The duality of Polymarket: a betting site for participants, a news site for everyone else
Over the past week, Polymarket has been a very effective source of information on the US election. Polymarket not only predicted a 60/40 chance of Trump winning (while other news sources predicted a 50/50 chance, which is not too impressive in itself), but also displayed other advantages: when the results came out, although many experts and news sources had been trying to entice the audience to hear news favorable to Harris, Polymarket directly revealed the truth: the chance of Trump winning was over 95%, and the chance of him taking control of all government departments was over 90%. But for me, this is not even the best example of what's interesting about Polymarket. So let's look at another example: the election in Venezuela in July. The day after the election ended, I remember seeing someone out of the corner of my eye protesting the highly manipulated election results in Venezuela. At first, I didn't pay much attention. I knew Maduro was already one of those "basically dictators", so I figured he would of course fabricate every election result to keep his power, of course there would be protests, and of course the protests would fail - unfortunately, many others have failed. But then, as I was scrolling through Polymarket, I saw this: People were willing to bet over $100,000 on a 23% chance of Maduro being overthrown in this election. Now I started paying attention. Of course, we know the unfortunate outcome of this situation. Ultimately, Maduro did remain in power. However, the market made me realize that this time, the attempt to overthrow Maduro was serious. The scale of the protests was massive, the opposition put out an unexpectedly well-executed strategy, and they proved to the world how fraudulent the election was. If I hadn't gotten the initial signal from Polymarket that "this time, something is worth paying attention to", I might not have started paying attention at all. You should never fully trust the Polymarket betting charts: if everyone believed the betting charts, then any wealthy person could manipulate the betting charts and no one would dare bet against them. On the other hand, fully trusting the news is also a bad idea. The news has sensationalistic motives, exaggerating the consequences of anything for the sake of clicks. Sometimes this is reasonable, sometimes not. If you see a sensationalistic article, but then go to the market and find that the probability of the relevant event has not changed at all, then skepticism is warranted. Or, if you see an unexpectedly high or low probability in the market, or an unexpectedly sudden change, that's a signal to read the news and see what's causing it. The conclusion is: by reading both the news and the betting charts, you can get more information than by reading either one alone.The broader meaning of "info finance"
Now, let's get to the important part: predicting election results is just the first application. The broader concept is that you can use finance as a way to coordinate incentives to provide valuable information to the audience. A natural reaction is: isn't all finance fundamentally about information anyway? Different participants make different buying and selling decisions because they have different views on what will happen in the future (in addition to personal needs like risk preferences and hedging desires), and you can infer a lot of knowledge about the world by reading market prices. For me, info finance is like this, but structurally correct. Similar to the concept of structurally correct in software engineering, info finance is a discipline that requires you to (i) start with the facts you want to know, and then (ii) deliberately design a market to best extract that information from market participants. Prediction markets are one example: you want to know a specific future fact, so you set up a market for people to bet on that fact. Another example is decision markets: you want to know which of two decisions A and B, according to some metric M, will produce a better result. To achieve this, you set up conditional markets: you require people to bet (i) which decision will be chosen, (ii) if decision A is chosen, the value of M, otherwise zero, (iii) if decision B is chosen, the value of M, otherwise zero. With these three variables, you can determine whether the market thinks decision A or decision B is more favorable for obtaining M. I expect that a technology that will drive the development of info finance in the next decade is AI (whether large models or future technologies). This is because many of the most interesting applications of info finance are related to "micro" problems: hundreds of millions of small markets, where individual decisions have relatively small impact. In fact, low-volume markets often cannot function effectively: for experienced participants, it doesn't make sense to spend time on detailed analysis just to make a few hundred dollars in profit, and many believe that without subsidies, such markets simply cannot function at all, because apart from the largest and most sensational issues, there are not enough naive traders for experienced traders to profit from. AI completely changes this equation, meaning that we may be able to obtain reasonably high-quality information even in markets with $10 in trading volume. Even if subsidies are required, the subsidy amount per issue becomes very affordable.Information finance needs human distillation
Judgment
Suppose you have a trustworthy artificial judgment mechanism, and this mechanism has the legitimacy of the entire community's trust in it, but making judgments requires a long time and high cost. However, you want to access at least one approximate copy of this "expensive mechanism" in real-time and at low cost. Here are the ideas proposed by Robin Hanson on what you can do:
Each time you need to make a decision, you will set up a prediction market to predict what result the expensive mechanism will make on the decision. You let the prediction market run and invest a small amount of funds to subsidize the market makers.
