Derek Thompson left The Atlantic after 17 years to go independent — timing it to a #1 NYT bestselling book

Summary

Derek Thompson left The Atlantic after 17 years to launch an independent Substack, timing the move to coincide with Abundance, the book he co-authored with Ezra Klein, hitting number one on the New York Times bestseller list. He was 22 when he joined in 2008 and is 39 now. The decision came down to age. Wait another 11 years and he'd be 50, likely past the window where starting something new feels like a genuine bet rather than a late pivot.

Independence

Money was not the driver. Thompson says The Atlantic pays well, the book paid well, and no retention offer would have changed his mind. What he wanted was the cognitive reset that comes with leaving an institution. After 17 years, how you process the world gets locked into what you're optimizing for at one place. He closes off certain kinds of experimentation. He now runs his own Substack as CEO, CTO, COO, and, as he notes, the comms intern.

Ezra Klein, whom Thompson told three days before Abundance launched, pointed out that The Atlantic is a strong platform. Thompson had already decided to go.

How became number one

The book sat in the two-to-three range for the first four weeks, losing badly to John Green's Everything Is Tuberculosis. Thompson describes the gap as a blowout. It spiked to number one on week five as discourse built around it. His book agent said she'd never seen that trajectory before.

Thompson argues that television outperforms podcasts precisely because it's a worse product. A 10-minute segment with Fareed Zakaria makes people want the book. Four hours with Lex Fridman exhausts the subject. Podcasts matter because the real product of a nonfiction book is the conversation it generates, not the object itself. But the conversion mechanism is still short-form TV.

Publishers control almost none of their own distribution channels. Authors are essentially begging Zakaria, Bari Weiss, Ben Shapiro, everyone else because all meaningful channels of exposure are owned by someone else. The one exception in this case was Klein, whose podcast is arguably the single most effective platform for launching nonfiction books.

SpaceX vs. California high-speed rail

Thompson pushes back on the clean private-good, public-bad comparison. SpaceX is extraordinary, but the proof of concept for rocket technology required moving dozens of people to space. Building transportation infrastructure for 50 million Californians is a categorically different problem. SpaceX and Tesla have collectively received billions in government contracts and loan guarantees. The private sector success story was partly publicly financed.

On California high-speed rail specifically, Thompson doesn't defend the execution. A $33 billion bond was approved by voters and several billion dollars were spent constructing track that connects nothing to anything. He reads that as a failure of execution and possibly priority-setting, not as evidence the government should exit infrastructure entirely. His framework is a mixed economy with muscular government for genuine public goods and private sector for things that can be monetized. The real gap is the absence of a SpaceX-style model applied to chronic disease, Alzheimer's, or carbon capture.

NIH cuts

Thompson calls the proposed 40% cut at NIH under HHS horrendous. That amounts to roughly $18 billion annually. You could achieve the same deficit reduction by trimming the corporate income tax cut by 0.4%.

GLP-1 drugs illustrate the point. The entire drug class traces back to a federally funded scientist examining the saliva gland of a Gila monster. No VC would have funded that research. The basic science wasn't patentable. The commercial application came decades later. Knowledge which can't be patented on discovery—DNA structure, neurological mechanisms, metabolic pathways—requires public funding precisely because the private sector can't capture the return.

Thompson addresses Jay Bhattacharya by name, saying he knows him and doesn't believe Bhattacharya actually wants these cuts deep down. He'd like to talk through how to reform NIH without gutting it.

Zohran Mamdani

Thompson released an interview with New York City mayoral candidate Zohran Mamdani the morning of this conversation. He describes Mamdani as unusually charming and a skilled listener, while disagreeing with him on two core issues.

On rent freezes, Thompson argues you cannot cap the price of a good whose supply you're trying to increase. The grocery store analogy is that capping receipts at $50 would just stop stores from stocking shelves. On public sector unions, NYC subway construction costs run four times higher per mile than the international average partly because union staffing levels are four times the global norm. Thompson asked Mamdani directly whether he'd engage with public sector unions on this. Mamdani disagreed.

What Thompson found notable was Mamdani's stated openness to changing his mind. He's moved from supporting defund the police to actively supporting increased police funding, arguing officers are being forced into social worker roles that crowd out actual crime-solving. He's also shifted on private developers, now acknowledging a role for private construction in housing supply rather than pushing for government-built housing.

Thompson describes himself as a liberal and Mamdani as a socialist. They won't agree on everything. But that outcome-first, process-flexible thinking is genuinely useful in a politician or executive.