PredictIt vs Kalshi vs CFTC saga continues. Future Fund anounced $1M+ prize for shifting their probabilities of AI timelines & dangers. Dan Luu looks at the track record of futurists.
>I particularly liked this article (a) on the Chicago Tribune on how prediction markets are an antidote to degraded public discourse.
It's a good article.
I think a more powerful pitch would be around (a) information overload and (b) expert identification.
Pre-internet, info available on any given topic was limited. Post-internet, there is limitless information, and you can find an "expert" arguing in support of pretty much any position you want.
The problem is on any given issue, there is too much info for the average person to want to survey it all -- reading info that conflicts with your priors, in particular, is basically work, and most of us are on the internet for fun, not work.
Prediction markets seem like a powerful way to incentivize a few people to get really in-depth on particular issues, in a way that allows anyone to quickly see the state of the debate and learn which "experts" are actually willing to bet their beliefs.
If the UK is *de facto* governed by the markets anyways, why don't they just implement conditional prediction markets and figure out what policy the market values the highest? I think this is the only way for Liz Truss to save her political career. The policy the market likes the most is probably much better than the policy Truss could come up with on her own (at least, assuming a sufficiently wide array of policies have associated conditional markets, so the policy search space is nice and large).
Or we could tell the next PM: "Don't be like Liz Truss. Test your policies out using conditional markets before you announce that you're officially going to implement them. Doing so is essential for your political survival."
>I particularly liked this article (a) on the Chicago Tribune on how prediction markets are an antidote to degraded public discourse.
It's a good article.
I think a more powerful pitch would be around (a) information overload and (b) expert identification.
Pre-internet, info available on any given topic was limited. Post-internet, there is limitless information, and you can find an "expert" arguing in support of pretty much any position you want.
Early internet days of massive argument databases https://astralcodexten.substack.com/p/the-rise-and-fall-of-online-culture have given way to the sort of nihilistic arguments you see today, where people don't even manage basics like citing their sources.
The problem is on any given issue, there is too much info for the average person to want to survey it all -- reading info that conflicts with your priors, in particular, is basically work, and most of us are on the internet for fun, not work.
Prediction markets seem like a powerful way to incentivize a few people to get really in-depth on particular issues, in a way that allows anyone to quickly see the state of the debate and learn which "experts" are actually willing to bet their beliefs.
- - -
BTW, did you hear about the UK political chaos because the PM made some moves the markets didn't like? https://www.cnbc.com/2022/10/14/uk-pm-liz-truss-fires-finance-minister-kwasi-kwarteng.html
If the UK is *de facto* governed by the markets anyways, why don't they just implement conditional prediction markets and figure out what policy the market values the highest? I think this is the only way for Liz Truss to save her political career. The policy the market likes the most is probably much better than the policy Truss could come up with on her own (at least, assuming a sufficiently wide array of policies have associated conditional markets, so the policy search space is nice and large).
Or we could tell the next PM: "Don't be like Liz Truss. Test your policies out using conditional markets before you announce that you're officially going to implement them. Doing so is essential for your political survival."
Never let a crisis go to waste I say!