Oct 30, 2025 · 1h 26m · news
OpenAI Restructuring: Who Wins and Who Loses & Mercor Raises $350M at a $10BN Valuation · 20VC with Harry Stebbings
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of 20VC, Harry Stebbings, Jason Lemkin, and Rory O'Driscoll dissect the current state of venture capital and tech, analyzing OpenAI's restructuring, the rise of multi-billion dollar mega-funds, and the economic realities behind massive late-stage valuations.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 14.7% of the talking time here. How this is scored →
Harry as informed peer 4.4 Guest teaching 4.5 Guest disagreement 3.6 Harry pushing back 2.4
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 21:44 Rory rejects podcast fame as the source of deal wins
Rory forcefully rejects Jason's suggestion that podcasting drives deal wins, exclaiming that winning isn't because he's 'on a fucking pod' instead of grafting for 30 years.
Hardest push from Harry ▶ 23:26 Harry challenges Jason on fund scale advantagesHarry directly counters Jason's argument that fund size doesn't matter, explaining how Andreessen's massive fee pool pays for talent, carry, impressive offices, and serendipity-generating events.
Biggest teaching moment ▶ 1:12:12 Rory explains the IRR vs multiple constraint formulaRory educates the room on portfolio management math, explaining that venture investing is not simple IRR maximization but rather maximizing overall fund multiple subject to a minimum IRR constraint.
Harry holds his own ▶ 29:43 Harry presents a 3-pillar breakdown of RLHF evolutionHarry demonstrates deep domain expertise on RLHF unit economics and customer concentration, earning explicit praise from Rory for his sharp market analysis.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Podcast Kickoff and Conversation Setup | 1 | 3 | 1 | 0 | Harry opens the conversation and prompts Rory to explain the breaking news around OpenAI's corporate restructuring. Rory provides a clear overview of the Microsoft deal and state Attorney General approval, while Jason adds commentary on Sam Altman owning zero equity. | |
| Winners and Losers of the OpenAI Restructuring | 2 | 3 | 1 | 1 | Harry asks who the winners and losers are, briefly noting that there seem to be no real losers. Rory outlines how Microsoft, the charitable foundation, and employees all won, while Jason notes the flexibility for LP ownership rules. | |
| OpenAI's Trajectory Towards a Historic Retail IPO | 3 | 4 | 1 | 1 | Jason and Rory explore OpenAI's potential for a massive retail IPO. Harry asks whether OpenAI can become a trillion-dollar company by 2026, prompting Rory to analyze revenue growth trajectories and valuation multiples. | |
| OpenAI's Web Browser Ambitions: Game-Changer or Hype? | 6 | 5 | 7 | 5 | A heated debate emerges over Andreessen Horowitz raising a $10B mega-fund. Rory passionately rejects Jason's idea that media presence outweighs track record, dropping an expletive, while Harry pushes back against Jason by enumerating concrete scale advantages like fees, carry, and events. | |
| Mercor's $350 Million Funding Round and the Realities of RLHF | 7 | 5 | 2 | 3 | Harry demonstrates strong market expertise by laying out a three-pillar framework for RLHF evolution and highlighting customer concentration risk at Mercor. Rory expands on the technical shifts from cat labeling to PhD-level math/physics RLHF. | |
| Ramp's $30 Billion Rumors and 'Private as Public' Valuations | 4 | 5 | 6 | 2 | When Jason dismisses Ramp's valuation jump from $23B to $30B as unimpressive, Rory directly pushes back and introduces the concept of 'private as public' pricing for mature growth startups. Harry adds context on fund deployment and check math. | |
| Spray and Pray vs. Picking and the Reality of Venture Disappointment | 5 | 6 | 6 | 3 | Rory directly challenges Jason's blog post interpretation of Carta Series B data, arguing that stock selection discipline is essential. Harry brings up seed betting numbers comparing Andreessen to Sequoia and shares personal venture losses. | |
| Synthesia and the $3 Billion Acquisition Dilemma | 4 | 6 | 3 | 2 | The group analyzes Synthesia rejecting Adobe's $3B acquisition offer. Jason details the structural disconnect between VC fund-return goals and founder personal liquidity, while Rory argues that Synthesia's rapid ARR growth justifies staying independent. | |
