The Teleprompter Bet and What Prediction Markets Actually Became
A White House teleprompter operator was betting on Kalshi about what the president would say — words he was scrolling onto the screen himself. He made about $100,000, and according to economist James Woodruff at Platypus Economics, the shock isn’t that he profited. It’s that he didn’t make more.
That’s the current state of prediction markets. Not the elegant information aggregation engines the academics dreamed up. A guy with a headset and a keyboard exploiting the most extreme asymmetric information advantage in modern finance.
The confession that hurts the most
Woodruff’s piece “I Championed Prediction Markets. Look What They’ve Become” is worth reading because he’s one of the people who helped build the intellectual case for freeing these markets from regulation. The dream was noble — market odds that beat the polls, beat Nate Silver, surface truthful probabilities about elections and product safety and disease outbreaks. And partly it came true. Prediction markets are better than polling at forecasting elections.
But go look at Kalshi or Polymarket today. Roughly 90% of the money is on sports. We built a superior information aggregation machine, and the ideas are being used to create a backdoor sportsbook.
Woodruff knows the cost personally. Before he was an economist, he was a teenager working for bookies on Australian racetracks. He once took five dollars from a coworker who “didn’t bet” and put it on number one in the next race. Six months later that coworker was betting thousands a race and stealing from his own family. Compulsive gambling destroys lives, and the families it wrecks don’t have lobbyists. The firms dominating the industry have piles of cash and Donald Trump Jr. on both of their payrolls.
The strongest objection, and why it still bothers me
The counterargument is real: prediction markets do surface useful information. The market odds for elections consistently outperform polls, pundits, and forecasting models. Maybe the sports-betting volume is just the subsidy that keeps the infrastructure alive for the high-signal stuff. A little gambling noise in exchange for genuinely better predictions about the world — that’s a trade worth making.
Maybe. But here’s what gives me pause: what happens when the asymmetric information problem scales?
The teleprompter guy is one person with one cheat code. Trump Media is now selling early access to Truth Social posts as a paid product. Think about that — the president announces policy, sometimes war, through Truth Social. Embassies maintain accounts just to stay informed. And now anyone active in markets will have no choice but to pay for early access, because the alternative is trading against someone who does have it. The Federal Register was the public, neutral channel. Now the president has privatized and monetized it.
Here’s what I think Woodruff gets right
He doesn’t pretend to have the answer. He says America never had a grown-up debate about prediction markets — just a regulatory vacuum that the gambling industry filled faster than the information-economists could build their case. That’s the pattern: build a beautiful tool, assume good faith, watch the incentive structure win.
I don’t think prediction markets are bad. I think unregulated prediction markets, in a world where the teleprompter guy exists and Truth Social posts are a paid feed, are a recipe for the same thing gambling always produces: a few winners with inside information and a lot of people who didn’t realize they were playing a rigged game.
Woodruff ended his piece: “I don’t claim to know the exact right policy answer for America.” Same here. But I know the right question: who is on the other side of your trade?
If you don’t know — and you can’t know — it’s not a prediction market. It’s just a casino with better marketing.
Sources: Platypus Economics — “I Championed Prediction Markets. Look What They’ve Become” (July 23, 2026); Hacker News discussion