 If you were somehow able to use detailed information about the future to make a ton of money, even if you donated it all to charity, I think that you still technically have to list yourself as being for profit. Stephen Hawking made several famous public wagers on various topics in astrophysics and cosmology during his career. He bet Kip Thorne that a source of x-rays in the constellation Cygnus wasn't a black hole. Which it was. He bet physicist Gordon Cain a hundred dollars that nobody would find evidence of the Higgs boson, which he ended up paying in 2012 when Cern announced their findings. The stakes were never really life changing, but Hawking was infamous for his willingness to put some stakes on scientific speculation. That's kind of a rare thing, which is odd considering how much time we spend making guesses about the future. Think about it. Every conversation about politics, every debate over where we should go for dinner, every argument about when to use the Oxford comma is essentially a dispute about what will happen in the future. Electing that person will have negative consequences. A meal at this restaurant will hit the spot. If you don't use the Oxford comma, people will be more confused. With so much prediction happening in everyday discourse, you might expect that we'd have a ton of respect for those folks with especially accurate mental models who are regularly right about what's going to happen. But Hawking's definite public declarations of his guesses are the exception, not the rule. Usually we're content to simply state or overstate our opinions, often using language that vastly oversells our confidence in them for rhetorical purposes. How many times have you heard someone claim to be 99% sure of something? Usually it's just a bit of hyperbole. They're pretty certain, but not positive. If that estimate of their confidence was literal, there's a wager that they should be willing to accept. If the prediction turns out to be true, they win a dollar. If it's not true, they have to pay 99 dollars. How many people do you think would be willing to take that bet? Prediction markets are essentially exchanges that operate on exactly that principle, basically functioning as stock markets or, if you like, casinos for forecasters. Rather than speculating on stock in a company or placing bets on hands of cards, prediction markets allow people to put money down on assertions about the future, or contracts. Contracts are a little different than stocks in that their value is determined as a probability distribution rather than a cash value, and when you invest, you buy in at a certain percentage. You might buy something like computer program passes Turing test by 2025 at 65%, meaning you're about 65% certain that that will happen. Or you could clean up suckers money on vaccines cause autism by putting as much cash as you can spare all the way down at 1%. The value of the contract changes in a very similar fashion to the value of a stock price. If a lot of people bet on something happening, its percentage value goes up. If they bet on it not happening, the value goes down. This causes a similar process of market value equilibrium, where speculators who think a contract's current percentage value is much higher or lower than it ought to be will invest according to that belief, shifting the curve in that direction. When the contract closes, everyone's final bids are locked in, and when the event either transpires or doesn't, it pays out according to their level of investment. If you were one of the only people to bet on an outcome that everyone else thought was impossible, you can make bank if it actually happens, or not. The upshot of this whole process is that the distribution of the contract price is a real-time index of the investor's certainty that is correct, not their cavalier conversational speculation, but what they really think in their heart of hearts. After all, it's a numbers game. If they guess right, there's money to be made, and money to be lost if they screw up. That means that the probability distribution of a contract represents an incredibly accurate crowd-sourced guess about its likelihood, often a much more consistently correct guess than even experts with access to tons of data can manage on their own. That's useful information. So useful that many large companies like Google, Microsoft, and Qualcomm use internal prediction markets to inform business decisions. Of course, no market is perfect, and there are problems with blindly trusting the wisdom of the crowd to accurately forecast everything. Prediction markets attract a certain kind of person, and that selection bias can manifest as skewed results. There's also the problem of specificity when defining a contract. It's fairly easy to buy and sell shares of a company at a particular price, but determining whether a particular event has actually come to pass can be a fussy process. And just as with stock markets, there are numerous devious ways to exploit prediction markets to make a quick buck, or worse yet, to influence the real world. After all, if you've got 20 large writing on some political leader being assassinated in the next year, but research into these prediction markets has led to an interesting development in the psychology of accurate forecasting. In 2011, a group of researchers funded by a US intelligence initiative started something called the Good Judgment Project, a study which asked several thousand people to take their best guesses on a set of geopolitical and economic events, and use those results to offer amazingly accurate projections of the future. Weirdly, while there were crunching numbers, the researchers discovered that a subset of their study participants were significantly better at guessing than the vast majority of people. These so-called super forecasters weren't mega-geniuses or prophets, but they did consistently better on their predictions than most, even better than some government analysts with access to top-secret information. Of course, the Good Judgment Project will happily sell you the secret sauce to be a better forecaster in their workshops or online training courses, but the general gist is outlined in the book Super Forecasting, written by two of the project's leads. It seems that most of the mental apparatus that super forecasters have in common are forms of debiasing, psychological tools and balances that allow them to compensate for the usual suspects that make human estimates more a matter of wishful thinking than rigorous analysis. Oh look, a whole video on debiasing! Crazy, right? What are the chances? Along with the study's findings came the idea of using the mojo that powers prediction markets and refining it with individual track records, using statistical techniques to weigh certain guesses more or less depending on the past performance of the guesser in question. The recreational online prediction market, Metaculous, uses such an algorithm to calibrate a prediction engine that operates parallel to the public market and has astounding accuracy, nailing its guesses more than 90% at the time, better than even super forecasters can manage. But hey, maybe you think you can do better. Metaculous is one of several online prediction market games that you can play with virtual currency, seeing how you stack up against other forecasters without actually gambling. There are also distributed cryptocurrency markets for those who think they're good enough to make real money. If you're an opinionated person who regularly makes pronouncements about what's certain to happen next, maybe try taking a page from Hawking's book and putting some stakes on that opinion. Hopefully you do better than he did. Out of four publicly stated scientific bets, he only ever won once. Do you think that prediction markets are useful for forecasting future events? Please leave a comment below and let me know what you think. Thank you very much for watching. Don't forget to blah, blah, subscribe, blah, share, and don't stop thunking.