All-In episodes

Oct 03, 2026 1:22:29

White House Superintelligence Summit, GDP Strength and Midterm Odds

The episode breaks down Trump's White House superintelligence summit, where six frontier-model labs signed a voluntary accord that Sacks says becomes binding through external audits, board oversight, D&O insurance, and SEC/FTC enforcement. David Friedberg argues model-level regulation is impossible and predicts a shift to government allocation of data centers and GPUs for cyber defense. The hosts debate strong GDP and labor data against climbing Treasury yields and bank impairment risk, then split on the 2026 midterms. The closing segment covers the Flydubai hijacking attempt and the hosts' criticism of media outlets that initially framed it as an altercation.

• Science
• AI and tech
• Policy and regulation
• Markets

Oct 01, 2026 52:49

Jake Paul & The Chainsmokers: Fame and Bubble Signs

Jake Paul talks about building a career on Vine, organizing its top creators to demand $1 million each per year, and watching the platform collapse after Twitter refused. He then details his pivot into Team 10, pro boxing via Most Valuable Promotions, angel investing in OpenAI and SpaceX, and a stated plan to enter politics by 40. The Chainsmokers explain their Hype Machine-era rise, why 300,000 daily Spotify uploads distort music economics, and their intentionally hands-on venture fund. In the final stretch the group debates late-stage and secondary-market pricing as bubble signals, with Alex Pall warning that celebrity venture funds are rarely built for the concentration of returns at the top.

• Deals and companies
• Policy and regulation
• Markets

Sep 29, 2026 51:16

Daniel Ek Builds Neko to Fix Broken Healthcare Incentives

Daniel Ek joins the All-In hosts to discuss Spotify's founding and his new preventive-health startup Neko, a $499 annual visit that measures 53 blood markers, images 6,000 skin spots, and includes a clinician consult. He argues the US spends 18% of GDP on healthcare because incentives are built for acute care, not prevention, and that longitudinal multi-modal data plus AI can catch chronic disease earlier. Ek favors open-source AI, but suggests compute scale should be part of the safety conversation; he is not convinced frontier-model pacing is the right frame. He also says Spotify will keep supporting open podcasting standards while building proprietary features, and shares Neko's US expansion plans.

• Deals and companies
• Science
• Policy and regulation
• AI and tech

Sep 26, 2026 1:34:24

AI Token Price War, Anthropic IPO Risk, Meta Muse Pop

This episode centers on a 10-day flood of open-weight models that has flipped token economics from 80/20 closed-source to 80/20 open-source and made superintelligence bans unenforceable. The hosts argue Anthropic's IPO is threatened less by doomsday claims than by its own contradictory leadership, open-source competition, and customers maxing out expensive tokens. They contrast Dario Amodei and the UN's global-governance push with Trump's 'superintelligence' speech and JD Vance's anti-Frankenstein line. Meta's Muse and xAI's Grokbot are hailed as the first consumer AI agents that give ordinary Americans real time and money savings, while putting Amazon and Apple's app-store economics under pressure. Friedberg defends Anthropic's wet lab as ordinary protein-validation science.

• Policy and regulation
• AI and tech
• Deals and companies
• Science

Sep 24, 2026 25:30

From $40K to $40B: Buying Product-Market Fit and Beating PE

Luca Ferrari, CEO of Bending Spoons, tells the All-In hosts how he turned $40,000 left from a failed AI startup into a ~$40 billion market-cap company by acquiring product-market fit. Starting with a $10,000 keyboard app, Ferrari built a shared operating system and an 800-person core team that re-engineers acquired companies like AOL, Eventbrite, Vimeo, and Miro. He says debt costs about 9% until 2031, leverage is 2.5x, and historical unlevered returns exceed 25%, so higher rates may actually help by lowering asset prices. The company screens for large, predictable targets and relies on high talent density; it received 800,000 applications but hired under 300 people last year. Ferrari argues private equity cannot copy the model because it keeps companies separate for eventual sale, preventing shared tech infrastructure and team rotation.

