Just finished the latest episode of the All-In Podcast: Democrats won big in the midterm elections, $4 trillion in AI stocks vanished in a single day, and OpenAI has overturned a foundational assumption in computer science. Midterm Election Recap: Democrats Sweep Both Chambers; Traditional Red States Flip Blue In this midterm election, Democrats secured majorities in both the House and Senate, along with multiple governorships—even flipping traditionally red states like Texas and Kansas. The four hosts analyzed, from their respective perspectives, the root causes behind the Republican Party’s collapse. Sacks attributes the defeat primarily to the Iran war and its ensuing economic fallout. The war triggered sharp spikes in crude oil and diesel prices, which drove up consumer goods and food costs, reigniting inflation and completely dashing the Fed’s expectations for rate cuts. Mortgage rates surged back to high levels, directly undermining housing affordability for ordinary Americans. He also noted that the war shattered the longstanding political alliance between Republicans and young male voters, as well as independent podcast circles (e.g., Joe Rogan, Tucker Carlson), resulting in a clear and significant loss of this voter bloc. Friedberg’s focus was on deeper fiscal imbalances. He emphasized that the core issue is unchecked government spending. After the proposed 2025 reform of the Department of Government Efficiency (DOGE) was blocked and ultimately abandoned, annual deficits of trillions of dollars have pushed up costs across society. Ordinary citizens have seen no share of economic prosperity, and this growing disparity drove voters to support candidates advocating for greater government intervention and wealth redistribution. Chamath argued that ideological labels no longer matter—what’s at stake is a fundamental breakdown in America’s economic infrastructure. Over the past several years, regardless of macroeconomic conditions, asset holders have seen their wealth consistently grow, while wage earners have suffered severe erosion in real purchasing power. This voter anger is essentially a retaliatory vote against economic inequality. Jason contends that voters rejected the incumbent party because Trump broke two key campaign promises: resolving inflation and avoiding new wars. Combined with inflationary volatility and enforcement controversies stemming from his tariff policies, his support among moderate voters has plummeted. Several Self-Rescue Strategies for Trump’s Remaining Term The hosts also offered concrete political recommendations for how Trump could navigate the next two years. First, immediately end the war to bring down global oil prices, which would reduce fuel, transportation, and food costs—creating room for the Fed to cut rates and fundamentally alleviating the affordability crisis. Second, shift toward pragmatic, transactional governance. Abandon pointless partisan battles; in exchange for preserving border control achievements, negotiate with Democratic lawmakers on populist economic policies—such as raising the minimum wage, exempting low-income households from federal taxes, stimulating new housing construction, and restricting public officials from stock trading. Sacks offered an intriguing historical analogy: both Clinton in 1994 and Obama in 2010 rebounded from midterm losses by adjusting policy and ultimately secured re-election. He warned against repeating Bush’s 2006 mistake—doubling down on troop surges in Iraq, which led to even greater political collapse. $4 Trillion in AI Stocks Vanish Overnight as Local Politics Block Data Center Approvals Markets lost nearly $4 trillion in market value in a single day due to fears that AI compute infrastructure is grinding to a halt. Anthropic’s and OpenAI’s IPO prospects have been dented as a result. The core conflict lies in local opposition to massive data centers. Communities and newly elected governors are protesting that these facilities are driving up local electricity bills and consuming vast quantities of water—while simultaneously benefiting from massive tax breaks. This dynamic—where tech giants profit while local residents bear the costs—is fueling widespread resentment. Chamath provided a crucial analysis of the industry’s shifting landscape. Slowed approvals will severely harm newer cloud providers and frontier foundation model companies reliant on leased compute power. Meanwhile, hyperscalers with their own energy and infrastructure reserves—Amazon, Google, Microsoft, Meta, SpaceX—will gain control over pricing and allocation of scarce compute resources. This political turmoil is likely to accelerate industry consolidation, further entrenching the dominance of incumbents. Jason and Sacks also warned of a longer-term risk: if U.S. states implement strict regulations or freeze approvals en masse, data centers and compute infrastructure will be forced to relocate to Canada, Mexico, Southeast Asia, or the Middle East. The exodus won’t just take high-paying engineering jobs—it will also drain tax revenue. The U.S. could transform from a net exporter of compute power into a net importer of tokens. While this may sound hyperbolic, the direction is worth heeding. Science Corner: OpenAI Has Overturned a Foundational Assumption in Computer Science At the end of the episode, Friedberg introduced a theoretical breakthrough. Since its proposal in 1999, the Exponential Time Hypothesis (ETH) has served as a cornerstone of theoretical computer science—establishing that NP-complete problems (e.g., 3-SAT) cannot be solved in sub-exponential time—and forming a critical pillar in the broader inquiry into whether P equals NP. OpenAI recently published a 175-page paper demonstrating, through an entirely novel methodology, that 3-SAT can be solved in sub-exponential time—mathematically refuting ETH. This implies that all problems capable of rapid verification might also be efficiently computable—reigniting serious consideration of P=NP. This breakthrough not only reveals AI systems’ transformative potential for solving extremely difficult pure mathematics and foundational scientific problems but also compels the academic community to fundamentally reevaluate computational complexity theory.