All-In PodcastTrump Takes On the Fed, US-Intel Deal, Why Bankruptcies Are Up, OpenAI's Longevity Breakthrough
CHAPTERS
- 0:00 – 9:00
Cold Open, Moose Visit, Bar Mitzvahs, and All-In Summit Promotions
The episode opens with Jason’s dog ‘the moose’ crashing the recording, followed by friendly banter among the besties about ad reads, sponsors, and the upcoming All-In Summit in Los Angeles. They reminisce about their bar mitzvahs and childhood birthday parties, mixing self-deprecating humor with a bit of generational commentary on Gen Z attitudes toward ethics and ‘square’ behavior.
- •Jason’s dog appears on camera, lightening the tone before the main topics.
- •Friedberg fumbles an ad read, leading to playful mockery and a rundown of Summit sponsors (Oracle, Circle, BVNK, Iren).
- •Discussion of Summit activations: bars, networking lounges, AI infrastructure dinners, a stablecoin dinner, and an arcade showdown idea.
- •Besties share bar mitzvah stories (rented arcade machines, low-budget Brooklyn parties, Reagan/Nixon jokes) as character color.
- •Chamath notes that younger generations react strongly against perceived unfairness (e.g., tipping maître d’s for line-skipping), tying into broader cultural shifts.
- 9:00 – 20:35
Travel, On Running Shoes, and Founder University Expands to Riyadh
Chamath briefly recaps his European trip, including a rough stretch on the island of Pantelleria and a decompression stop in Milan, which leads to a mini-endorsement of On Running shoes. Jason then announces the expansion of his Founder University program to Riyadh, describing its pre-accelerator model and deal-flow logic.
- •Chamath describes visiting Pantelleria and Milan; his partner fell ill, limiting vacation time.
- •He praises On Running’s comfort, says he’s ‘ditching all my Nikes,’ and half-jokes about helping the brand via Roger Federer.
- •Jason announces Founder University’s first international expansion to Riyadh in partnership with Sanabel, targeting MENA founders.
- •The program filters thousands of applicants down to ~50 participants and invests in roughly 10% after a 12-week period.
- •They position Founder University as both ecosystem-building and high-quality deal flow for Jason’s fund.
- 20:35 – 33:40
Trump vs. Fed Governor Lisa Cook and the Myth of Fed Independence
The group dives into Trump’s unprecedented attempt to fire Fed Governor Lisa Cook ‘for cause’ over an alleged mortgage-fraud issue predating her appointment. This sparks a wider debate over whether the Fed is truly independent or inherently political, whether presidents should be able to remove governors at will, and what the Fed should even do in a 2025+ economy.
- •Context: Lisa Cook, a Biden-appointed Fed governor, is accused of listing two primary residences; FHFA refers the case to DOJ, Trump demands her resignation and then announces her firing.
- •Chamath argues all major federal appointments—including Fed and Supreme Court—are inherently political, and we should stop pretending otherwise.
- •He proposes a narrower Fed: keep supervision/regulation and payments/clearing; move lender-of-last-resort and principal rate-setting to Treasury plus market mechanisms.
- •Chamath highlights Commerce’s move to put GDP data on-chain as a step toward real-time economic data feeds and blockchain-based pricing oracles.
- •Friedberg defends the necessity of Fed independence and 14-year terms to avoid short-term political manipulation of rates that could harm long-term debt sustainability.
- •He suggests revisiting term lengths and clarifying the executive’s removal power for governors rather than back-dooring it through selective investigations.
- •Jason underscores that FOMC voting is more balanced than narratives suggest (Trump and Biden appointees both on board, some documented dissents).
- 33:40 – 45:50
Powell, ‘Transitory’ Inflation, and Politicized Rate Cycles
Sachs lays out a detailed case that Jerome Powell’s decisions around inflation and rate moves were politically timed—to get renominated and to aid Democrats—while Jason pushes back using FOMC voting structure and inflation data. They spar over the ‘transitory’ period, QE continuation, and the 2024 pre-election 50-basis-point cut.
- •Sachs claims Powell went along with Biden/Yellen’s ‘transitory’ narrative after a 5% inflation print in 2021 to secure renomination, delaying necessary tightening and exacerbating asset bubbles and later inflation.
- •He points out Powell dropped the word ‘transitory’ just a week after being renominated, then launched a historically rapid tightening cycle.
- •Sachs emphasizes Elizabeth Warren’s 2024 letter demanding aggressive cuts when inflation was ~2.5%, and calls out her later demand that Powell resist Trump as hypocritical.
- •Jason counters: Powell is only one of 11–12 FOMC votes, there have been multiple dissents, and the Fed’s pivot timing can be partly explained by shifting inflation data and uncertainty around Trump’s tariff impacts.
- •Chamath reiterates that markets, via instruments like SOFR and Treasury auctions, already price risk and rates far more dynamically than a monthly committee.
- •Sachs accuses Powell of cutting 50 bps pre-election to favor Kamala Harris and then pausing the cutting cycle once Trump won, dubbing him ‘Too Late Powell.’
- 45:50 – 55:30
What Should the Fed Actually Do in a Modern Economy?