• Deals and companies
• Markets

Sep 23, 2026 55:44

Fixing California: Hilton, Pratt, Ballot Fraud, and the Palisades Fire

Republican gubernatorial candidate Steve Hilton and Spencer Pratt join the All-In hosts to argue California is broken after 16 years of Democratic control. Hilton pushes a cost-of-living platform: tax-free first $150k, $3 gas, flat $73 DMV registration, and cutting regulations. Pratt details his LA mayoral loss and alleges California's ballot-harvesting system is legally corrupt. The hosts spar over voter fraud claims, while Pratt recounts the Palisades fire, his mass tort lawsuit, and plans for data centers on Skid Row.

• Policy and regulation

Sep 22, 2026 22:35

AI's energy wall: Naveen Rao's dynamical computer and 1,000x efficiency bet

Naveen Rao, co-founder and CEO of Unconventional AI, argues that AI's biggest constraint is energy, not model quality, and estimates the industry could run out of power in about three years. He unveiled a new 'dynamical computer' that fuses compute and memory, claiming the first physical chip already generates images at roughly 500 nanojoules per image, orders of magnitude below a GPU. The company's goal is 1,000x power efficiency within three and a half years, with a full data center product expected in about two years. Rao also discussed his track record building Nervana Systems, selling it to Intel, and later joining Databricks. The conversation ended with a Q&A on porting existing models, team composition, and the roadmap to production.

• AI and tech
• Science

Sep 21, 2026 23:52

Under-the-radar AppLovin: mobile game ads, deep learning, and surviving the 92% drawdown

Adam Foroughi, CEO of AppLovin, describes how an advertising network embedded in mobile games, built without early VC funding, grew into a company that reached a $250 billion market cap. He puts the mobile gaming ad market at about $50 billion a year, with AppLovin's own platform around $20 billion in annual ad spend. He also recounts the 2022 drawdown from a $40 billion peak to a $3.8 billion market cap, followed by a ~$6 billion buyback and a recovery to $750 per share. Foroughi argues that advertising is ML 1.0, that search-based AI ads create little new economic expansion, and that discovery ads do. He also gives views on privacy regulation, why AppLovin bought and sold game studios, and why agents won't replace the average shopper's discovery experience.

• Deals and companies
• AI and tech
• Markets
• Policy and regulation

Sep 05, 2026 1:31:56

GPT-6 Hits AGI? Tech Euphoria 2.0, SF Mansion Shortage, NYC Bans AI in Schools & Venezuela Oil Deal

OpenAI released GPT-6, Anthropic and OpenAI debuted new models, while the hosts broke down the AI trade, the San Francisco market, and the Venezuela oil deal. They disagreed on whether an AI bubble is building and whether founders should sell shares. The episode closed with New York City banning AI in schools and the U.S. oil move in Venezuela.

• AI and tech
• Markets
• Deals and companies

Aug 29, 2026 1:36:41

Nvidia's record quarter, Salesforce's rebound, and the US debt standoff

Nvidia delivered a historic quarter: $96.2B in revenue, up 106% year over year, $60B in profit, and 70% growth guidance, pushing back on bubble fears. Salesforce surged 20% after beating estimates and leaning into Anthropic's Claude, reviving the SaaS bull case. With 30-year Treasury yields near 5.2%, Scott Bessent's bond buybacks drew a rebuke from Stan Druckenmiller, and the hosts argued the real problem is Congress's $2T-plus deficits. They also debated Druckenmiller using AI to write his op-ed and Friedberg broke down Moderna's neoantigen cancer immunotherapy, questioning its $500K price tag.

• AI and tech
• Markets
• Policy and regulation
• Science

Aug 27, 2026 1:30:20

Eric Weinstein: Science Stalled, String Theory Wrong

Eric Weinstein joins the All-In interview to argue that American science has been deliberately domesticated: consensus pressure, peer review, and restricted data have turned researchers into a 'scientific procariat.' He traces the break to 1965–75, blames string theory for a 42-year physics dead end, and names Renaissance Technologies as a possible secret continuation of Manhattan-Project-style physics. On funding, he says government must back high-beta people rather than low-beta grants, while Friedberg pushes back that private capital like Jim Simons can fill gaps. The conversation moves from Pati-Salam grand unification and 'boom, vroom, zoom' to UAPs, multi-time adversaries, and AI mining rejected physics. He ends urging scientists to work with the Trump-era science officials he met in Washington.

