The Year in Tech w/ Jathan Sadowski & Brian Merchant

Episode Summary

Executive Summary: The episode reviews 2025’s tech landscape through a critical lens: the deepening alliance between Silicon Valley and the Trump administration, the use of AI to justify state and labor cuts, the failure of digital sovereignty efforts outside the U.S., and the fragility of the AI bubble. It closes by highlighting 2026 watchpoints: AI politics, data centers, and whether the bubble’s financial structure can hold.

Main Topics: Silicon Valley–Trump alliance and techno-fascism (Priority: 5/5): The hosts argue that 2025 formalized a durable political-economic alliance between the Trump White House and Silicon Valley power brokers like Thiel, Sachs, Andreessen, and Musk, blending state power with tech capital in increasingly authoritarian ways. Doge, administrative-state destruction, and labor fallout (Priority: 5/5): They describe Doge as a technocratic cover for mass federal layoffs and the hollowing-out of U.S. state capacity, with long-term consequences for institutions, public services, and labor. AI as an instrument of restructuring and automation (Priority: 5/5): AI is framed not just as a product trend but as a managerial and political logic used to justify layoffs, reduce institutional knowledge, and degrade creative and entry-level work. International failure of digital sovereignty (Priority: 4/5): The discussion broadens to Canada, Europe, and Australia, where attempts to regulate Big Tech or assert digital sovereignty are portrayed as weak, compromised, or outmatched by U.S. leverage. AI bubble, circular finance, and infrastructure overbuild (Priority: 5/5): The hosts debate the fragility of the AI boom, noting circular investments, depreciating GPU assets, data center spending, and the possibility that OpenAI and peers may be propped up by bailout expectations or liquidity games. Worst person in tech contest: Musk vs. Thiel (Priority: 3/5): The episode ends with the annual audience vote, where Elon Musk narrowly defeats Peter Thiel for worst person in tech, underscoring how both remain central villains but in different ways. 2026 watchlist: AI politics, data centers, and bubbles (Priority: 4/5): The speakers identify the coming year’s key fronts: political opposition to AI, electricity and data center backlash, and whether OpenAI/industry valuations can survive financial stress.

Key Arguments: Silicon Valley’s relationship with the Trump administration is no longer incidental; it is now a durable alliance shaping policy, contracts, and propaganda. The administrative state is easy to destroy and very hard to rebuild, so the damage from Doge and related austerity efforts will outlast the current political moment. AI is being used as a justification for layoffs and for replacing human judgment with low-quality automation that often creates more work rather than less. The U.S. tech stack is being exported globally through geopolitical pressure, making digital sovereignty efforts in Canada, Europe, and elsewhere difficult to sustain. The AI market is structurally bubble-like because it depends on expensive, rapidly depreciating GPU assets, circular financing, and ever-growing capital commitments. OpenAI’s and peers’ real goal may be not durable profitability but reaching a liquidity event or bailout before the bubble bursts. Political opposition to AI is becoming a cross-ideological issue, with the left focusing on labor/environment and the right focusing on censorship, minors, and state power. Peter Thiel remains the more strategic and ideologically consistent architect of the current tech-right ecosystem, even if Elon Musk is the more visible and hated figure.

Data Points: Worst person in tech final vote: Elon Musk won 52 to 48 over Peter Thiel - Annual audience poll deciding the worst person in tech for 2025 Sam Altman vs. Mark Zuckerberg third-place vote: 1,225 vs. 1,223 votes - The third-place runoff was decided by only two votes OpenAI fundraising effort: $40 billion raised in 2025; aiming for $100 billion next year - Used to illustrate the escalating scale of AI bubble financing Projected future AI spending: Trillions of dollars on data centers - Discussion of the capital intensity needed to sustain AI growth AI-related GDP contribution: More than consumer spending - Point made about AI spending propping up U.S. GDP growth DeepSeek timing: February 2025 - Referenced as an early-year challenge to the AI growth narrative Data center moratorium call: Bernie Sanders called for one just days before recording - Used as evidence that AI/data-center politics are shifting

Pivotal Quotes: "Voting matters, folks. Voting matters. You could be the deciding vote in a horse race election." — Host (Paris Marks): Opening remark after referencing a very close vote count "The center cannot hold." — Jathan Sadowski: Describing the collapse of moderate consensus around tech, politics, and public acceptance "AI will do it for you." — Jathan Sadowski: Summarizing how AI is sold as a shortcut while eliminating institutional capacity

Implications: Listeners are urged to see AI less as neutral innovation and more as a political tool tied to austerity, labor displacement, and state power. In 2026, battles over data centers, regulation, and bubble risk may define both tech policy and broader politics.