| Amazon-Roomba Antitrust Block and M&A Friction | 6 | 5 | 4 | 3 | Discussing the blocked Amazon-Roomba deal, Rory severely criticizes FTC regulatory overreach. Harry argues strongly for early-stage managers utilizing secondary markets to shorten fund duration, prompting Rory to detail the IRR versus multiple math equation. | |
| Amazon's AI Lag and Leadership Transitions | 4 | 4 | 5 | 2 | Jason gives Jeff Bezos an 'F' for stepping down right before the AI boom, contrasting him with Sergey Brin. Rory defends Bezos's personal decision, while Harry adds key facts regarding Google's 14% ownership stake in Anthropic. | |
| "Agree or Disagree" Quickfire Round | 6 | 4 | 4 | 4 | In the quickfire round, Harry strongly defends Andreessen Horowitz's operational dominance, particularly when expanding into Europe. Jason humorously reflects on why he avoids investing in defense weapons companies like Anduril despite their brag value. |
Statements from this episode (36)
O'Driscoll: Andreessen Horowitz is the Red Army of venture capital
“And recent words is the red army of the venture industry.”
O'Driscoll: OpenAI finalizes restructuring deals with Microsoft and state regulators
“The big news today is OpenAI cut their deal. They cut their deal with Microsoft, and they cut their deal with the attorneys general, the plural of attorney general of Delaware and California, which means they have been able to implement their restructuring, wh…”
O'Driscoll: OpenAI's restructuring brings a public offering significantly closer
“This is a real honest, honest to goodness American corporation, a different kind of American corporation, but they can go public with this. They've gotten out of the straitjacket. So that's the big, and it's big news. It means, you know, what, for example, it …”
Lemkin: Sam Altman receives zero equity in OpenAI's corporate restructuring
“They're a little interesting, but the craziest thing in this deal, because it's unprecedented, I think in our lifetimes is OpenAI said that Sam Altman will still have no shares in the combined entity.”
O'Driscoll: Microsoft Achieved a 10x Return on its $13B OpenAI Investment
“They put in thirteen million, thirteen billion dollars. They got a 10 X on their money as of today.”
O'Driscoll: Bret Taylor Forced Elon Musk to Complete $44B Twitter Acquisition
“For the second time in the last five years, he totally wanted as board chairman of Twitter, where he jammed that down Elon's throat for forty four billion dollars despite his opposition, and he won here today because he unraveled the mess and set it all up for…”
O'Driscoll: Microsoft was the only investor to give OpenAI $1B in 2019
“No one else was writing open AI, a billion dollar check in 2019. Microsoft did, and they got that return.”
O'Driscoll: VC Ownership Targets Matter Usually, But Not for Outliers
“90% of the time, your ownership target is a really meaningful metric, and it should run your business on it. And, you know, 10%, one percent of the time, who the hell cares? One percent of the biggest company on the planet is five billion dollars.”
Lemkin: OpenAI will be the most popular retail IPO in history
“I just can't think of a retail IPO that would be more popular than open AI, right? Just bringing everyone out of the woodwork to put a little bit of their life savings to do it, to ignore the, even if the valuation makes no sense. I just, this would have to be…”
O'Driscoll: SoftBank is funding OpenAI at $300B and $500B valuations simultaneously
“SoftBank is closing two separate deals right now. They're closing their roughly three hundred billion pre direct investment, and they're also doing a share buyback from exist from some existing employees of five hundred billion.”
O'Driscoll: OpenAI will reach a $1 trillion valuation in two years
“On the current trajectory, maybe all I can answer is on the current trajectory and without the euphoria Jason mentioned, it's probably two years because, you know, you do get some attenuation of growth at scale and, you know, at the current thing that, I mean,…”
Lemkin: OpenAI's Web Browser Will Not Be a Major Game-Changer
“I mean, obviously the browser is where we live a lot of our lives, but it could be like MCP. It could end up being interesting, but not not really the game changer that that VCs and others think it'll be.”