The conversation moves from Powell’s record to first principles about central banking. Chamath outlines a reduced-scope Fed whose core monetary functions are supplanted by highly data-driven markets, while Friedberg maintains that expert judgment and independence are critical for balancing short- and long-term tradeoffs.
- •Chamath lists four Fed functions: lender of last resort, monetary policy/price stability, banking supervision, and payments/clearing.
- •He argues Treasury and markets already do a better job on liquidity and rate-setting, while the Fed is still valuable for supervision and payments.
- •He envisions a future where macro metrics (GDP, employment, etc.) are published on-chain and banks use oracles to set rates in real time; Fed funds becomes less relevant compared to market benchmarks like SOFR.
- •Friedberg insists macro tradeoffs—short-term demand vs. long-term inflation and debt-service risk—require a stable, independent board of economists, not purely market dynamics.
- •They briefly touch on RFK and COVID/vaccine policy at the end of the episode, but defer detailed debate to a future show pending more data.
- 55:30 – 1:08:40
U.S. Buys 10% of Intel: Industrial Policy, China, and Sovereign Wealth Funds
Attention shifts to the Trump administration’s move to convert CHIPS Act grants into a 10% non-voting equity stake in Intel, following public criticism of Intel’s CEO. The besties compare this model to China’s state-directed investment and golden shares, debate when equity-for-subsidy is appropriate, and explore how such stakes should be managed for the public good.
- •Intel had been allocated ~$9B in CHIPS Act grants and additional loans to onshore advanced chip manufacturing; Trump and Lutnick renegotiated the support as non-voting equity.
- •Chamath contrasts this with China’s model of state-backed ‘national champions’ in semis, rare earths, and APIs, where the government often takes golden shares and uses price dumping and capital-market intimidation.
- •He criticizes prior U.S. crises (TARP, 2008 bailouts, Buffett–Goldman) where taxpayers took massive downside risk but captured little to no equity upside.
- •All three largely endorse the Intel structure as ‘much better’ than free grants: transparent, no golden vote, market-finance compatible, and taxpayer-upside-friendly.
- •Sachs and Friedberg stress criteria: only in clear national security contexts and market failures, not routine subsidization; equity should impose a real cost on firms seeking state support.
- •Jason praises the substance but criticizes the optics and ‘bullying’ style of publicly attacking and ‘threatening’ the CEO before striking the deal.
- 1:08:40 – 1:23:25
Should America Build a Sovereign Wealth Fund—and Tie It to Social Security?
The Intel discussion segues into a broader exploration of whether the U.S. should operate a sovereign wealth fund. Chamath suggests seeding it with tariff-linked inbound capital and strategic equity stakes, while Friedberg argues it should be integrated with the Social Security trust fund to address looming insolvency. Jason warns that any new asset pool risks being raided by Congress.
- •Chamath: Trump’s tariff deals purportedly bring >$1T of committed investment from allies (Japan, Korea, Europe), often on 90/10 profit splits; he wants that capital and equity from deals like Intel to seed a national fund.
- •Friedberg outlines three options for holding equity: (1) on the general federal balance sheet (too easy to spend), (2) a new standalone sovereign wealth fund, or (3) within the OASI Social Security trust fund.
- •He prefers option 3, arguing that Social Security’s projected insolvency around the early 2030s could be mitigated by holding diversified public and private equity rather than only special-issue Treasuries.
- •Sachs is wary of ‘using’ Social Security contributions as VC capital, but agrees equity from bailouts/strategic investments should be pooled and could help offset Social Security obligations.
- •All three worry Congress will treat any new income stream or marked-up asset value as license to increase spending (citing California’s boom-time budget blowout as an example).
- •Jason brings up Grover Norquist’s Colorado model, where state spending is constrained by population plus inflation and surpluses are returned—suggesting a similar cap could be the core plank of an ‘America Party’ focused solely on fiscal restraint.
- 1:23:25 – 1:32:30
Corporate Bankruptcies, Creative Destruction, and the End of ZIRP
The panel examines S&P data showing 2025 on pace for the highest number of large corporate bankruptcies since 2010. They reject the idea that recent tariffs are the primary driver and instead trace failures back to a decade-plus of artificially cheap money, over-levered buyouts, and suppressed M&A that delayed necessary restructuring.
- •S&P defines ‘large bankruptcies’ as public firms with ≥$2M in debt and private firms with ≥$10M in assets or liabilities; 2025 is pacing ahead of all years since 2010.
- •Chamath calls media efforts to blame this on Trump’s recent tariffs ‘nonsense’ given the timeline; large firms don’t fail in 30–60 days from such shocks.
- •He argues that ZIRP allowed structurally weak businesses—especially retail chains and PE-owned companies overloaded with debt—to survive far longer than fundamentals warranted.
- •M&A and consolidation were heavily constrained by antitrust/regulatory regimes, limiting ‘creative destruction’; as those constraints ease, more bankruptcies and restructurings are a healthy recalibration.
- •Friedberg underscores how brick-and-mortar retail is effectively leveraged through long-term leases and has been structurally disrupted by e-commerce and on-demand delivery.
