All-In Podcast on Venture Capital, Ranked

All-In Podcast venture capital episodes, ranked and summarized: the $4T AI IPO wave, secondaries at 31% of venture activity, SpaceX at $1.75T, and what Gerstner, Chamath, and Sarah Friar say about liquidity.

Start with Thomas Laffont's Coatue presentation from the All-In Summit. It is the clearest single explanation on the show of what has happened to venture capital: capital concentrating into fewer and much larger companies, and the finding that once a business clears $100 billion in value it still carries a 31% chance of compounding another 10x. From there, the two All-In Liquidity panels cover the plumbing, which is how founders, funds, and employees actually get paid in 2026.

These eight episodes come from our summary library, which covers the show's recent run across roughly the past year. Treat the list as the current All-In conversation about venture capital and public listings, with every figure traceable to a specific episode and guest.

One structural pattern runs through all eight: the exit has moved. IPOs are back as a headline event, secondary markets now absorb 31% of all venture activity, and the largest private rounds (OpenAI's $120 billion-plus) have grown bigger than most listings. The themes section below pulls those threads together.

1. Thomas Laffont: The $4T AI IPO Wave Is Coming

All-In Podcast · with Thomas Laffont of Coatue Management · 32m · June 2026

The only episode on this list built from a full data set. Laffont brought Coatue's unicorn-economy analysis to the All-In Summit and used it to re-price the entire venture asset class: fewer companies are raising, the ones that do raise far more, and a business past $100 billion in value carries a 31% probability of another 10x. That one statistic inverts the standard seed-stage power-law pitch, and it explains why crossover funds keep writing larger checks at entry multiples that look indefensible in isolation.

Key takeaways

  • Coatue's data shows the unicorn economy consolidating: fewer companies raise capital, and the ones that do raise significantly larger rounds.
  • Companies above $100 billion in value (centacorns) show a 31% chance of another 10x, a higher hit rate than earlier venture stages.
  • Laffont's SpaceX analysis found launch cadence is the variable most correlated with valuation, so operating throughput drives the multiple.
  • As major liquidity events approach, the ecosystem rebalances, and the cost of missing a winner rises in step with how fast winners compound.
The winners are compounding faster than ever, which means the cost of not being in a winner are higher than ever. — Thomas Laffont, Coatue Management

Read the full episode summary

2. Why Secondary Markets Are Eating the IPO: All-In Liquidity Secondary Markets Panel

All-In Podcast · with Brad Gerstner, Gavin Baker, and Kelly Rodriques · 39m · June 2026

The best 39 minutes on this page for understanding where venture liquidity actually comes from now. Gerstner, Gavin Baker, and Forge CEO Kelly Rodriques walk through a market that has flipped: private shares that cleared at roughly 20% discounts in recent years now trade at a 6% premium, and secondaries make up about 31% of all venture activity. Baker's line about a CEO going from the most special flower to one of thousands of public tickers is the most honest account anywhere of why strong companies stay private, and Gerstner's warning about retail buying in at pre-IPO peak marks is the risk that platform marketing tends to leave out.

Key takeaways

  • Secondary volumes hit records and now represent about 31% of all venture activity, competing with the IPO as a primary exit.
  • Pricing flipped from roughly 20% discounts to a 6% premium, which signals real demand for late-stage private exposure.
  • Position sizing matters more than conviction: the panel's advice is to size for staying power through inevitable drawdowns.
  • Entry timing drives the return, so accessing a company early beats buying at peak private marks right before a listing.
  • Gerstner flags the political stakes: trillions in private value created while most of the public feels locked out is destabilizing.
When you're the CEO of a private company, you are the most special flower to all of your investors. Once you're public, you're one of thousands of companies. — Gavin Baker

Read the full episode summary

3. The IPO Comeback: Why Tech Giants Are Finally Going Public

All-In Podcast · with Brad Gerstner, Andrew Feldman, and Will Marshall · 32m · June 2026

Founder-side truth about going public, from two CEOs who had just done it. Andrew Feldman of Cerebras and Will Marshall of Planet Labs describe the IPO as a funding event with new reporting obligations attached, and Feldman makes the case that most of the money in a great company gets made after the listing, in percentage and in absolute terms. His answer on market timing (they got it wrong for a decade before getting it right) is the most useful thing a late-stage founder can hear about windows.