O'Driscoll: Mega-fund VC dominance will continue for next four to five years
“This is the world we live in today, and as investors, we'll live in for the next four to five years, right?”
O'Driscoll: Mega-funds overpay at seed to purchase Series B options
“The first is they can literally decide they're not pricing this round, they're buying an option on the next round. If Harry's trying to make his money on seed, and they're simply trying to set themselves to make the money on the B, they can, by definition, pay…”
Stebbings: Top VC firms force LPs into 2:1 fund stapling requirement
“If you're an LP and you want to be in the early stage fund, you will often, very often with top tier brands, have to put in two times that into another fund to get that one dollar in the other, and you don't get your pick of which fund. Very often you have to …”
Stebbings: Mercor hit $500M revenue run rate in 17 months
“Now this company has gone to five hundred million in revenue, faster than anyone else, I believe, in history. I think it was 17 months.”
O'Driscoll: AI model companies spend $300B to $400B on compute
“Because if you think about it, these model companies have grown faster than any company in human history. They're spending three, four hundred billion dollars on compute. They're probably spending three, four billion dollars on RLHF, and they were spending zer…”
Stebbings: Two buyers generate over 50% of AI labeling revenues
“The other thing I will say is you have concentration of buyer unlike any other industry. Two buyers are 50% plus of every one of these labeling providers revenue.”
Lemkin: $23B to $30B valuation bump is not a real up-round
“Going from twenty three billion to thirty billion, I don't even consider it an up round. It's not enough. It's not enough.”
O'Driscoll: Non-AI late-stage VC returns will match public small-cap equities
“As you think about these late, super late growth fund things, the return they will realistically get absent the AI lift is some version of what was the small cap high growth public market return.”
O'Driscoll: AI and delayed IPOs are the biggest recent VC shifts
“The two biggest changes of the last three to five years have been the advent of what's happening in AI and the transformation of the late stage and IPO marketplace to this ultra, ultra late stage where companies are way beyond the IPO threshold and still priva…”
O'Driscoll: Carta data shows 18% of 2018 Series Bs returned over 5x
“What it told people for listeners is they looked at all 547, 2018 series B investments, and then they did a histogram of where they come out in less than a one X, you know, one to a two X. Oh, that was series B. So 547, and then, you know, 20, 18% of them grea…”
O'Driscoll: At most 50% to 60% of fund capital goes to follow-ons
“Most of my money goes in on my initial round. At most, 50, 60% comes in follow ons. So I can't afford to be wrong on two thirds of my money to be right on one third.”
O'Driscoll: Y Combinator is the only firm with a mass-production seed model
“The only people on the planet who have a structural business where they can de-emphasize picking because they can write checks, option checks at scale is Y Combinator because they have a structured advantage in terms of their economics. They are the only peopl…”
Lemkin: A $10B outcome benefits VCs, not founders
“Even 10 is not really worth it for a founder. It is worth it for the VCs.”
Lemkin: Strong VC managers face minimal DPI pressure from LPs
“I just don't know that there's this massive DPI pressure. If you have a hot, if you have a hot hand.”
O'Driscoll: Scale backed a 2x runner-up over a 15x market leader
“We did B. We got a two X. We didn't do A. We left a 15 X on the table.”
O'Driscoll: FTC's block of the Amazon-Roomba deal was absurd and destructive
“And the FTC led by Lena Kahn chose to block that deal because I think Benedict Evans said in his place because of an insipid monopoly in the house vacuum cleaner marketplace, very tongue in cheek, right? It was an absurd decision at the time. It only got more …”
Lemkin: Salesloft exit at $2.5B in December 2021 was market peak
“My last deal sales off December, 20, 21, you know, two and a half billion.”
Stebbings: Ramp reaches $1B revenue as Brex hits $700M near breakeven
“Ramp is at a billion in revenue. Brax is at 700. They're both around the breakeven to moving profitable phase.”
Stebbings: Bullish on Andreessen Horowitz as partner quality matches boutique firms
“They are as good as the partners at smaller firms with three to five partners, and they have been fantastic. Like, I am a Big bull on Andreessen now where I was not before.”