grave and payout those yes votes.

Why a weird 3% jump from 2028 to 2029 and 2029 to 2030, when almost all other years get just 1%?

Theoretically there could be a good reason for this, but realistically there is not. It is just a bias towards 2030.

bought Ṁ0 of NO

@DavidBolin the jump is 3x smaller than when @DavidHiggs made the same point below. I think this market should be lower, but unfortunately I don't want to sell because of loan mechanics.

Currently a 9% spread between this and the identical 2030 market. Of course they're both "optimally vote no regardless" resolutions, so can't really exploit it. 🙃

predicts YES

@DavidHiggs I have a little yes here for that reason but I think the extra year is significant to the odds.

@MartinRandall It's significant, but there's no possible way that 1/7 of the total years difference is worth over 1/3 of the odds change. Especially since LLMs or at least MassTransformers scaling "directly" into AGI is really the only plausible way to get superintelligence extinction in less than a decade, and that would probably imply a slightly shorter timeline than 2029/2030.

Well, other than massively important unpredictable insight chain into basement FOOM which is basically uniform or gradually growing small (shallow not necessarily small total area) probability distribution over the next several decades/maybe past the end of the century.