Models (3)

• Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, while OpenAI released GPT-6 Cyber; the labs restricted their highest-risk cyber capabilities to vetted organizations, and Fable remained broadly available with cyber and biology safeguards.
• Alex Finn called Fable 5.2 the best model he had used, citing faster, cheaper, stronger recent Anthropic releases and a complex video game built over two weeks without his input. Sarah Hooker argued the changes mainly reflect cost pressure, while Manu Sharma credited broad scaling across reinforcement learning, multimodal capability, and model behavior.
• Alex Finn said Grokbot's harness delivers more value for everyday knowledge work than Claude for his team, while Sarah Hooker said Anthropic deliberately keeps its model and harness tightly coupled. Manu Sharma distinguished routine work suited to smaller models from long-horizon science and engineering tasks that still benefit from frontier models.

Safety (2)

• GPT-6 Cyber was rated critical, OpenAI's highest preparedness risk tier, and reportedly stayed within its stated scope 99.93% of the time, up from 94.49% for GPT-5.6. The host framed the release as a response to rising cyber threats and companies' demand for stronger defenses.
• Fireworks AI said attackers accessed customer datasets and unfinished models after using internal credentials, then attacked computers holding research data. The company blamed model-access theft and resale; CEO Lynn Xiao said the attackers' behavior suggested AI agents were involved.

Enterprise (1)

• Manu Sharma said enterprise integrations, IT controls, memory, and accumulated context create substantial application-level lock-in, even as small teams switch tools freely. Sarah Hooker warned that open-weight models and automated self-improvement are making API dependencies less sticky.

Markets (2)

• Anthropic's updated IPO filing set a $70 to $78 share range, implying up to $1.405 trillion fully diluted value; strategic investors agreed to buy $26 billion of stock. Alex Finn questioned how a model company could grow tenfold without a durable moat, while Sarah Hooker argued Anthropic must move into vertical products such as legal, tax, healthcare, and finance.
• Manu Sharma said Anthropic generates more than 80% gross margins and that Claude writes 90% of its own training code, giving the lab an internal advantage. He said improved compute efficiency supports those economics, while lower prices could drive usage growth above the price reduction.

China (2)

• Sarah Hooker said export controls have been difficult to enforce because Chinese firms can use data centers in nearby countries, while China is also developing domestic chips and optimizing models for efficiency. Manu Sharma added that Chinese labs can use remote compute through APIs, so chips need not sit on the mainland.
• The host reported allegations that Nvidia executive Alice Wong helped divert chips to Chinese companies, including an alleged $7.5 billion shipment, and said Nvidia's board found no reason to believe secrets or national security were compromised. The board opened an independent investigation, and Reuters had not independently verified the reported figures.

Chips (1)

• Alex Finn argued Nvidia should not remove Jensen Huang absent evidence he deliberately harmed US security, calling Huang central to the company. Sarah Hooker said the export-control story highlights compute as a strategic resource, while Manu Sharma questioned whether restrictions can stop Chinese AI development.

Regulation (1)

• The host described HR 3447, the Chip Security Act, as requiring location-verification mechanisms for covered chips and reporting when hardware is diverted. Alex Finn argued export controls will be circumvented when demand is strong, while the host predicted the bill would advance after the Nvidia allegations.

Adoption (1)

• Alex Finn said his parents became enthusiastic about AI after ChatGPT helped them repair a broken door, and argued that practical AI education would build lasting support for data centers better than giveaways. The discussion also cited a Geely model priced at $9,000 that had displaced Tesla's Model Y as China's top-selling car.

Startups (2)

• Sarah Hooker said Adaption Labs automates fine-tuning and continuously updates models using a company's context, aiming to make smaller models useful without the manual tuning process she said many firms found difficult. Manu Sharma described Labelbox's Recursion as an enterprise agent platform for recurring technical work such as underwriting, research, and model training.
• Alex Finn described Henry Intelligent Machines as an agent that learns a user's skills and helps build businesses around problems that user can solve. Finn said the product was preparing a waitlist and envisioned agents creating value while their users pursue other work and interests.

Elections (1)

• The guest says the Democratic midterm gains will make it harder for Trump to advance his agenda through Congress, while leaving substantial executive power intact. The guest also describes Democratic-socialist influence in major cities and a Democratic Texas governorship as potentially consequential shifts.

