All-In PodcastNew SEC Chair, Bitcoin, xAI Supercomputer, UnitedHealth CEO murder, with Gavin Baker & Joe Lonsdale
At a glance
WHAT IT’S REALLY ABOUT
Trump’s Deregulation Agenda, Bitcoin Surge, xAI Bet Reshape Power, Markets
- The episode features guest hosts Gavin Baker and Joe Lonsdale with David Friedberg and Jason Calacanis dissecting the post‑election “Trump bump,” focusing on deregulation, debt, energy policy, and America’s competitive position vs. China. They analyze Trump’s nomination of Paul Atkins as SEC Chair and what a more pro‑market, pro‑crypto regulator means for capital markets, accreditation rules, and Bitcoin’s role relative to the U.S. dollar and BRICS. A major portion of the discussion dives into AI: Elon Musk’s xAI supercomputer in Memphis, GPU coherence breakthroughs, scaling laws, and what Grok‑3 could do to the competitive landscape versus OpenAI and others. The show closes with the shocking assassination of UnitedHealth’s CEO, using it to explore anger at the healthcare system, the “oppressor vs. oppressed” mindset, and the risks of personalizing corporate responsibility.
- Throughout, the guests return to a few core themes: deregulation as the primary lever for growth, energy abundance (especially nuclear) as the foundation of prosperity and AI, and competition—both market and geopolitical—as a disciplining force on governments, corporations, and technology trajectories.
IDEAS WORTH REMEMBERING
5 ideasDeregulation and bureaucratic reform are seen as the core lever for U.S. growth and debt stabilization.
Baker and Lonsdale argue the U.S. is like late‑Ballmer Microsoft: structurally advantaged but mismanaged, with layers of counterproductive regulation preventing infrastructure build‑out (rural broadband, EV chargers) and energy expansion. Lonsdale calls for automatic sunsets and data‑driven renewal for regulations, plus reintroducing rigorous civil‑service testing and accountability to reverse decades of bureaucratic decay. Friedberg ties this to the debt ‘death spiral’: you can’t fix the arithmetic without materially raising GDP growth while restraining deficits.
Energy abundance—especially nuclear—is framed as a national security and AI imperative where China is pulling ahead.
Friedberg emphasizes China’s planned jump in electricity capacity (roughly 2 to 8 TW vs. 1 to 2 TW in the U.S.) and links cheap power directly to middle‑class prosperity, industrial competitiveness, and AI scaling. Lonsdale and Baker call nuclear the most practical, green baseload option available now, arguing U.S. regulatory barriers—not physics or know‑how—are the main bottleneck. They see rapid nuclear build‑out as essential to avoid ceding manufacturing and AI leadership to China.
A pro‑market SEC under Paul Atkins could materially alter crypto, fund access, and capital formation dynamics.
Lonsdale, who knows Atkins personally, describes him as fair, rule‑respecting, and innovation‑friendly, in sharp contrast to Gary Gensler’s adversarial ‘gotcha’ enforcement. Baker highlights a 2007 Atkins speech criticizing raising accreditation thresholds, arguing it locks ‘non‑rich’ investors out of upside while protecting them only from downside. While Friedberg defends the SEC as one of the few competent agencies and warns about retail fraud in opaque private markets, others suggest sophistication tests (not wealth tests) and better‑defined rules as a way to expand access without unleashing scams.
Bitcoin is positioned as both a check on fiscal excess and a long‑term threat to dollar dominance.
Friedberg distinguishes Bitcoin from the broader ‘crypto’ space, casting BTC as digital gold and a safe‑haven asset that ultimately could pressure the U.S. dollar, especially if U.S. deficits remain uncontrolled. He notes the irony of Trump vowing 100% tariffs on BRICS currency experiments while simultaneously celebrating Bitcoin’s rise. Lonsdale argues BRICS currency coordination is an authoritarian threat that should be opposed, whereas Bitcoin is a pro‑liberty check on state overreach. Baker ultimately believes BRICS won’t displace the dollar, but Bitcoin eventually might become a serious competitor, forcing better policy discipline.
Elon Musk’s xAI supercomputer is portrayed as a decisive technical and competitive breakthrough in AI scaling.
Baker explains that no one in the industry believed more than ~30k NVIDIA H100 GPUs could be made coherent (each GPU effectively ‘knowing’ what the others are doing) given current networking limits. By redesigning the data‑center architecture from first principles, xAI reportedly made ~100k H100s coherent in Memphis, with plans for 200k and eventually ~1M GPUs. Grok‑3, now training on this cluster, will provide the first real new data point on whether training‑time scaling laws still hold beyond GPT‑4‑class systems; if they do, Baker expects Grok‑3 to become state‑of‑the‑art and potentially overtake OpenAI’s models.
WORDS WORTH SAVING
5 quotesAmerica is like Microsoft under Ballmer: a monopoly that’s been horribly mismanaged. Satya’s first win was just to stop doing really dumb things.
— Gavin Baker
It’s not like our bureaucracies are kind of sort of bad. It’s like the worst company you know went bankrupt 30 years ago, and we just kept pumping money into its worst department.
— Joe Lonsdale
We’re basically on the verge of unlimited energy for all time—if we can just get out of our own way.
— Jason Calacanis
No one thought it was possible to make more than 30,000 Hoppers coherent. Elon figured out how to make over 100,000 work together. What he did was superhuman.
— Gavin Baker
This mindset that there are only oppressors and oppressed—and if you’re an oppressor there’s no limit to what should be done to you—is a really stark and sad commentary on where we are.
— David Friedberg
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome