I'm interested in attempting to use Manifold's prediction markets and platform to start defining effective paths for kickstarting the adoption in liberal democracies of futarchy-like and other participative prediction-based systems for the purpose of improving collective decision-making, truthseeking and accountability.
I think I'd like to try achieving this through a movement (which I'm currently tentatively calling Predictive Democracy) that makes extensive use of prediction markets and other participative prediction-based systems for its own internal decisions, idea exploration and evaluation for at least two reasons:
1• the movement's success achieved through the tools it promotes would itself be rather strong evidence of the legitimacy and usefulness of said tools
2• the movement would have a certain amount of skin in the game, and an interest in diagnosing the limits of these tools it uses; patching them up or refining them where possible, and abandoning them where not
BOUNTIES OFFERED FOR ANY GOOD IDEAS/COMMENTS RELEVANT TO WHAT OUTLINED ABOVE.
I'm most interested in ideas regarding how to best go about this project and set up this initial phase to get the ball rolling, including — importantly — suggestions for further questions that could be asked on Manifold to start giving it a more defined shape.
The manifesto itself is still rather vague, needs work, and I certainly don't want to define it top-down on my own. I just tried to gesture at what I'm trying to do.
But I appreciate and could reward criticism too: e.g. are this project's objectives even desirable? Possible unintended consequences?
I can be fine with a fool's errand, but I wouldn't want to help create a monster.
And any other kind of interesting, relevant comment I didn't think about could be rewarded too, of course!
Will add more to the bounty pool if this turns out interesting.
1,000
3.00If you are going to build a Futarchy based upon prediction markets then anyone should have the capability to easily and cheaply run their own which is why I am building socialpredict … if you grant me mana I will use it to have a well known third party set up markets relevant to building this and I will subsidize those markets with that mana so they will be higher stakes and draw attention to the effort. https://github.com/openpredictionmarkets/socialpredict
Let me relay something that I think is an important point which I have learned from Manifold and from discussion in the discord.
A big challenge to futarchy, and in general to markets that try to predict the causal impact of some events on others, is that events in the real world are complicated and often polycausal. Let's say you define some goal for your movement to futarchically try to achieve. There are probably dozens of actions that you/your organization/your movement can take that affect whether you achieve your goal. If the choices are independent in their effects and genuinely uncertain, the average probability increase in achieving your goal that each of these actions affords will be at most ~1/(number of actions), and this doesn't even account for the influence of random factors outside your control.
It is therefore important to consider how to handle, when you ask the market which action is best, the answer coming back: "most actions are about the same in terms of their impact".
Yeah there's probably some value in prediction markets in a lot of areas. I think they would be good for things like helping to predict a virus, like where it could spread or how many cases there will be. It could at least help get a probability, as a tool in combination with other models. So maybe as an idea, you could try to use prediction markets to predict different aspects of a virus that you start tracking.
Like for the flu, maybe you could make markets for "If you get X flu shot, will you get the flu this season?" The percentage might be too low for prediction markets to be good at, so maybe it could be like "If this group of 20 people all get the flu shot, will at least one get the flu?" And same for if they don't get the shot.
If we make lots of markets on economic/welfare outcomes conditional on electoral outcomes, we can gather data about how calibrated Manifold is at predicting the consequences of political events. I just went ahead and made some about outcomes following the 2024 US presidential election.
Maybe we could also make markets about outcomes conditional on the passing of bills (closer to Hanson's original formulation), but I expect them to gather less interest.
In my understanding, Futarchy requires a crystal-clear goal. Clear like a stock price where the reaction to news become visible within seconds as a hard number.
Another important aspect is a clear commitment to give power to the markets. The easiest form of power would be money. This is not a manifesto but rather a rule set.
Going to give 50 Mana to the first up-to-15 accounts to contribute a comment written with a minimum of effort.
On Sunday I'll distribute all remaining Mana from the 2750 pool between the comments in some way (likes received will likely be taken into account).
Then I'll decide whether it's worth iterating again.