Markets (2)

• The guest says crowded AI positions amplified the post-election market selloff, with hedge funds cutting teams and strategies. Marty Bent cites a record single-day volume of put options, suggesting investors bought protection after volatility had already hit.
• The guest warns that AI investment and Treasury-market health are increasingly linked: leveraged hedge funds buy long-dated government debt, but may pull back when volatility forces them to reduce positions. The weak Treasury bid during the equity selloff raised concern about that connection.

Macro (2)

• Marty Bent cites stronger-than-expected payrolls alongside falling oil prices, while equity markets recovered after the sharp selloff. The guest says the market's limited Treasury bid suggests investors may see AI-driven growth as important to tax receipts and the government's debt burden.
• The guest argues that blocking data-center construction could weaken growth, tax receipts, and Treasury demand, pushing the United States closer to debt stress and renewed monetary debasement. That chain of risks makes a prolonged buildout stall a potential tailwind for hard assets.

AI Infrastructure (1)

• AI-company valuations and fundraising expectations have come down even as demand for compute and model capabilities keep rising. The guest sees continuing economic utility in AI, while Marty Bent points to rising GPU-hour prices as evidence that compute demand remains strong.

Chips (1)

• The guest treats the timing of reports about Jensen Huang's personal life as potentially strategic, not as established evidence of a plot. The guest speculates that a leak could increase pressure on Huang to align NVIDIA with U.S. national-security priorities after his opposition to tighter chip export controls.

Iran (1)

• Marty Bent says J.D. Vance attributed the midterm shift toward Democrats partly to voter anger over the Iran war, its perceived connection to Israel, and higher energy costs. The guest cautions that voters may oppose losing wars more than war itself, and says Trump may have incentives to resist a quick withdrawal.

China (1)

• The guest says geopolitical competition with China is unlikely to ease after the U.S. midterms: Chinese producers keep exporting chemicals despite losses, while automakers in the U.S. and Germany seek more protection from Chinese electric vehicles. A proposed U.S. vehicle ban had more than 100 House co-sponsors.

Diplomacy (1)

• Marty Bent says China and Oman called for safe navigation through the Strait of Hormuz, while the guest notes that the conflict's outcome remains uncertain. The guest also flags a Polish dispute over whether central-bank gold should be placed under closer European Central Bank custody, and U.S. sanctions on Canadian businesses.

BTC Markets (3)

• Bitcoin closed below its 50-week moving average after seven consecutive weekly closes above it. Alex Thorn's comparison with earlier post-bottom recoveries in 2011 and 2019 found that Bitcoin did not make a lower low after losing the average, though the guest says that pattern may not repeat.
• The guest says AI investment has competed with Bitcoin for marginal capital, as investors favored high-growth technology stocks and expected AI-driven growth to ease debt concerns. If that growth disappoints, the guest argues, reduced competition for capital and renewed concern about sovereign debt could improve Bitcoin's appeal.
• Marty Bent argues that commoditized AI tokens could still support Bitcoin if inference margins compress and investment shifts away from the labs. The guest adds that rapid automation could disrupt workers with mortgages and student debt, potentially increasing pressure for a new debasement policy or universal income.

Elections (5)

• Democrats swept both chambers and won at least 29 governor races, while Republicans lost ground across voter groups and several deeply red states, including Texas, Kansas, Iowa, and South Carolina. The Senate and House counts were still unsettled.
• David Sacks blamed the Republican defeat on the Iran war and its effect on household costs: fuel and transport pushed inflation higher, erased the prospect of rate cuts, and made affordability the dominant voter concern. He argued that ending the war could reverse those pressures.
• Chamath Palihapitiya read the result as a rejection of an economy that rewards asset owners while wage earners struggle, rather than a clear ideological turn. He cited failed spending and redistricting strategies, and Gallup's finding that Americans were far less optimistic about AI than Chinese respondents.
• Jason Calacanis said Trump's approval fell before the Iran war amid tariffs, ICE enforcement, and the Epstein-files dispute; David Sacks countered that the Senate forecast barely moved after the tariff rollout and shifted sharply after the war began. Their disagreement centered on whether the war or the broader record drove the collapse.
• David Sacks argued that presidents can recover from midterm losses by changing course, citing Bill Clinton and Barack Obama; he warned that doubling down on the Iran war could repeat George W. Bush's political trajectory after Iraq. Chamath Palihapitiya said Trump should preserve border controls and seek economic concessions.

Macro (1)

• David Friedberg argued that the end of DOGE and continued federal spending left the cost-of-living problem untouched. He linked the post-COVID rise in spending to higher prices and argued that the 2025 budget did not materially change the government's spending trajectory.

