My very high level goal: I want people to be doing “fortified chat”, where when they are having an argument, they think “there should be a market for this”, and they find one if there is, or create one if there isn’t, and then the market is part of their argument.
Example user flows I want to be good:
A) Two people are arguing about whether it was good or bad that Bing produced their insane Bing chat bot. James says “I bet they pull the plug on it this week.” Steve thinks, “James is full of shit. I wonder if there’s a market on this.” Steve can figure out quickly if there is a market and then if there is, he can link it. If not, he can create it. Then he can peer pressure James into betting on his beliefs.
Example utopian transcript:
[James] the bing chat bot sucks ass. they will delete it instantly
[Steve] /manifold search bing chatbot turn off
[Manifold bot] (visible only to Steve) Search results:
[1] Will the Bing chatbot be turned off in 2023? [link]
[2] Will Microsoft kill the Bing chat bot before Google creates a search assistant? [link]
[3] Will I think the Bing chat bot is a failure by June? [link]
[Steve] /manifold 1
[Manifold bot] Will the Bing chatbot be turned off in 2023? [link] P = 64%.
[Steve] the market demonstrates that you are full of shit
[James] i believe what i said
[James] /manifold bet 1000 yes
[Manifold bot] James bet yes on "Will the Bing chatbot be turned off in 2023?" [link]
Previous P = 64%. New P = 96%.
B) Some people are arguing about something in chat and an onlooker wants to make a market as a kind of funny sideshow to their discussion. The thing they are arguing about is really parochial and specific (e.g. it’s about a fight that people are having right there) so there’s no chance there is already a market on it. James wants to make a market right on the spot and bet on it as a way of expressing how he feels about the argument, and other people might follow on.
Example utopian transcript:
[Ian] fuck you, you know that i didn't literally mean there are lizards on mars
[Austin] i think you literally meant that because you are too stupid to know that
lizards need water so there can't be lizards there
[Ian] ok but there is water on mars so maybe there could be lizards
[James] /manifold create Will Ian admit lizards cannot live on Mars in the next hour
[Manifold bot] Will Ian admit lizards cannot live on Mars in the next hour [link] P = 50%
[James] ian will never admit it
[James] /manifold bet 100 no
[Manifold bot] James bet no on "Will Ian admit lizards cannot live on Mars in the next hour" [link]
Previous P = 50%. New P = 17%.
[Alice] /manifold bet 100 no
[Manifold bot] Alice bet no on "Will Ian admit lizards cannot live on Mars in the next hour" [link]
Previous P = 17%. New P = 8%.
[Ian] /manifold create Will I quit my job in 2023
[Manifold bot] Will I quit my job in 2023 [link] P = 50%
[Ian] /manifold bet 100 yes
[Manifold bot] Ian bet yes on "Will I quit my job in 2023" [link]
Previous P = 50%. New P = 83%.
[Austin] /manifold bet lizards 100 no
[Manifold bot] Austin bet no on "Will Ian admit lizards cannot live on Mars in the next hour" [link]
Previous P = 8%. New P = 4%.
C) Someone is streaming a game and people are using a market to have fun betting on the outcome of the game. By convention it’s someone’s job to make the market, and everyone else bets a lot on it, and it’s slightly cool to accumulate the play money and be the best at the betting minigame. There is a lot of churn where people are creating markets and then rapidly closing and then resolving them after a series of bets.
Total set of features that seem important to do a really good job with:
Features I’m not sure if they are important: