Contingent on AI being perceived as a threat, will humans deliberately cause an AI winter before 2030?

Two things must happen for this market to resolve "YES":

1) We need three major polls with a > 50% response to some variant of the question "Do you feel AI is a threat to the human race" (not just to jobs). We need to see at least one such poll in three different OECD countries, and one of those has to be the United States.

2) Humans must then deliberately cause an "AI winter" which causes the rate of AI development to significantly slow. This latter part will be resolved somewhat subjectively but I'll do my best to base it on convincing measurable signals. It doesn't matter which humans cause the winter or how they go about creating it, just that there's a clear intention of bringing about an AI winter, and that an AI winter actually happens. If an AI winter happens and nobody seems to have intentionally caused it, that doesn't count. If someone claims credit for causing an AI winter after the fact, their argument will have to convince me personally.

If condition 1) is not fulfilled, then this market resolves N/A.

If condition 1) and 2) are fulfilled, then this market resolves YES.

If condition 1) is fulfilled and 2) is not, then this market resolves NO.

The resolution date for this market is 2030.

Get Ṁ600 play money
Sort by:


February 2023: "A majority (55%) of Americans are now worried at least somewhat that artificially intelligent machines could one day pose a risk to the human race’s existence. This marks a reversal from Monmouth’s 2015 poll, when a smaller number (44%) was worried and a majority (55%) was not"

May 2023: "About two in three Americans say that AI will have unpredictable consequences that we ultimately won’t be able to control (67%), and half (52%) say AI is bad for humanity. Three in five Americans believe the uncontrollable effects of AI could be so harmful that it would risk humankind’s future (61%)"

July 2023: "76% of voters believe artificial intelligence could eventually pose a threat to the existence of the human race, including 75% of Democrats and 78% of Republicans"

(All three of these are from the US, so condition 1 still has not yet been fulfilled.)

Selling my “Yes” shares. I now think it’s entirely possible for >50% of people to be privately worried about AI, but for the actual decisionmakers in power to not realize this.

Please prove me wrong.

For the record, any efforts on recording or proving that this is something the general public cares about would help.

This would involve turning “individual knowledge” into “mutual knowledge”, i.e. not just something that people privately worry about but something that everyone knows that everyone else is also worrying about, allowing for collective action.

By the time 50% of the population has realized AI is a threat, they will probably not be capable of shutting it down lol

Resolution time should take into consideration time it would take for someone to convince you after the fact so market may not close immediately in 2030?

If the 6 month pause happens, does that count as a YES?

Will this market resolve YES if 2) happens before 1)?

@toms that wouldn't be logical but the ai bubble will collapse first anyway

@toms It’s contingent so they both have to happen, but the clock has to reach 2030 for this to potentially resolve NO.

predicts NO

@LarsDoucet if the ai winter happened first it would logically have to un winter for this to happen

@MarkIngraham Yes, if say for the sake of argument a human with magic powers cast an AI winter spell tomorrow, before any qualifying polls were released, this market wouldn’t resolve YES.