Regulation (2)

• Jason Calacanis proposed a deal-focused response: tax large capital-gains-backed loans, raise the minimum wage, bar lawmakers from stock trading and foreign business, and encourage homebuilding. Chamath Palihapitiya favored eliminating federal income taxes below roughly $125,000 to $150,000.
• David Friedberg warned that states may restrict AI development, deployment, and use, while David Sacks predicted federal rules requiring frontier models to undergo government review. Both raised concerns about state power over AI, though Chamath Palihapitiya argued that broad regulation could instead entrench large technology companies.

Markets (1)

• A roughly $4 trillion market-cap decline hit AI and technology stocks, including Nvidia, SpaceX, Tesla, AMD, and Micron. The selloff also lowered prediction-market odds for OpenAI going public, while reports suggested its potential valuation could be nearer $1 trillion than earlier estimates.

AI Infrastructure (1)

• Chamath Palihapitiya said slower data-center construction would favor hyperscalers with existing capacity and squeeze cloud renters and frontier-model companies. He highlighted SpaceX's 90-day termination option in its compute deal with Anthropic as leverage to reprice scarce capacity or redirect it to larger customers.

Agents (1)

• Jason Calacanis said his firm was spending about $50,000 a year on AI subscriptions and estimated roughly $2 million in value, with agents increasing investment and media-production capacity. He predicted more local-model use and offshore data-center construction if U.S. permitting slows.

Reasoning (1)

• David Friedberg said OpenAI's 175-page paper challenged the Exponential Time Hypothesis by presenting a subexponential solution for a class of 3-SAT problems. He stressed that outside verification was still underway, while arguing that the result keeps open the possibility that P equals NP.

Markets (1)

• The host expects Anthropic to proceed with its IPO: Polymarket odds looked clear when the roadshow began, though public reports say bankers are pressing for a lower valuation. A speaker compares the pressure to Meta’s IPO, which priced down sharply before its shares later rose many times over.

Models (4)

• Alex says frontier labs need consumer-facing benefits that make AI tangible, proposing that they use superintelligence to tackle the top 5,000 diseases. The host adds that a personal health tool or widely loved AI product could build public support faster than abstract benchmark gains.
• Alex expects frontier-model improvement to move from scheduled releases toward continuous updates around early 2027. He predicts AI systems will explain new architectures, while the host argues rapid releases will force companies to rethink how people absorb new capabilities.
• The speakers expect recursive self-improvement to redirect compute toward science, engineering, medicine, and products such as drugs. The host predicts many people’s first vivid AI encounter could be a cheap, personalized talking avatar; Alex also predicts AI could enable relativistic probes and force renewed attention to the Fermi paradox.
• A paper measuring AI economic value per joule tested professional tasks across occupations: GLM 5.2 led in average task value, while Gemma 4’s 31-billion-parameter version delivered more value per joule. Alex favors measuring economic output against economic input, since watts may not remain the right denominator.

Labor (1)

• Alex frames AI’s social bargain as a trade between capital and labor: superintelligence could bring disease treatment, universal basic benefits, and broader human uplift. He warns municipalities that banning data centers may forfeit leverage before compute shifts to space.

Energy (1)

• The host says power constraints were already pushing developers toward efficiency, and political backlash adds urgency. The host also urges European energy executives to build generation, arguing that Europe’s focus on gas supply and local politics obscures rising AI power demand.

Reasoning (2)

• An independent METR study found Anthropic’s AI development accelerating at roughly 1.9 times its historical rate, with researchers attributing about 80% of the acceleration to AI-assisted research. Alex says METR’s charts and internal compute allocation indicate recursive self-improvement is already well underway.
• Alex defines recursive self-improvement partly by whether new compute is assigned to AI researchers rather than human researchers; he says the balance may already have crossed 50%. The speakers cite AI-generated research ideas, large agent runs, and scientific problem-solving as evidence that AI can improve its own research process.

AI Infrastructure (1)

• The host says major buyers have reserved much of the country’s compute and chip capacity, leaving smaller firms unable to obtain cloud access. Speakers warn that compute access could become a basic requirement for business, while Alex says scarcity should raise prices and create incentives to build elsewhere.

Regulation (1)

• The White House’s NSTM-6 policy requires federally funded researchers to use approved DNA and RNA synthesis providers, with customer verification and record retention. Alex favors controls on physical synthesis over government rules on what people or AI systems may think.

Robotics (1)

• Alex expects similar physical-world rules for general-purpose robots, potentially requiring approved manufacturers and logs of commands. He argues that regulation should focus on actions and safety controls, rather than restricting access to compute or AI research.

