Jan 15, 2026 · 1h 30m · news
Anthropic’s $10B Raise | a16z’s $15B Fund: Is the Middle Dead in VC? | How OpenAI Could Go to Zero? · 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, host Harry Stebbings and co-hosts Rory O'Driscoll and Jason Lemkin analyze the changing landscape of venture capital, debating the massive valuations of AI giants, the viability of mega-funds, and the structural threat of California's proposed wealth tax.
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 8.5% of the talking time here. How this is scored →
Harry as informed peer 3.7 Guest teaching 6.1 Guest disagreement 2.7 Harry pushing back 3.0
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 21:01 Jason's Bear Case: OpenAI as Detroit or AOL
Jason forcefully rejects Rory's cash-buffer defense, arguing OpenAI risks becoming like Detroit or AOL if capital freezes while fast-improving models make past architectures obsolete.
Hardest push from Harry ▶ 46:33 Harry Confronts Rory: 'Are You Not The Middle?'Harry directly challenges Rory's positioning by citing Alex Rampell's premise that the VC middle is dead and explicitly asking Rory to account for his own fund size.
Biggest teaching moment ▶ 28:06 Rory's Macro VC Math on a16z's $15B FundRory delivers an detailed breakdown of macro exit values, capital deployment equilibrium, and Series A market share required to make a mega-fund return.
Harry holds his own ▶ 24:12 Harry Cites 22% ChatGPT Drop DataHarry counters Rory's assumption about consumer stickiness by introducing specific data showing a 22% drop in ChatGPT usage following Gemini's update.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Analyzing Anthropic's $10B Raise and Valuation | 1 | 6 | 1 | 0 | Harry introduces the topic and asks open questions about Anthropic's pricing and IPO timing. Rory provides a detailed mathematical breakdown of Anthropic's ARR trajectory and revenue multiples compared to public comps like Palantir and Cloudflare. | |
| Developer Tools Battle: Claude Code vs. Cursor | 6 | 4 | 3 | 6 | Harry brings data from CPOs on his 20 Products show, noting a sharp drop in Cursor usage in favor of Claude Code, and presses whether Cursor investors should be nervous. Rory counters that Harry is getting lost in details before Jason introduces the frog and scorpion metaphor regarding API access. | |
| Apple's Partnership with Google Gemini and Enterprise Privacy | 4 | 5 | 1 | 3 | Harry asks a structured question about Google Gemini securing the Apple Siri relationship over OpenAI. Rory explains ad monetization search dynamics versus LLM deal economics, while Jason highlights the critical importance of enterprise privacy in cloud infrastructure. | |
| The Existential Risks and Bear Case for OpenAI | 7 | 6 | 7 | 7 | Harry challenges OpenAI's stability, citing high SBC and churn. Jason lays out a forceful bear case comparing OpenAI to Detroit or AOL if capital dries up, while Harry hits back with data showing a 22% drop in ChatGPT usage following Gemini's release. | |
| Evaluating Andreessen Horowitz's $15B Mega-Fund | 4 | 7 | 3 | 3 | Harry asks how a16z can return a $15B fund. Rory delivers a comprehensive monologue analyzing total VC exit volume, market equilibrium, and the 10% Series A market share required for top returns. | |
| Is the Middle Dead? Boutique vs. Platform VC | 7 | 5 | 5 | 8 | Harry directly confronts Rory by citing Alex Rampell's statement that the middle in VC is dead and asking Rory if his firm is the middle. Rory defends boutique focus while Jason explores finding glitches in the matrix outside the dominant discovery channels. | |
| Valuation Risk in Late-Stage VC | 2 | 8 | 1 | 1 | Rory educates on the difference between uncorrelated early-stage business risk and 100% correlated late-stage valuation risk, using Databricks' growth deceleration scenario as an example. | |
| The Reality of Eleven Labs' Pricing & Jason's "Vibe Coding" Game | 2 | 5 | 1 | 1 | Jason shares his personal experience vibe-coding Founderscape.ai, explaining how he burned $30 in ElevenLabs voice credits in 48 hours across just a few players. | |
| Debating Eleven Labs' $11B Valuation & The Threat of Substitution Risk | 5 | 5 | 4 | 6 | Harry pushes Jason on whether he would invest in ElevenLabs at an $11B valuation and whether a 3-5x return exists. Jason explains why substitution risk and high usage costs make him hesitant. | |
| Rory's Historical Perspective: Nuance, Enterprise Spend, and Eleven Labs' Scaling Path | 2 | 8 | 1 | 1 | Rory brings historical context from Nuance in the 1990s and calculates that an $11B valuation requires $5B in steady revenue, questioning if the speech market can support that without high customer concentration. | |