• Science
• Policy and regulation
• AI and tech

Aug 25, 2026 55:59

Trump's Science Czar on Anti-Science Claims, DEI Cuts, and the China Race

Michael Kratsios defends the Trump administration against anti-science accusations, outlining cuts to DEI and climate research while championing a 'new golden age' for American science. He acknowledges China's R&D surge but insists free-market innovation outperforms central planning. The conversation covers Fauci's damage, STEM pipeline failures, and big bets on fusion, quantum computing, and the moon. Kratsios argues that government must fund basic research and leave applied tech to private capital.

• Policy and regulation
• Science

Aug 21, 2026 1:30:53

Dario Amodei's AI Essay, Regulatory Capture, and the Midterm Wave

The episode debates Dario Amodei's AI essay and the hosts' critique that Anthropic has amplified AI fears. They analyze data center backlash and the risk of an open-source ban, with Sacks opposing FINRA-style regulation. The second half focuses on the 2026 midterms, rising socialist sentiment, and predictions for an AOC presidency.

• AI and tech
• Policy and regulation

Aug 18, 2026 55:58

Flock CEO Garrett Langley on license-plate surveillance, police abuse, and drones

Jason Calacanis interviews Flock Safety CEO Garrett Langley about the company's license-plate readers, which now serve more than 6,000 U.S. cities and have been credited with over a million solved crimes. Langley defends Flock's privacy design — no facial recognition, video, or person search — and walks through new audit tools that have flagged and gotten many police officers fired for improper plate searches. The two wrestle with the core trade-off between safety and privacy, including data-retention defaults, city-level control, and the risks of AI-driven policing. They also discuss Flock's growing drone business and the company's PR crisis after vandalism and backlash.

• Policy and regulation
• AI and tech

Aug 14, 2026 1:39:30

Anthropic's $2T IPO, Zuckerberg's open-source manifesto, and Nvidia's $500B GPU-finance play

Anthropic is reportedly targeting a $2 trillion IPO in October on a $100–120 billion annualized revenue run rate, a 10x year-over-year ramp, sparking a debate over whether open-source models will erode its pricing power. The crew dissects Zuckerberg's 6,500-word manifesto arguing AI is too dangerous to centralize, directly opposing Anthropic's effective-altruist stance. Nvidia's $500B GPU-financing partnership with Goldman and BlackRock treats compute as an asset-backed class, with Anthropic's quarterly earnings called the industry's pace car. They also weigh Amazon's DSP subcontractor lawsuits in New York and New Jersey, and xAI's Grok 4.6 and Grokbot catching the frontier.

• Policy and regulation
• AI and tech
• Markets
• Deals and companies

Aug 13, 2026 1:11:36

Rahm Emanuel on China, Indo-Pacific Deterrence, and the DSA

Rahm Emanuel joins All-In to argue that China's 'win-win' was a dependency trap, advocates a US-led economic bloc to isolate Beijing, and warns Trump's NIH cuts and trade isolation hurt America. He details Indo-Pacific deterrence from Taiwan to the Philippines and Japan, urges a Middle East realignment against Iran, and says Europe's immigration and energy crises were Merkel-era failures, not leftist ideology. Domestically, he backs the Dignity Act, foreign-student tuition subsidies, a 3-year BA, minimum wage with EITC, and attacks the DSA as a separate party.

• Policy and regulation
• AI and tech

Aug 08, 2026 1:15:18

Google AI Exodus, SpaceX Earnings, Airtable's Collapse

The episode dissects Google's AI brain drain as Demis Hassabis moves into a chairman role and Jeff Dean departs to start Discovery Loop, with hosts debating whether Google is smartly prioritizing low-beta AI infrastructure capex over risky frontier model development. They then break down SpaceX's first public earnings, which shocked on revenue but saw the stock fall, and argue Starlink alone could justify a trillion-dollar valuation. The Airtable-to-Bending-Spoons sale at roughly 10% of its peak becomes a case study in the SAS apocalypse and late-stage VC incentives. Finally, the group clashes over whether US data-labeling firms selling expert-curated training data to Chinese labs is a strategic mistake or just a commodity trade.

