1 month after GPT-5 comes out, I will create a poll that will run for a month with three options:
GPT-5 exceeded the general expectations for it.
GPT-5 exactly met the general expectations for it.
GPT-5 fell below the general expectations for it
This market will resolve "Exceeded" if more people vote that GPT-5 exceeded the general expectations around it then vote that it fell below general expectations. It will resolve "fell below" if the opposite happens. Yes, this effectively means the votes of the people who say it exactly met general expectations won't do anything. The price of neutrality I guess.
Respondents will be asked about general expectations around GPT-5, not whether it exceeded or fell below their own expectations. Exactly how they interpret general expectations I will leave up to them.
I have decided this poll is sufficient basis to resolve upon https://manifold.markets/GabeGarboden/did-gpt5-exceed-expectations
@Philip3773733 yes, this market should have been at exactly 50%, and well-calibrated manifolders should have been 50/50 split in the poll - but the actual poll was far from that!
@retr0id ah I should've paid more attention to this part... "Respondents will be asked about general expectations around GPT-5, not whether it exceeded or fell below their own expectations."