Can the U.S. Rein in Prediction Markets? + Joanna Stern on Her Year of A.I. Experiments + Our Producer Goes to Attention School

Episode Summary

Executive Summary: This episode centers on two big tech themes: the rapid, poorly regulated rise of prediction markets and a pair of reflective segments on AI and attention. The hosts argue prediction markets are being undermined by insider trading, gambling incentives, and weak oversight, then pivot to Joanna Stern’s year-long experiment living with AI and Rachel Cohn’s experience at Attention School, exploring how people are trying to make technology more useful, humane, and less attention-fracturing.

Main Topics: Prediction markets and insider trading (Priority: 5/5): The hosts discuss the explosive popularity of prediction markets and the growing evidence that people with privileged information—military personnel, insiders, and possibly event participants—are using them to profit unfairly. Regulation and market integrity (Priority: 5/5): They debate whether prediction markets should be regulated more like securities markets, with stronger enforcement, age verification, self-exclusion, and ad limits, rather than relying on company self-regulation. Joanna Stern’s AI immersion and book (Priority: 4/5): Joanna Stern joins to discuss I Am Not a Robot, a year-long reporting experiment using AI for work, parenting, travel, dental research, and even self-reflection, highlighting both utility and hype. AI in daily life: useful, flawed, and manipulative (Priority: 4/5): The conversation covers where AI is genuinely helpful (agentic tasks, note-taking, writing support, wearables) and where it can mislead or be exploited, like dental upselling and inaccurate child-friendly advice. Attention School and reclaiming focus (Priority: 4/5): Rachel Cohn describes attending the Struther School of Radical Attention in Brooklyn, where participants practice unusual attention exercises to resist commodified attention and rebuild shared human presence. Tech resistance as a cultural movement (Priority: 3/5): The episode frames attention school as part of a broader counterculture to tech saturation—one that resembles earlier anti-industrial or environmental movements and seeks community, meaning, and political resistance. Human identity in the AI era (Priority: 3/5): Across the Joanna and Rachel segments, the hosts probe what remains distinctly human—play, curiosity, embodied perception, and social connection—as machines become more capable.

Key Arguments: Prediction markets are increasingly compromised by insider information, making them feel rigged and discouraging ordinary participation. Insider trading on prediction markets is not just unfair to traders; it undermines trust and market liquidity, just as it would in stock markets. The CFTC is too small and ill-suited to oversee a fast-growing prediction market sector, so a stronger regulatory framework is needed. Prediction markets should regulate for both gambling harms and market integrity, including self-exclusion, age checks, advertising limits, and anti-insider-trading enforcement. Joanna Stern’s AI experiment shows that AI can be genuinely useful for autonomous tasks, but it is also prone to flattering users and reinforcing what they already want to hear. AI’s strongest near-term use cases may be in agents, productivity, and wearables, while humanoid robots remain mostly hype. Dental AI illustrates a dark side of AI adoption: diagnostic overlays can be used to upsell expensive treatment and pressure patients unnecessarily. Attention School’s exercises may look odd, but they create community and a shared practice around resisting the commodification of attention. The school frames attention as political: spending time in non-commodifiable, embodied, communal activities can be a form of resistance to big tech. Technology resistance is becoming a cultural movement that gives people alternatives to constant partial attention and digital overload.

Data Points: Average win rate on polymarket for military-related long-shot bets: about 52% - Cited by the Anti-Corruption Data Collective when discussing possible insider betting Average win rate on the platform overall: 14% - Used as a baseline to show military-related bets outperform unusually often Users who lose money on Polymarket: more than 70% - Reported by the Wall Street Journal as part of the risk profile of prediction markets Unprofitable users per profitable user at Kalshi: 2.9 to 1 - Based on data from the past month, cited to show structural losses for most users Prediction markets settled analyzed on Polymarket: more than 400,000 - Anti-Corruption Data Collective analysis over the last five years States/countries blocking prediction market sites: 27 sites blocked in Brazil - Brazil blocked Kalshi, Polymarket, and others for illegal gambling Age range at Attention School: 7 to 70 - Rachel describes the school as serving participants across ages Cost of one class: $250 - The Times paid for Rachel to attend one class; most others were free Year of Attention School founding: June 2023 - Rachel notes the school was founded shortly before AI became a mainstream concern Temperature change in Paris sensor allegation: 18°C to 22°C - Recorded jump at Charles de Gaulle airport on April 15 amid suspicious betting activity

Pivotal Quotes: "I think these insider trading scandals just show like right now, we are sort of at a preregulatory Wild West moment for these prediction markets." — Casey Newton: Used to characterize the current state of prediction market oversight "What we're getting actually is just people just betting on the military operations that they're involved in." — Kevin Roose: Summarizing the real-world incentive problems in prediction markets "I thought it was a little bit of a full circle moment because the whole book, I kind of am saying, like, AI is this mirror and it's going to tell you basically what you want." — Joanna Stern: Explaining how AI influenced her career decision and the broader thesis of her book

Implications: Prediction markets likely need firmer rules to survive as credible information tools. AI is becoming useful but also manipulative, and attention itself is emerging as a political and cultural battleground where people seek ways to stay human.