Key takeaways

  • Feldman frames the IPO as a starting gun: the morning after, the product line and the engineering roadmap sit exactly where they were.
  • He cites studies showing more money is made post-IPO than pre-IPO, in both percentage and absolute terms.
  • Good timing is usually the output of a decade of being early, so durability matters more than the window itself.
  • Marshall's thesis: today's models absorbed the text of the internet and remain blind to the physical world, which makes real-world data the next index.
  • Latency is the product in AI inference, which is why Cerebras built around real-time delivery.
I think historically more money's made after IPO than before. I think every single study shows that there is more money to be made both in percentage and in what we care about which is absolute. — Andrew Feldman, CEO of Cerebras

Read the full episode summary

4. OpenAI CFO Sarah Friar on IPO, AI Rivalries, and Spending $100B+ on Compute

All-In Podcast · with Sarah Friar, CFO of OpenAI · 32m · June 2026

The clearest window on this page into how the largest private financing in history was actually constructed. Friar walks through a round north of $120 billion and the Rubik's cube logic behind it: multiple cloud providers, multiple chip partners, multiple product surfaces, all engineered so a single dependency can stall nothing. Her framing of an IPO as a milestone inside a funding sequence is the most grounded statement in this whole set, and it quietly reframes the trillion-dollar listing talk that fills the other episodes.

Key takeaways

  • OpenAI raised north of $120 billion, and Friar treats a future IPO as one more fundraising mechanism along that path.
  • The Rubik's cube strategy spreads dependency across cloud providers, chip partners, and products to manufacture optionality.
  • Her investor lesson: the market weighs results, and being first to market confers no durable advantage.
  • Staying close to the value-creation layer is where margin concentrates, which shapes where OpenAI chooses to build.
  • Product direction comes from what users actually ask for, since top-down mandates get politely ignored.
In the end, an IPO, I say this to the team all the time, it's a milestone. It is not a destination. Do not run your company as if that's some sort of destination. It's just another way to fund raise. — Sarah Friar, CFO of OpenAI

Read the full episode summary

5. SpaceX IPO, Iran War Fallout, Quantum Bitcoin Hack, The Space Opportunity

All-In Podcast · Chamath, Jason, Sacks, and Friedberg · 1h 20m · April 2026

The episode that put a number on the biggest venture outcome ever written: SpaceX filing at $1.75 trillion, which would rank it the eighth-largest company on earth. The besties then do what they are genuinely best at, reasoning through second-order effects: a Tesla merger that creates a company worth more than $3 trillion, a queue of OpenAI, Anthropic, and Databricks listings behind it, and Chamath's sequencing argument that in a crowded IPO window being consumed first beats being priced perfectly.

Key takeaways

  • SpaceX's $1.75 trillion filing would make it the eighth-largest company globally, setting the benchmark for this IPO wave.
  • Chamath expects SpaceX and Tesla to merge, producing a company valued above $3 trillion.
  • Sequencing matters in a crowded window: the earliest large listings absorb the available demand.
  • Friedberg's space thesis: robotic lunar mining plus mass drivers could move processed material to Earth more cheaply than any terrestrial shipping method.
  • Chamath's AGI fork: if AGI arrives, incumbent durability collapses; if it stalls, hundred-billion-dollar raises deserve hard scrutiny.
If AGI is real, the durability of most companies is slim to none. If AGI is not real, then the fundraising capacity of these companies that are now raising hundreds of billions of dollars needs to get questioned and inspected thoroughly. — Chamath Palihapitiya

Read the full episode summary

6. More Trillion Dollar IPOs, Anthropic $3T, Zuck's Price War, China Ends Open Source?

All-In Podcast · Chamath, Jason, Sacks, and Friedberg · 1h 42m · July 2026

The most important disagreement on this page, and it lands on the number every limited partner will eventually demand. Chamath reports his own CTO measuring token costs doubling every 45 days against productivity gains of 5% at most, with realized ROI closer to zero to 2%. Brad Gerstner counters that intelligence is the largest addressable market ever measured, and that frontier revenue moving from $100 billion to $300 billion swamps whatever any single sophisticated buyer optimizes. Which side is right determines whether the trillion-dollar listings queued behind SpaceX hold their marks.

