The academics and researchers who spent decades arguing that prediction markets would make us smarter are watching their creation go mainstream — and some of them aren’t sure they like what they see.
The Uncomfortable Success Story
Robin Hanson has been evangelizing prediction markets since before most current traders could spell “probability.” The George Mason University economist helped pioneer the theoretical framework that underlies every contract traded on Kalshi and Polymarket’s latest markets today. He should be celebrating. Instead, he’s expressing something closer to philosophical unease.
The problem isn’t that prediction markets failed. The problem is they succeeded — spectacularly, rapidly, and in ways that have little to do with the epistemic revolution Hanson and his intellectual fellow travelers imagined.
When Hanson and other researchers pushed for prediction markets in the early 2000s, the pitch was elegant: let people bet on outcomes, and you’ll aggregate dispersed information more efficiently than any poll, any pundit, any committee. The market price becomes the best available estimate of truth. Information flows to where it’s needed. Society gets smarter.
What actually happened is that prediction markets became another vehicle for speculation, another place where people chase dopamine and action, another corner of finance where the line between investing and gambling is whatever the lawyers say it is.
The Volume Numbers Tell a Different Story Than the Headlines
The explosion in prediction market trading over the past eighteen months has been genuinely stunning. Polymarket has crossed thresholds that seemed impossible two years ago. Kalshi fought the CFTC to a standstill and won the right to offer election contracts that generated massive attention during the 2024 cycle. New platforms keep launching with venture backing and aggressive marketing.
But look closer at the composition of trading activity. How much is driven by people genuinely seeking to hedge risk or express informed views? And how much is simply sports betting’s cousin — recreational gambling dressed in the language of probability and forecasting?
The original prediction market boosters didn’t imagine a world where crypto traders would pile into Taylor Swift wedding contracts or where the biggest volumes would cluster around whatever political controversy generated the most cable news chatter. They imagined corporate prediction markets helping companies allocate resources. They imagined policy prediction markets helping governments anticipate problems. They imagined something closer to a public utility than a entertainment product.
The utilities never really materialized. What materialized was action.
The Regulatory Dance Nobody Predicted
There’s a deep irony in watching regulation debates unfold around prediction markets in 2024 and 2025. The academic advocates spent years arguing that prediction markets weren’t gambling — they were information aggregation tools deserving protection and encouragement. The distinction was always somewhat theoretical, but it mattered enormously for legal purposes.
Now the industry is caught between competing narratives. To regulators and skeptics, these platforms look an awful lot like sportsbooks with better PR. To true believers, the regulatory backlash confirms what they always suspected: incumbents will crush anything that threatens their information monopolies.
The reality, as usual, sits somewhere messier. Wall Street’s biggest names are circling prediction markets not because they believe in epistemic democracy but because they see a new asset class emerging. The CFTC’s approach to event contracts reflects genuine uncertainty about how to categorize instruments that don’t fit neatly into existing buckets.
And the original researchers who championed these markets? They’re watching from the sidelines as lawyers and lobbyists fight battles they never anticipated.
What the Founders Actually Wanted
The intellectual case for prediction markets was always about something grander than entertainment. The core argument went like this: human institutions are terrible at aggregating information. Committees are political. Polls are manipulable. Experts are overconfident. Markets, by contrast, give people skin in the game and let prices reveal what the crowd actually believes.
Hanson’s work on prediction markets grew from a deeper interest in “futarchy” — the idea that governments might someday use market predictions to guide policy. Bet on outcomes, not intentions. Let the collective intelligence of distributed traders replace the guesswork of politicians and bureaucrats.
This was always a utopian project, and the utopians knew it. But they believed that even partial implementations would demonstrate the power of market-based forecasting. Corporate prediction markets would help businesses avoid disasters. Academic prediction markets would help researchers identify promising directions.
Some of this happened. Prediction markets famously outperformed polls in certain elections. Internal corporate markets at companies like Google showed promise before mostly fading away. The research community produced genuine insights about how market mechanisms could improve collective judgment.
But the mainstream breakthrough came via a different route entirely — through crypto speculation and culture war controversy, through meme markets and celebrity gossip, through everything the original boosters never imagined or wanted.
The Sports Betting Convergence Problem
The current moment in prediction markets is defined by convergence with sports betting — a convergence the intellectual founders never sought and cannot control.
DraftKings’ recent moves into event contracts signal where the industry is headed. The company views prediction markets as a natural extension of its sports betting platform, a way to keep users engaged during the offseason, a new revenue stream built on existing infrastructure.
This framing horrifies the purists. For them, prediction markets were supposed to be the opposite of gambling — they were supposed to be useful, to generate information value, to serve purposes beyond entertainment. Watching prediction markets become another product in DraftKings’ portfolio feels like a betrayal of the entire project.
But the purists never controlled the narrative. And as Brian Armstrong and other crypto advocates make their case for prediction market expansion, they’re not emphasizing epistemic hygiene or information aggregation. They’re emphasizing freedom and opportunity and, implicitly, the chance to profit.
The Trust Problem Nobody’s Solving
The deeper issue the original prediction market advocates never fully grappled with is trust. For market prices to serve as truth signals, participants must believe the market is fair — that prices reflect information rather than manipulation, that contracts will settle honestly, that the whole system operates with integrity.
Recent events have not inspired confidence. Polymarket’s security breach exposed vulnerabilities that undermine the platform’s credibility. Allegations of market manipulation on political contracts have generated congressional scrutiny. The offshore, lightly-regulated nature of major crypto prediction platforms creates exactly the kind of opacity that erodes trust.
If prediction markets were going to serve the epistemic function their advocates envisioned, they needed to be boring. They needed to be regulated, transparent, accountable. They needed to feel more like the Bureau of Labor Statistics than a poker room.
What emerged instead is something thrilling and chaotic and probably inevitable. The money came before the infrastructure. The speculation came before the trust. And now the industry is scrambling to build credibility while simultaneously scaling at venture-capital speed.
Where This Leaves the True Believers
Hanson and his fellow travelers face an uncomfortable question: is this what success looks like?
On one hand, prediction markets have never been more visible or more liquid. Ideas that seemed fringe two decades ago are now debated in Congress and covered by mainstream media. The basic concept — that markets can aggregate information — has penetrated public consciousness in ways the researchers never achieved through academic papers.
On the other hand, the implementation looks nothing like the vision. The markets that attract the most attention are often the least useful. The regulatory environment grows more hostile in some jurisdictions even as markets expand. And the core promise — that prediction markets would make institutions smarter — remains largely unfulfilled.
Maybe the best prediction markets can hope for is cultural legitimacy. Even if they never become the policy tools Hanson imagined, even if they never replace polls or punditry in systematic ways, they might gradually earn acceptance as one more valid form of probabilistic expression.
Or maybe this is the typical trajectory of any idealistic technology: invented by dreamers, deployed by merchants, regulated by bureaucrats, and eventually absorbed into the ordinary fabric of commerce. The dreamers always lose control. The merchants always win. And the thing that emerges looks nothing like what anyone originally wanted.
The prediction market pioneers are learning what every revolutionary eventually discovers — that success has a way of changing everything, including the meaning of success itself.