Space (1)

• A New Zealand photographer combined long exposures to show thousands of satellite trails, which Alex calls evidence of an emerging “baby Dyson swarm.” He says orbital compute could reduce reliance on politically exposed terrestrial data centers, with sun-synchronous orbit offering room before expansion elsewhere in the solar system.

Politics (1)

• Alex sees political swings as cyclical and superintelligence as a continuing secular trend; he cites Ray Kurzweil’s law of accelerating returns to argue that wars and depressions have had little effect on compute’s long-term price. The host warns that institutions must adapt actively rather than let technological change unfold unchecked.

Elections (9)

• Michelle Goldberg credits Texas Democrats’ gains to a full slate of serious candidates, whose events, ads, and turnout efforts reinforced one another instead of relying on a single marquee campaign. Democrats won four of five redistricted House seats and made major gains in the Texas House.
• Michelle Goldberg says James Talarico paired a religious appeal to love and neighborliness with populist attacks on billionaires and corporate power. His campaign offered a different political tone without asking Democrats to stop fighting over economic inequality.
• Michelle Goldberg says Ossoff drew on Georgia’s civil-rights tradition and built support among Black voters and suburban women who had become politically active after 2016. His attacks on Mike Collins tied election denial and racial provocation to constituent costs and corruption.
• Jon Ossoff’s reserved, carefully written speeches stood out against politicians copying Donald Trump’s loose, constant-media style. Michelle Goldberg says Ossoff believes restraint can make voters more attentive when he speaks, and noted that he writes his own speeches.
• Michelle Goldberg says Democratic gains in the South made Republican seats newly contestable, raising the stakes for candidate recruitment and sustained party investment. She cited Scott Colom’s Mississippi race after his blocked federal judicial appointment, and questioned how Republicans can hold a coalition together across a changing electorate.
• Michelle Goldberg criticizes Democratic leaders for favoring Haley Stevens’s establishment backing over the visibility and energy of Abdul El-Sayed and Mallory McMorrow. Goldberg also saw a sharp drop in enthusiasm between Graham Platner’s campaign and Troy Jackson’s later Maine events.
• Abdul El-Sayed ran as a working-class populist, arguing that attacks on his name and Muslim faith distracted from the cost of gas, groceries, and medical care. Michelle Goldberg says his win showed that a candidate can face heavy identity-based attacks and still succeed with a focused economic message.
• Adam Hamilton combined a pastor’s appeal to shared values with populist criticism of Roger Marshall’s treatment of patients’ medical debt. Hamilton described himself as both liberal and conservative and called himself pro-life with a heavy heart, a culturally moderate style Goldberg compared with older Democratic campaigns.
• Michelle Goldberg says Democrats are reconsidering which voters they can win: some MAGA voters may respond to candidates who address economic hardship with respect and practical help, not only to appeals aimed at educated suburbanites. She contrasts that approach with the resistance strategy that followed Trump’s 2016 victory.

Education (1)

• Michelle Goldberg found anger among conservative Texas voters over Greg Abbott’s school vouchers, which residents said were draining public schools. One former superintendent who had voted Democratic only once since the 1990s joined Gina Hinojosa’s campaign over the issue.

Immigration (1)

• Michelle Goldberg says immigration enforcement drove a sharp Democratic swing in heavily Hispanic Texas border counties, after residents saw relatives and neighbors detained and heard accounts of citizens being roughed up by ICE. Bobby Pulido won a Trump district by 25 points.

Corruption (2)

• Jon Ossoff connected Donald Trump’s corruption to household costs, arguing that public money and policy were benefiting Trump-aligned donors while constituents paid more for health care, gas, and groceries. Michelle Goldberg says Ossoff made complex allegations, including a Kazakhstan tungsten deal, into clear political stories.
• Michelle Goldberg sees the emerging Democratic coalition as focused on oligarchy, corporate power, and corruption, with investigations into the diversion of public goods as common ground between accountability and governing. She warns that holding a broad coalition together in office will be harder than winning with one.

Health (1)

• Michelle Goldberg says Medicare for All, once treated as an electoral liability, drew little effective attack in Michigan as health-care costs shaped voters’ concerns. She expects health reform to return to Democratic debate and noted that several doctors are entering the Senate caucus.

History (1)

• Michelle Goldberg recommended three novels: Patmina Sabbat’s Good People, about an Afghan American family and assimilation; Mario Vargas Llosa’s The Feast of the Goat, about Dominican dictator Rafael Trujillo; and Leonardo Padura’s The Man Who Loved Dogs, about Leon Trotsky’s assassination.