Key takeaways

  • Chamath's internal data: token spend doubling every 45 days while measured downstream productivity stayed under 5%.
  • Realized enterprise AI ROI in his portfolio landed between zero and 2%, which sets up a CFO reckoning over EPS lift.
  • Gerstner's counter: incremental frontier revenue moving from $100 billion to $300 billion is unprecedented in Silicon Valley history.
  • The operator fix is strict ROI tracking plus forward-deployed engineers redesigning processes with department heads.
  • Token prices fall roughly 95% over time, which changes the arithmetic once agents finish hourly tasks overnight.
The actual ROI was somewhere between zero and 2%. So I don't know. I mean I think that enterprise looks really good. The problem is that very smart investors like Brad and Gavin and others at some point will start asking companies what's your ROI? What's the actual EPS lift? — Chamath Palihapitiya

Read the full episode summary

7. Anthropic IPO at Risk, Meta's Muse Pop, Token Prices Fall, Open Source Gains Share

All-In Podcast · Chamath, Jason, Sacks, and Friedberg · 1h 34m · September 2026

The valuation-risk episode, and the bear case stated by people who are long the sector. The hosts take apart whether a frontier lab can hold a premium multiple while capable low-cost models proliferate and token prices keep falling, which is the question sitting under every AI listing on this list. Chamath's demand that these businesses be called companies and carry product liability like companies is a governance argument with direct pricing consequences, and Sacks reframes a safety decision as a shipping decision a board can own.

Key takeaways

  • Anthropic's IPO path carries real valuation risk as open-source models close the capability gap and token prices fall.
  • Chamath's governance point: frontier labs are companies and belong under company standards, including product liability.
  • Sacks reframes release decisions as shipping decisions, which keeps accountability with whoever puts the product out.
  • Value is migrating from benchmark scores toward reliable consumer agents (Meta's Muse, Grokbot) that finish real work.
  • The efficiency wave compresses model pricing on a timescale measured in days, which pressures premium valuations.
We have to stop calling these companies labs, they're companies. And we have to be judged on the same standards, including product liability. — Chamath Palihapitiya

Read the full episode summary

8. Anthropic's $2T IPO, Zuck's AI Manifesto, Nvidia's $500B AI Bet, Grok's Comeback

All-In Podcast · Chamath, Jason, Sacks, and Friedberg · 1h 39m · August 2026

The capital-formation episode: where the money behind a $2 trillion listing physically comes from. The panel digs into Nvidia's effort to turn GPU infrastructure into a financeable asset class, which would open a new pool of capital for AI capex. Gavin Baker supplies the constraint that separates AI build-outs from software build-outs, which is that power, land, and data centers are atoms with real lead times. His preference for distributed capacity is a direct challenge to the single-winner thesis priced into the largest private rounds.

Key takeaways

  • Anthropic IPO reports arrive alongside AI revenue growth that outruns available power, data center, and GPU supply.
  • Nvidia is working to make GPU infrastructure a financeable asset class, which would unlock new capital for capex.
  • Baker's constraint: AI expansion is physical, so lead times on atoms govern the pace more than software cycles do.
  • He argues for distributing and decentralizing capacity whenever a choice exists, which widens the set of viable winners.
  • xAI and Grok's improving position shows the frontier stays contested, and the total market expands as entrants compound.
The pie becomes gigantic. — Gavin Baker

Read the full episode summary

What these episodes have in common

Theme 1: The exit moved, and venture math moved with it

Three episodes describe the same structural shift from three different seats. The All-In Liquidity secondary markets panel documents the mechanics: private shares now carry a 6% premium after years of roughly 20% discounts, and secondaries absorb about 31% of all venture activity. Thomas Laffont's Coatue presentation supplies the reason capital tolerates those prices, since a company past $100 billion in value still shows a 31% chance of another 10x. Sarah Friar closes the loop from the issuer's side, describing a round north of $120 billion and calling an eventual IPO one funding mechanism among several.

Put together, the three reframe what a venture outcome looks like. Ownership changes hands for years before any listing, the largest private rounds now exceed most IPOs in size, and the public offering functions as a liquidity event for an already-mature cap table. Gavin Baker's most-special-flower line names the behavioral cost of that arrangement, because private markets flatter founders, and Gerstner's reminder that investing is the search for truth is the discipline that has to substitute for public-market scrutiny.

