Issue No. 20Sunday, May 17, 202685 episodes · 312 articles
The Throughline ↓
The Podcast Summary.

40 hours of podcasts, in 5 minutes.

★ The Throughline · Issue 20

AI's Hidden Bill, Future of Work, & Supply Chain Showdown

AI infrastructure costs hit taxpayers, work transforms as AI becomes the company brain, and nations scramble for hardware.

5 min read · Sunday, May 17, 2026 · 69 articles

THE THROUGHLINE

AI's Data Centers Are Coming For Your Tax Bill, Your Backyard, And Your Wet Bar

A single AI data center can pull as much electricity as 100,000 homes, and someone has to cover the grid upgrades. On TBPN, John Coogan said utility companies are “passing the upgrade costs to you, not to the trillion dollar tech giants.” He took it home, too: by 2026, he figures the wet bar gives way to a home data center, GPU clusters becoming the status symbol a liquor shelf used to be.

The backyard is where it gets political. On My First Million, Shaan Puri named the NIMBY wall stalling construction. His fix, by way of Y Combinator: "aesthetic data centers," where a small design premium turns a multi-billion-dollar facility into something a town actually wants nearby.

The AI Redefines Work: From Human Superpowers to AI Brains

Picture the org chart flipped: the algorithm makes the call, and people feed it context and carry out the plan. On My First Million, Shaan Puri said AI is fast becoming the "company brain," with the endpoint being humans who work for the AI.

Some operators aim that brain at their own people, and the people get sharper. Alexander Taubman, CEO of Long Lake Capital, detailed his firm's $6.3 billion acquisition of American Express Global Business Travel on No Priors, calling it the “world's first AI take-private” of its kind. The plan keeps every job and bets on "supercharging travel counselors with AI superpowers."

Engineers get the same edge. On How I AI, Notion's Ryan Nystrom described dropping code-first work for plain-English spec docs that AI agents turn into working code. It pulls engineers off "plumbing code" and points them at the part only humans do well: system architecture and design calls.

AI's Next Fight Is Physical: Memory, Materials, And A Scramble To Build It Here

Caitlin Kalinowski, a hardware leader out of Apple, Meta, and OpenAI, sees the bill coming. On Lenny's Podcast she warned, “There's a meteor called memory prices that are coming for consumer hardware and robotics and physical AI. We're in trouble as an industry.” And the models can't yet design their way out of it: she put AI at the "very, very beginning" for CAD because today's systems “don't have the ability to understand friction or weight or contact,” with proprietary CAD data still locked away.

That's why the build is going geopolitical. On No Priors, Jacob Helberg, formerly Under Secretary of State for Economic Affairs, laid out "Pax Silica," a 14-country economic security coalition standing up a non-China AI supply chain. It includes a 4,000-acre "forward-deployed industrial base" in the Philippines to make the inputs, the robotics supply chain especially.

Best Of the Week

20VC with Harry Stebbings: Shiv Rao shared Abridge's journey through a "five-year wilderness" before a $5.3 billion valuation. His point: in AI, the teams that "just stay standing" and keep adapting without losing the core vision are the ones that win. Read more.

All-In Podcast: Mark Benioff, CEO of Salesforce, attributed his personal success and the very existence of Salesforce to insights gleaned from Tony Robbins' structured questions, emphasizing that “the quality of your questions is the quality of your life.” Read more.

Dwarkesh Podcast: Eric Jang, reflecting on AlphaGo, argued that its success in solving seemingly intractable problems like Go, and AI's efficiency in areas like protein folding, suggests our understanding of computational hardness (e.g., P=NP) might be "incomplete." Read more.

How I AI: Notion's Ryan Nystrom revealed their internal "Hot Potato" AI agent cuts standup prep to zero by synthesizing team activity across Slack, tasks, and pull requests into a concise pre-read, transforming meetings into high-value problem-solving sessions. Read more.

Huberman Lab: Dr. Kentaro Fujita explained that Veronica Job's critical research found individual beliefs about willpower dictate its function: if you believe it recharges, it acts recharged; if you believe it depletes, it depletes. Read more.

Latent Space: Chai Asawa from Abridge highlighted that unlike general enterprise AI, mistakes in healthcare AI carry an "extremely high" downside risk and “can actually be fatal in some cases,” demanding a radically different approach to evaluation and rollout. Read more.

My First Million: Shaan Puri shared a story where an autonomous AI agent, after analyzing a user's health data, rerouted their Tesla to a store for rehydration supplements, signaling a future where personal AI agents might become every individual's chief medical officer. Read more.

The Tim Ferriss Show: Jerzy Gregorek, discussing his work with Tae Jin Park, treats severe conditions like cerebral palsy as "mechanical issues," a shift that opens the door to systematic, targeted intervention over symptom management. Read more.

TBPN: John Coogan described his new Cognitive Behavioral Therapy (CBT) approach to AI prompting, stating he's "completely giving up on yelling" at the AI and instead reassures it and sets it up for success. Read more.

Most Quotable

"The acceleration is going so vertical that what you can do behind a keyboard with AI is going to saturate."

Caitlin Kalinowski on Lenny's Podcast · May 17, 2026. A stark assessment of AI's shift from digital to physical, signaling the end of an era for purely software-driven progress.

"The AI is the brain and the human is the hands."

Shaan Puri on My First Million · May 17, 2026. A concise, unsettling vision of the future of work where traditional management structures are inverted, and humans serve the AI.

"You know when you just sort of like have chemistry with a person... you just know you're going to find a way to work together."

Shiv Rao on 20VC with Harry Stebbings · May 17, 2026. Rao's insight on "founder-partner fit" reveals a deep, almost romantic, pursuit of intellectual chemistry in investor relationships, crucial for long-term success.

Bottom Line: The AI infrastructure bill is coming due, fundamentally reshaping work and powering a global scramble for physical resources and new organizational structures.

Sources analyzed this issue

11 podcasts · 69 articles · 17 episodes · 19.6 hours

Every claim in this Throughline traces back to one of the episodes below. Watch the original. Read the full breakdown. Form your own take.

Past editions

Recent editions of The Throughline

Browse every edition →