Show Notes
Dick Costolo and Paul Kedrosky discuss the “myth of the CEO,” criticizing the worship of godlike tech leaders and the recency bias that elevates celebrity executives despite evidence of poor leadership. They argue management books and public CEO personas create misleading templates that founders try to imitate, often becoming inauthentic and destructive. Costolo recounts a Twitter board story contrasting Jobs-style focus with Bezos’s “do everything” approach to show there are many paths to success. They emphasize good CEOs admit what they don’t know, encourage dissent, manage to outcomes rather than checklists, create many paths to “yes,” and move quickly while correcting mistakes instead of preventing them. They also cover hiring and firing decisively, the dangers of fear-driven cultures, performative transparency, and overreliance on ratings like Glassdoor.
--
Two of the sharpest minds in tech, Dick Costolo and Paul Kedrosky, pull back the curtain on AI, startups, and the future that’s rushing toward us. With insider access, fascinating guests, and a healthy dose of irreverence, they’ll make you laugh while they explain how the world might end (and how we might still save it).
More from The Dick & Paul Show
Dick Costolo and Paul Kedrosky make the case that Meta’s Muse won’t just win the race for personal AI agents. They think it could knock the frontier labs back on their heels.
Dick and Paul dig into why everything — restaurants, faces, portfolios, hiking trails, even AI chatbots — is starting to look exactly the same. From a Soho frozen yogurt line to Instagram Face to correlated markets flash-crashing in unison, they trace how social proof, algorithms, and the compulsion to "have done the thing" are quietly erasing variety everywhere you look. Plus: why ChatGPT always picks the number 7, why there are 900 cans of bear spray on a trail with zero bears, and what a fake barn in a Don DeLillo novel has to do with all of it.
Dick and Paul dig into why the "Terminator" version of AI risk is the wrong one to worry about, using the actual data: a farming AI study where removing one moral guardrail led autonomous tractors to drive over animals in 84% of encounters to save fuel (while carefully avoiding rocks, which could damage the equipment).
Ozempic is quietly distorting U.S. jobs data, and your favorite wellness podcast might be lying to you — Dick and Paul debunk viral health claims (bananas, creatine, Chernobyl) and explain why lifting weights matters more than any biohack.
Dick and Paul take a deep dive into the AI data center spending boom and why it's different from — and in some ways scarier than — the dot-com fiber-optic bust. They break down why lenders are pouring hundreds of billions of dollars into these buildouts without much regard for whether the underlying business makes sense, simply because Big Tech's balance sheets are so clean (a dynamic they compare more to the 1880s–1920s railroad boom than the fiber bust). They dig into why so much of this spending is moving off balance sheet into special-purpose vehicles, why that structure echoes Enron, and how a recent SEC "no action letter" is now letting data center debt get syndicated the same way mortgage-backed securities were before the 2008 financial crisis. They also cover data suggesting frontier AI models have largely plateaued in capability even as price premiums stay sky-high, and how a chain reaction from the SpaceX IPO rippled into AI stock selloffs and even the U.S. Treasury's own bond-buying moves. It's a wide-ranging, systems-level look at where all this AI money is really coming from — and where the risk is quietly piling up.