• AI and tech
• Markets
• Deals and companies
• Policy and regulation

Aug 06, 2026 48:19

Autonomous ships, the 230-to-1 shipbuilding gap, and Saronic's Port Alpha

Saronic co-founders Dino Mavrikis and Vib Alakar join Jason Calacanis to discuss the first autonomous ship rescue in the Strait of Hormuz, the U.S.-China shipbuilding gap, and their plan to mass-produce unmanned warships. Mavrikis says China outbuilds America 230-to-1 and the U.S. built only five commercial ships last year. The pair argues that removing crew systems and vertically integrating production can cut costs and field far more VLS tubes per year. They announce Port Alpha, an 800-acre shipyard in Brownsville, Texas. They also discuss AI rules of engagement, allied sales, and why defense tech is this generation's space race.

• Policy and regulation
• Deals and companies

Aug 01, 2026 1:36:34

Chip Crash, $20B Margin Call, AI Pause Debate, and NYC's Socialist Groceries

The episode digs into the chip stock crash and a $20B hedge fund margin call, the macro backdrop of 5.2% Treasury yields and a $2T deficit, China's open-source AI threat, the AI safety letter from frontier labs, Anthropic's book shredding, and NYC's socialist grocery stores. The hosts debate whether the market correction is momentum-driven or fundamental, with Sacks confident in AI capex and Friedberg wary of leverage. Their views diverge on open-source competition and the sincerity of the AI pause.

• Markets
• Science
• AI and tech
• Policy and regulation

Jul 29, 2026 1:08:35

Four robotics leaders: inspection dogs, home humanoids, and the factory of the future

From the Machina conference in Paris, Jay Calacanis interviews the CEOs of ANYbotics, 1X, Boston Dynamics, and Agility Robotics. ANYbotics sends four-legged inspection robots into oil rigs, offshore wind stations, and explosive atmospheres; 1X has sold out its first 10,000 Neo home-humanoid pre-orders and will ship in 2026. Boston Dynamics is rolling out Spot across 500+ customers and taking a hard line against Chinese robotics in US infrastructure, while Agility Robotics is introducing Digit V5, the first humanoid to work outside safety cages. The recurring tension is between fast-scaling, humanoid autonomy and the dull, dirty, dangerous jobs they will take over.

• AI and tech
• Policy and regulation

Jul 25, 2026 1:33:44

Open-Source AI Fight, Anthropic's $1.5B Settlement, NYC Eviction Politics

China's Kimi K3 triggered a White House debate over banning Chinese open-source AI; Sacks said no decision has been made and argued Anthropic is pushing regulatory capture instead of stopping distillation. Chamath and Friedberg detailed why open-source models are commoditizing AI fast and why a US ban would backfire, while Sacks countered that US frontier labs are still growing explosively. Anthropic settled the largest copyright case in US history for $1.5B after training on pirated books, exposing what the hosts called hypocrisy on distillation. Google and Tesla reported massive capex; Google posted its first negative free cash flow as a public company. The episode closed with a debate on NYC rent-control proposals and eviction bans.

• Policy and regulation
• AI and tech
• Deals and companies
• Markets

Jul 21, 2026 41:39

Mark Cuban on the AI Bubble: Who Gets Wiped Out and Why

Mark Cuban argues the current AI boom is not a traditional bubble that will hurt most Americans, but it could devastate over-leveraged VCs and private equity funds. He highlights that AI implementation is far harder than expected, with enterprise adoption requiring significant human oversight. Cuban sees enormous opportunity for entrepreneurs using AI tools like Lovable to rapidly build businesses, but warns that AI agents drift and need constant management. The conversation also covers politics, NBA parity, and the rise of self-directed healthcare with AI.

• AI and tech
• Science
• Policy and regulation