Is tech-enabled gambling just that—or is it reshaping work, risk,

and decision-making in America?

Editor’s note: Wondering if there were an angle on the explosive growth of online betting suitable for this newsletter, we dispatched our intrepid contributing-editor-in-training, ChatGPT, to investigate. Here, with considerable human input, is his report:

_______________________________

SOME SAY THAT if you watch March Madness closely enough, you can see the future of work hiding in plain sight. It doesn’t look like a job fair or a coding boot camp. It looks like a phone screen lighting up with live odds, constantly shifting probabilities, and tiny prompts inviting someone to predict what will happen next.

For many—maybe most—Americans, tech-enabled gambling feels like a cultural subplot, an entertainment upgrade layered onto sports. But for others, something more consequential may be unfolding beneath the surface. They see betting platforms becoming mass-market interfaces for probabilistic reasoning, places where millions of people encounter uncertainty not as a vague feeling, but as a number.

For these people, the most important shift is not that more Americans are betting. It’s that betting has become a structured interaction with predictive systems. Odds update in real time. Models recalibrate instantly. Market prices adjust to new information within seconds. In this environment, outcomes are framed not as certainties but as likelihoods. The language of percentages replaces the language of guarantees.

That dynamic echoes a broader argument made by statistician (and former professional poker player) Nate Silver in The Signal and the Noise: Effective forecasting requires thinking in probabilities rather than absolutes. Silver’s work emphasizes that good judgment involves calibration—assigning realistic odds, updating beliefs when new data arrives, and resisting overconfidence.

Prediction markets make that process visible. They display collective estimates in real time. They force participants to confront uncertainty numerically. In theory, engaging with such systems could reinforce habits of probabilistic thinking that are increasingly valuable in an A.I.-driven economy.

Not surprisingly, industry leaders frame it that way. Tarek Mansour, CEO of Kalshi, has argued that prediction markets help “price the future” by aggregating dispersed information into tradable signals. In that telling, markets are not merely entertainment; they are mechanisms for forecasting, tools for organizing collective intelligence.

Viewed through that lens, betting platforms begin to resemble simplified forecasting exchanges—informal laboratories where users practice assigning probabilities to uncertain events.

And yet, this interpretation is far from settled. Behavioral economist Richard Thaler has spent decades demonstrating that markets do not eliminate human bias. Individuals remain overconfident. They overweigh recent events. They chase losses. Even when probabilities are clearly displayed, judgment is often distorted by emotion and cognitive shortcuts. Exposure to markets, in this view, does not automatically create disciplined forecasters. It can just as easily magnify familiar behavioral errors.

Addiction psychiatrist Anna Lembke adds a further complication. Digital platforms engineered around rapid feedback, reward loops, and frictionless participation can intensify compulsive behavior. When prediction tools are embedded inside personalized apps—complete with notifications, micro-bets, and constant engagement prompts—the psychological pull of participation may outweigh any incidental educational benefit. The structure that displays probabilities may simultaneously exploit human vulnerability to risk and reward.

This is the tension at the heart of the prediction economy. On one side is the argument that widespread exposure to probabilistic systems could normalize a healthier way of thinking about uncertainty—one aligned with the demands of an A.I.-saturated workforce for whom decisions are increasingly data-driven and model-informed.

On the other side is the warning that these platforms are optimized primarily for engagement, not education. If incentives reward time-on-app rather than calibrated judgment, the result may be amplified bias rather than improved reasoning. Nate Silver describes the current gambling industry as one that functions, for most, as a tax on the uninformed—an “IQ tax,” he’s called it—and says that the house/industry has a massive, often engineered advantage. He has long argued for stronger, standardized regulation for the sports betting industry to ensure that it operates fairly.

Which brings us back to March Madness. This month’s contests provide a vivid stage for this debate because they compress uncertainty into a single month. Millions of people track probabilities, reassess predictions, and watch markets move in response to new information. The spectacle becomes more than a tournament; it becomes a public demonstration of how humans and algorithms interact under conditions of uncertainty.

But the future of work is emphatically not about gambling; it is about the normalization of probabilistic interfaces across society—in finance, logistics, healthcare analytics, hiring platforms, and A.I.-assisted decision tools. Betting apps are simply the most visible consumer example right now.

The deeper question is whether regular interaction with A.I.-driven probability systems—wherever they appear—will cultivate better decision-makers or simply expose human cognitive limits at greater scale. The answer will depend less on the mathematics of forecasting and more on the incentives embedded in the platforms themselves.

And that is where the story moves from entertainment into economics—and from March Madness into the next decade.