- •They briefly reference new competition from unexpected places (e.g., Travis Kalanick’s CloudKitchens launching a Chipotle competitor) as further pressure on legacy formats.
- 1:32:30 – 1:57:30
CRE Refinancing Stress, Zombie Buildings, and the Case for Faster Cuts
Sachs zooms in on commercial real estate as a key locus of rate-sensitive distress. He explains how higher refinancing rates and lower valuations are wiping out equity, discouraging investment in tenant improvements, and keeping buildings half-empty. He blames Powell’s slow reaction for prolonging the pain.
- •Roughly $2.2T of commercial real estate debt must be refinanced by 2028.
- •Higher refinance rates can turn previously cashflow-positive buildings into cashflow-negative, effectively bankrupt assets.
- •Lower valuations reduce allowable LTV and therefore refinance proceeds; sponsors often need to plug multi–million-dollar gaps with fresh equity, which many cannot or will not do.
- •This leads to ‘pretend and extend’ deals morphing into actual foreclosures or voluntary surrenders of properties back to lenders.
- •A glut of vacant, underinvested office buildings (e.g., ~one-third of SF office vacant) cannot be re-leased without new capital for tenant improvements, which underwater owners have no incentive to provide.
- •Sachs argues that meaningful, timely Fed cuts are needed to ease this rollover burden; until then, CRE remains a risk pocket beneath otherwise-strong GDP prints.
- 1:57:30 – 2:11:00
OpenAI’s GPT-4B Micro and the AI-Designed Rejuvenation Proteins
Friedberg leads a deep dive into OpenAI’s collaboration with Retro Biosciences using a specialized model, GPT-4B Micro, to design more effective variants of Yamanaka factors for cellular rejuvenation. The results show massive gains in conversion efficiency, illustrating both how AI can accelerate drug/protein discovery and how close we might be to impactful anti-aging therapies.
- •Yamanaka factors (O, S, K, M) are four proteins that can reprogram somatic cells into induced pluripotent stem cells; partial, transient application can ‘rejuvenate’ cells without fully de-differentiating them.
- •The O protein alone is 360 amino acids long; with 20 amino acids, the combinatorial design space is 20^360—astronomically large and far beyond brute-force or classic modeling.
- •OpenAI trained GPT-4B Micro on protein sequences, biological texts, and tokenized 3D structural descriptors, then tasked it with proposing modified O/S/K/M sequences predicted to be more efficient.
- •Retro Biosciences synthesized and tested those designs; new constructs boosted conversion from <0.1% of cells to >30% at day 7 and ~85% by day 12 based on stem-cell marker expression.
- •Friedberg estimates 7–12 years to first approved human drugs based on these mechanisms, beginning with specific disease indications before ‘aging’ itself can become an indication.
- •He warns that overdosing or full reprogramming creates uncontrolled stem-cell proliferation that looks like cancer, so careful dosing, targeting, and control will be critical.
- •The experiment also showcases the power of small, fine-tuned models in biology: you don’t always need a giant general model—domain-specific LLMs can uncover emergent structure-function patterns from sequence data alone.
- 2:11:00 – 2:22:30
Speculating on Clinical Timelines, Offshore Clinics, and Vaccine Politics
The besties probe the practical implications of rejuvenation tech—how soon it reaches humans, whether offshore clinics will move faster, and what safety issues remain. They briefly touch on RFK’s evolving vaccine policies and the social psychology around revisiting COVID-era decisions.
- •Friedberg reiterates timing: current trials are Phase I safety/dosing; indications-specific approvals are likely a decade out, with aging as a formal indication later.
- •Chamath muses about people flying to permissive jurisdictions (e.g., Costa Rica) in 3–4 years to access early-stage therapies outside FDA channels; Friedberg says this is plausible, especially given current stem-cell medical tourism.
- •Key safety risk: pushing cells too far into a stem-cell state can create tumor-like growth; mitigating that is a central research problem.
- •They mention RFK’s moves to limit federal funding for mRNA vaccines and change CDC guidance on COVID shots for healthy kids/pregnant women, but postpone deep analysis until they can review his autism-related evidence.
- •Jason and Chamath reflect on cognitive dissonance and dogmatism around vaccines—how parents resist even questioning past decisions about their children’s health.
- 2:22:30
Closing Banter: Fed ‘High Priests,’ Clubs, and the All-In Club Idea
The episode winds down with jokes about the Fed as robe-wearing high priests, membership clubs, and the idea of an ‘All-In Club’ for backgammon, poker, and cigars. They tease surprise speakers for the upcoming All-In Summit and sign off with their usual self-aware bravado.
- •Sachs lampoons the Fed as cloistered ‘high priests’ reading economic scrolls, while Jason reiterates his skepticism of large government agencies post-COVID.
- •They riff on exclusive clubs (Shadow Creek, Zero Bond, Link’s Club), including wine-cellar pricing quirks and membership hacks.
- •The idea of a multi-city ‘All-In Club’ with poker and backgammon tables is floated as a half-serious aspiration.
- •They promote the All-In Summit, hint at last-minute big-name speaker additions, and urge listeners to subscribe and share with friends.