99.99% of the time, you actually don't call the expensive mechanism: maybe you will "cancel the trade" and return everyone's input, or you just give everyone zero, or you look at whether the average price is closer to 0 or 1 and treat it as a basic fact. 0.01% of the time - it may be random, it may be for the markets with the largest trading volume, or it may be a combination of the two - you actually run the expensive mechanism and compensate the participants accordingly.
This provides you with a credible, neutral, fast and inexpensive "distilled version" that reflects the behavior of your original highly trusted but extremely costly mechanism (using the term "distilled" by analogy to "distilled" in LLM).
A possible model of prediction markets + community note combination.
This applies not only to social media, but also to DAOs. One of the main problems of DAOs is that there are too many decisions, and most people are unwilling to participate in them, which leads to either widespread use of delegation, with the risks of centralization and delegation failure common in representative democracy, or easy to be attacked. If actual voting in the DAO rarely occurs, and most things are determined by prediction markets, with humans and AI combined to predict the voting results, then such a DAO may run well.
As we have seen in the example of decision markets, information finance contains many potential paths to solve important problems in decentralized governance, and the key is the balance between markets and non-markets: markets are the "engine", and some other non-financialized trust mechanisms are the "steering wheel".
Other use cases of information finance
Personal tokens - such as Bitclout (now deso), friend.tech and many projects that create tokens for each person and make them easy to speculate on - are a class I call "primitive information finance". They deliberately create market prices for specific variables (i.e. expectations of a person's future reputation), but the exact information revealed by the prices is too vague and subject to reflexivity and bubble dynamics. It is possible to create improved versions of such protocols and solve important problems like talent discovery by more carefully considering the economic design of the tokens (especially where their ultimate value comes from).
Advertising - the ultimate "expensive but trustworthy signal" is whether you actually buy the product. Information finance based on this signal can be used to help people determine what to buy.
Scientific peer review - the scientific community has long had a "replication crisis", where some famous results have become part of folk wisdom in certain cases, but ultimately cannot be replicated in new research. We can try to use prediction markets to determine which results need to be re-checked. Before re-checking, such markets will also allow readers to quickly estimate the extent to which they should trust any particular result. Experiments with this idea have already been done and seem to have been successful so far.
Public goods funding - one of the main problems with the public goods funding mechanism used by Ethereum is its "popularity contest" nature. Each contributor needs to conduct their own marketing campaign on social media to gain recognition, while contributors who lack the ability to do so or have more "background" roles naturally find it difficult to obtain large amounts of funding. An attractive solution is to try to track the entire dependency graph: for each positive outcome, which projects contributed how much to it, and for each project, which projects contributed how much to it, and so on. The main challenge in this design is to find the weights of the edges so that they can resist manipulation. After all, such manipulation has been happening. The distilled human judgment mechanism may be helpful.
Conclusion
These ideas have been theorized for a long time: the earliest writings on prediction markets, even decision markets, date back decades, and similar arguments in financial theory are even older. However, I believe the current decade provides a unique opportunity, mainly for the following reasons:
Information finance solves the trust problems that people actually have. A common concern of this era is the lack of knowledge (and worse, the lack of consensus) about who to trust in political, scientific, and business environments. Information finance applications can help become part of the solution.
We now have scalable blockchains as a foundation. Until recently, the costs were too high to truly realize these ideas. Now, they are no longer too high.
AI as a participant. When information finance had to rely on human participation in every issue, it was relatively difficult to function. AI greatly improves this situation, allowing effective markets even on small-scale issues. Many markets may have a combination of AI and human participants, especially when the number of specific issues suddenly changes from small to large.
To fully seize this opportunity, we should go beyond just predicting elections and explore what else information finance can bring us.
Special thanks to Robin Hanson and Alex Tabarrok for their feedback and comments.
