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

What It Covers

Hard Fork covers three topics across 72 minutes: the regulatory crisis facing prediction markets like Polymarket and Kalshi amid insider trading scandals, journalist Joanna Stern's year-long experiment outsourcing life decisions to AI (documented in her book *I Am Not a Robot*), and producer Rachel Cohn's month attending Brooklyn's Struthers School of Radical Attention.

Key Questions Answered

• •Prediction Market Manipulation: An Anti Corruption Data Collective analysis of 400,000+ Polymarket trades found military and defense long-shot bets win at 52% versus the platform's 14% average win rate — a statistical signal that insiders with privileged government information are actively trading. An Army sergeant allegedly netted over $400,000 betting on Nicolas Maduro's removal from power while involved in related operations.
• •Platform Economics Favor the House: Over 70% of Polymarket users lose money, and Kalshi reports 2.9 unprofitable users for every profitable one. Before participating in any prediction market, recognize the structural disadvantage: you are likely competing against insiders with non-public information, making normal retail participation statistically closer to a casino than a knowledge market.
• •Regulatory Gap Creates Risk: The CFTC, which regulates prediction markets via a historical accident involving futures contract classification, is significantly smaller than the SEC and lacks resources for active surveillance. Prediction markets have lobbied to remain under CFTC jurisdiction precisely because enforcement is weaker — a dynamic identical to how financial firms historically avoided SEC oversight.
• •AI Agents Crossed a Practical Threshold: Stern tracked AI research assistant capabilities throughout her year-long experiment. Tasks requiring a human assistant at the start of 2024 were partially automatable by mid-year via tools like Perplexity. By the book's completion, she assessed that AI agents could handle 100% of those same research, email, and scheduling tasks — a concrete benchmark for evaluating current agent capability.
• •AI Dental Upselling Is Systematic: Dental practices using Pearl AI or Overjet overlay AI diagnostics onto X-rays, and Dental Service Organizations (DSOs) that own multiple practices track whether dentists convert AI findings into procedures. Employees reported pressure to sell treatments flagged by AI. Before accepting expensive periodontal treatment recommended alongside AI diagnostics, seek a second opinion from an independent practice not owned by a DSO.
• •Attention School's Three-Pillar Framework: Brooklyn's Struthers School of Radical Attention, founded June 2023, structures its programming around study, sanctuary, and coalition building. Its core argument is that time spent in activities technology cannot commodify — shared observation exercises, farmer's market sensory walks, character-embodiment seminars — constitutes material political resistance against the attention economy, not merely personal wellness hygiene.