Pick any major news event from the last decade. Brexit. Trump in 2016. The 2024 US election polls. There’s always the same cast of characters, five experts on a panel, all confident, all contradicting each other, all completely unaccountable when they’re wrong.
The analyst who gave Trump a 12% chance in 2016 was back on television the next morning with a new theory. The economist who predicted six rate cuts last year, when there were two, is still writing columns. The pollster who called Remain in a landslide is still getting booked.
Scroll through the comments when those hot takes were published on social media, and people debate these points with real energy and an air of authority.
I mean, how many people became experts on COVID overnight? A lot.
We live in the hot take economy. Opinions are free, abundant, and consequence-free. The internet didn’t just give everyone a platform; it made being wrong an industry. There is no better business model in media than confident wrongness, because confident wrongness generates outrage, which generates clicks, which generates more bookings, and round we go.
Prediction markets showed up and ruined everything.
The premise is offensively simple. If you believe something is going to happen, back it with money. Not metaphorically, literally. Because the moment you’re asked to stake something real on your conviction, the noise clears remarkably fast. Accountability, it turns out, is a brutal filter.
What people are actually betting on tells you everything about the cultural moment. Sports, obviously, as 90% of Kalshi’s volume and are familiar territory for this industry. But scroll further and it gets interesting. The 2024 US presidential election attracted $3.7 billion in trading volume on Polymarket alone. There are currently 572 active markets on Trump with over $60 million being actively traded right now.
The US-Iran war attracted $529 million on a single contract, the largest geopolitical market in the platform’s history. People are betting on Federal Reserve rate decisions, cabinet departures, tariff levels on Chinese goods, and the exact words Elon Musk will say on a given day. We’ve even had a market on whether Jesus Christ would return before 2027. It’s trading at 4%. Which is higher than I’d have priced it, but what do I know?
The range is the whole point. From a Sunday NFL game to a regime change in Tehran. People want to bet on everything they have a strong opinion about. They always did. Nobody gave them the venue.
Here’s what makes prediction markets work where punditry fails. On a TV panel, the incentive is to sound right, confident, quotable, contrarian enough to be memorable. On a prediction market, the incentive is to be right, because being wrong costs you money. Those two incentive structures produce completely different outputs. Pundits optimise for performance. Traders optimise for accuracy.
And crucially, when you put your position on the line publicly… screenshot it, share it, build a following around your track record, you’re not just betting. You’re staking your credibility on your conviction. Social media didn’t create prediction markets, but it gave them rocket fuel, because humans desperately want to be publicly right about things. The position is the content. The win is the proof.
The numbers bear this out. When polling aggregates had the 2024 US presidential race as essentially a dead heat, Polymarket had Trump trading at 61-66% probability for weeks. Trump won decisively. Prediction markets didn’t get lucky, they were aggregating what thousands of traders with real money at risk actually believed, rather than what respondents told a pollster over the phone. The bullshit filter, when applied correctly, is brutal and it works.
Except. this is where it gets uncomfortably interesting.
This week, newly created Polymarket accounts placed bets on a US-Iran ceasefire in the hours before Trump announced one. They walked away with $200,000, $125,000 and $48,000 in profit. One account was created twelve minutes before Trump’s post. Similar patterns emerged before the US strikes on Iran in February, before the capture of Venezuela’s Nicolás Maduro in January. Clusters of brand new accounts, suspiciously well-timed, profiting on events that hadn’t happened yet.
The accountability thesis only functions as a truth machine when everyone starts from the same information. When insiders are trading on Polymarket before announcing military operations, you don’t have wisdom of crowds. You have a mechanism for people in power to profit from their own decisions, which is arguably worse than a pundit being wrong on television.
Of course, this problem isn’t unique to prediction markets and it doesn’t invalidate the mechanism, it reveals exactly what needs fixing. Both Kalshi and Polymarket have already announced enforcement measures. Congress is moving on insider trading legislation specific to prediction markets. The accountability engine works; it just needs the same guardrails that financial markets took decades to develop. The pundits on television have no such guardrails, and nobody is building them.
For anyone in this industry, the trajectory is impossible to ignore. DraftKings, FanDuel and Fanatics all launched prediction market products within the same three-week window in December. Robinhood exposed them to 27 million funded brokerage accounts.
These companies don’t move simultaneously by accident. They’re following consumer attention, and consumer attention is moving toward a product that sits at the intersection of news, identity and wagering in a way traditional sportsbooks never have. A sportsbook is somewhere you go. Prediction markets are slowly becoming part of how people process the world.
The hot take economy needed accountability. It found it. The mechanism isn’t perfect (nothing is on the first iteration) but the direction of travel is clear, the numbers are real, and the cultural fit is undeniable.