Theme 2: The besties agree on the wave and split on the earnings

Every episode here accepts the premise that trillion-dollar listings are coming. SpaceX files at $1.75 trillion, Anthropic gets discussed at $2 trillion in August and $3 trillion in July, and Laffont sizes the full AI IPO wave at $4 trillion. The argument starts at the income statement. In More Trillion Dollar IPOs, Chamath puts numbers on his own company: token costs doubling every 45 days, downstream productivity gains under 5%, realized ROI between zero and 2%. Brad Gerstner's reply is that intelligence is the largest addressable market ever measured, and that $100 billion of frontier revenue becoming $300 billion dwarfs whatever an individual buyer optimizes away.

Anthropic IPO at Risk is where that tension gets priced. If capable open-source models keep compressing token prices, premium multiples on frontier labs compress alongside them. Chamath's answer is governance: call them companies, hold them to product liability, and let the public market do what he calls the great antiseptic. The practical through-line for anyone underwriting these deals is to underwrite the earnings claim specifically, because the wave itself is already consensus.

Theme 3: Going public is a financing event, and the operators say so

The two founder-side episodes are unusually consistent, and both push back on treating a listing as an achievement. Andrew Feldman of Cerebras says that the morning after the IPO you have sold no more product and shipped no more engineering, and he cites the studies showing most of the money in a good company gets made after the offering. Sarah Friar tells her own team the same thing in CFO language, that an IPO is a milestone inside a funding sequence. Will Marshall of Planet Labs spends his segment on the underlying business, which is indexing the physical world for models that only ever absorbed text.

The practical read for founders and employees is that time horizon beats timing. Feldman's answer about getting the window right was that they got it wrong for a decade first. Laffont's SpaceX finding points the same direction, since the variable his data correlates with valuation is launch cadence, meaning operating throughput. Build the cadence and the window arrives on its own.

Every episode referenced

Frequently Asked Questions

Which All-In Podcast episodes cover venture capital?

Eight episodes in our library cover venture capital and public listings directly. The strongest is Thomas Laffont's Coatue presentation on the $4T AI IPO wave, which lays out the unicorn economy with real data. Follow it with the two All-In Liquidity panels (the IPO comeback with Andrew Feldman and Will Marshall, and the secondary markets panel with Gavin Baker and Forge CEO Kelly Rodriques), then OpenAI CFO Sarah Friar on the company's $120 billion-plus round.

What is the All-In Liquidity panel?

All-In Liquidity is the show's event series on how private companies and their shareholders get paid. Two panels appear on this page: an IPO panel hosted by Brad Gerstner with Cerebras CEO Andrew Feldman and Planet Labs CEO Will Marshall discussing their recent listings, and a secondary markets panel with Gerstner, Gavin Baker, and Forge CEO Kelly Rodriques on private share trading. Both run a little over half an hour, which makes them the fastest route into current liquidity mechanics.

What does All-In say about secondary markets?

The secondary markets panel reports a market that has inverted. Private company shares that traded at roughly 20% discounts in recent years now clear at a 6% premium, and secondaries account for about 31% of all venture activity, competing with the IPO as a primary exit path. Gavin Baker and Brad Gerstner both flag the risk for retail investors entering at pre-IPO peak marks, and their guidance is to size positions for staying power through drawdowns.

What is All-In's view on the IPO comeback?

The hosts treat the window as genuinely open, anchored by SpaceX's $1.75 trillion filing and by Anthropic discussions ranging from $2 trillion to $3 trillion, with OpenAI and Databricks queued behind them. Thomas Laffont sizes the full wave at $4 trillion. Chamath Palihapitiya adds the sequencing point that in a crowded window the earliest large listings absorb available demand, and Sarah Friar supplies the issuer's framing of a listing as one more way to raise capital.

What venture capital advice do the All-In hosts and guests give founders?

Three pieces of advice recur across these episodes. Andrew Feldman tells founders the IPO changes the bank balance and leaves the product and engineering work exactly where it was, and that getting market timing right usually follows a decade of getting it wrong. Sarah Friar builds optionality by diversifying cloud providers, chip partners, and products so a single dependency can stall nothing. Chamath Palihapitiya's version is to measure ROI on AI spend before the public market measures it for you.

Keep exploring