Billionaire Game Theory + We Are Not Ready for A.G.I. + Election Betting Markets Get Weird

Episode Summary

Executive Summary: The episode examines three intertwined power-and-influence stories: Jeff Bezos’s intervention to stop Washington Post presidential endorsements, revealing tensions between billionaire ownership and journalistic independence; Miles Brundage’s warning that AI progress is outpacing society’s readiness; and the rise of Polymarket, whose election odds and crypto-driven betting may be shaping, and potentially distorting, public expectations ahead of Election Day.

Main Topics: Bezos, the Washington Post, and election endorsements (Priority: 5/5): The hosts unpack Jeff Bezos’s decision to block a planned Harris endorsement and end the Post’s presidential endorsement tradition, debating whether the move was about trust, neutrality, business hedging, or avoiding retaliation from Trump. Billionaire incentives and political risk (Priority: 5/5): They argue tech CEOs are not just donors but gatekeepers of major information platforms, making their election behavior consequential for moderation, retaliation risk, antitrust pressure, and post-election legitimacy disputes. OpenAI and AGI readiness (Priority: 5/5): Miles Brundage explains his role on OpenAI’s AGI readiness team, why he left, and why he believes neither frontier labs nor society are ready for rapidly advancing AI systems. How individuals should prepare for AI (Priority: 4/5): Brundage offers practical advice: use current AI tools, think about career resilience, save money, and protect against deepfakes—while stressing that policy guardrails are still lacking. Polymarket and election forecasting (Priority: 5/5): The final segment analyzes the rise of prediction markets as election barometers, especially Polymarket, and questions whether betting odds are genuinely predictive or just another channel for hype and manipulation. Crypto, wash trading, and market credibility (Priority: 4/5): The discussion raises concerns about crypto-linked betting markets, including geo-blocking evasions, a large French bettor influencing odds, and possible wash trading inflating activity and perceived legitimacy.

Key Arguments: Bezos’s announcement was poorly timed and undermined the Post’s credibility more than it restored trust. A newspaper’s opinion section can create perception-of-bias concerns, but eliminating endorsements two weeks before an election looks suspicious and selective. The real worry about media trust is not just bias but the sense that outlets are controlled by powerful owners. Tech billionaires may be hedging against Trump because of retaliation risk, but also because Democratic antitrust pressure threatens them too. If Trump returns, tech leaders are likely to be far more compliant than in 2016, with less resistance and soul-searching. The aftermath of the election could be highly volatile, so platform moderation decisions by Meta, X, Google, and others may be critical. Brundage argues AI progress is real, fast, and not merely hype, even if outsiders remain skeptical. He believes no frontier lab is fully prepared for AGI and that society lacks the governance and economic planning needed. Prediction markets may sometimes be informative, but Polymarket’s odds can be distorted by large bets, wash trading, and self-selected users. The hosts contend that election betting markets may be more useful for creating narratives than for producing better forecasts than polls.

Data Points: Washington Post subscriber cancellations: more than 250,000 - Reported after Bezos ended presidential endorsements; described as roughly 10% of the subscriber base. Washington Post subscriber base share: roughly 10% - Estimated share represented by the 250,000 cancellations. Blue Origin NASA lunar lander contract: $3.4 billion - Cited as another business interest Bezos has before the government. Amazon Web Services contract at stake: $10 billion - Referenced in the 2019 lawsuit tied to Trump administration animus toward Bezos. Polymarket electoral college betting volume: in the region of $100 million - Amount wagered on the U.S. presidential election market as of about a week before the episode. Large Trump bet on Polymarket: about $30 million - A single French trader’s bet materially moved the odds toward Trump. Polymarket odds shown in episode: Trump 66% / Harris 34% - Current top presidential market odds at the time of discussion. Predict It betting cap: $850 - Contrasted with Polymarket’s uncapped betting to show why Polymarket attracts larger wagers. OpenAI tenure for Miles Brundage: 6 years - He said he worked at OpenAI for six years before leaving. AI systems can solve reasoning tasks: examples include LSAT problems and paper feedback - Brundage used this to argue AI is already practically powerful. Global gambling losses projection: $700 billion by 2028 - Mentioned from a Lancet paper about the broader expansion of gambling.

Pivotal Quotes: "“democracy dies in darkness”" — Casey Newton: Used to describe the Washington Post’s identity and the question of what values the paper should embody. "“it would be premature to slam the brakes, but we should be installing some brakes.”" — Miles Brundage: His view on AI governance: not a full pause, but stronger safeguards and policy constraints. "“AI is a real thing and not science fiction.”" — Miles Brundage: His bumper-sticker summary of why people should take current AI capabilities seriously.

Implications: Listeners are being warned that billionaire ownership, AI acceleration, and election gambling platforms could all shape post-election outcomes. The episode suggests scrutiny of platform governance, media independence, and AI policy will matter more after Election Day than before.