Decision Market Talks from Manifest

The Manifest recordings came out a week or two ago, so I thought I would share some of my favorite talks I saw. These two happen to be about decision markets, but you can find some other great ones on the Manifest YouTube channel.

Bo Cowgill - Beyond Forecast Accuracy: A Research Agenda for Prediction Markets

In this talk, Bo, a Columbia professor and former Google employee, discusses the highest-value verticals in prediction market research. The talk centers around two central ideas:

1. How does the presence of a market change the real world outcomes?

   1. It’s proven that having a prediction market about a specific event leads to accurate forecasts about it, but the key unproven piece is that having accurate forecasts leads to better outcomes in general.

   2. He proposes running prediction markets on whether company KPIs will be achieved. A randomly assigned half of the KPIs would have a market, and the other half would be a control group. If the experimental group achieved their goals faster, you could prove this was a causal chain.

2. How do we improve internal corporate decision markets?

   1. Classic market design considers the only incentive market participants have is to make money, but in the environment of an internal corporate decision market, some traders don’t want to reveal the information that they’re project might be behind schedule for fear of it being shut down or facing other potential consequences.

   2. If people have incentives to lie, the markets become thin and inaccurate, and potentially even have an Akerlof spiral, meaning they would start to be ignored by the employer.

   3. He proposes how to redesign this mechanism to align incentives — give it a watch!

Robin Hanson & Proph3t - Futarchy: A Status Report

Many avid Supercycle readers know that I am futarchy-pilled, meaning I think we should replace our representative democracy with prediction markets.

Robin seems to have a standard presentation he gives often to explain futarchy — the standout feature of which is a slide called “Imagine Autistic Person in C-Suite,” which explains how many high-functioning autistic people would be right about a certain prediction, but may not realize that the authority they are answering to may want to hear a sycophantic answer and not the accurate one. Futarchy, Hanson explains, solves this. By making honesty profitable, the futarchy market will almost always expose the truth rather than the answer that wants to be heard. The market is the autistic person in C-suite.

Returning from that completely unnecessary digression: he actually gave a different version of this presentation that’s focused on the implementation aspect rather than just the theory. A major critique of futarchy is “well, sure, it sounds good, but we’d need massive societal change to be able to implement this!”

They’re right. There’s not going to be a way to turn on futarchy with a switch. So Hanson lays out a plan for the best possible way that futarchy could be implemented.

He asserts that most innovation becomes mainstream by having a fringe group pilot it and, when the innovation is good, that group will suddenly achieve their goals much faster than expected.

Slowly, their innovation will spread because, naturally, people would like their group to prosper as well. Therefore, Hanson would like to have futarchy govern companies. Once the futarchy-powered companies outperform the rest, it’ll become the gold standard that VCs would require. Now, when all of the Fortune 500 companies govern with futarchy, maybe a city would, then a government, etc…

This is a natural implementation because of the preliminary experiments that have already been done amongst startups using futarchy for corporate governance.

And, of course, be sure to check out my session with Robin as well!