| The Product Paradox: Does Ease of Adoption Breed High Churn? | 2 | 6 | 4 | 1 | Jason argues that a 5-minute setup time makes switching vendors equally fast, while Rory counters that extreme ease of adoption is a competitive moat similar to Stripe. | |
| The "Trojan Horse" Strategy: Why Paper Wealth Taxes Will Kill Silicon Valley | 3 | 7 | 2 | 1 | Jason breaks down how California's proposed tax proposal acts as a Trojan horse intended to eventually lower the threshold to $25M in paper wealth. | |
| California's Referendum System & The Economic Damage of Proposed Taxes | 1 | 6 | 2 | 0 | Rory notes that while California referendums usually get voted down, the mere threat of illiquid wealth taxes creates self-inflicted economic harm. | |
| The Threat of a "Detroit in Silicon Valley": Exit Memes & Winning States | 4 | 7 | 3 | 3 | Harry asks what happens if the tax passes. Jason warns of a 'Detroit in Silicon Valley' dynamic where founders participate in YC in SF but relocate before their Series B. | |
| Would Jason Leave California? High Taxes & The Shift Back to Remote Work | 4 | 5 | 2 | 4 | Harry asks Jason directly if he would leave California. Jason candidly admits he would leave if annual paper wealth taxes were enacted, predicting a return to distributed teams. | |
| The AI Wealth Gap: Nvidia Millionaires vs. SaaS Layoffs & Social Unrest | 5 | 7 | 3 | 3 | Harry brings up labor displacement concerns in the UK and US. Jason highlights the stark divide between 18,000 Nvidia employees worth $25M+ and laid-off SaaS workers facing an impossible job market. |
Statements from this episode (45)
O'Driscoll: Late-stage VC involves correlated valuation risk, unlike early-stage
“In the early stage, you're taking uncorrelated business risk, and in the late stage, you're taking a hundred percent correlated valuation risk.”
Lemkin: Andreessen Horowitz has the strongest founder brand in VC
“It's pretty interesting that Andres not only raised the most capital, but on a two by two, I think has the strongest founder brand.”
O'Driscoll: Anthropic's $10B raise will likely be its last before IPO
“I think it probably will be the last one before the IPO.”
O'Driscoll: Anthropic grew ARR from $100M in 2023 to ~$10B in 2025
“They went from a 100,000,023 at the end of 23 run rate to a billion at the end of 24 run rate. To allegedly about between nine and ten billion at the end of 25.”
O'Driscoll: Anthropic's revenue multiple is lower than Palantir's
“It's a much lower revenue multiple than Palantir. It's kind of comparable with Cloudflare for God's sake in the public markets.”
Lemkin: Anthropic cut off xAI's access to its models
“Even little things like Anthropic cutting off XAI's access to Anthropic this week.”
Lemkin: Anthropic generates roughly $1B annually from Cursor
“Right now, it's great for Anthropic to get an extra billion or so a year from Cursor.”
O'Driscoll: AI startups face existential disruption every six months
“In fast land, you could compound for seven or eight years. Now there's existential risk every six months. And if you can't live with that, you probably need to find a different job.”
Lemkin: Google is currently a much more stable partner than OpenAI
“Certainly today, Google feels like a far more stable partner than OpenAI. It's just the reality. It's not the, OpenAI is not the only game in town anymore for Apple.”
O'Driscoll: Anthropic has closed its valuation gap with OpenAI to 2:1
“The way I keep score is the relative value of Amtropic to open AI is kind of the ratio of let's call it management success from over the last three years. And the truth is it's gone from, you know, 10 or eight to, you know, 10 plus to one to much more converge…”
Lemkin: The shelf life of a state-of-the-art LLM is under 100 days
“There is a very simple bear case that it goes almost to zero which is that shelf life of an LLM is less than a hundred days.”
Stebbings: ChatGPT usage dropped 22% after Google released new Gemini models
“Since the new Gemini models came out, you've had a 22% drop in ChatGPT usage.”
Lemkin: Andreessen Horowitz has achieved top-decile VC fund returns
“The returns were published. The returns are top decile. So you have the biggest fund, the top decile returns or quartile, whatever, top tier, which used to be a knock. And founders love this brand.”
O'Driscoll: a16z backed 10% of $5B+ Series A outcomes over last decade
“That work that we talked about way back last year that the partner from DST did showed by over the last decade, Andreessen did roughly 10% of all series A's that became five billion dollar outcomes.”
O'Driscoll: Total value of all private companies reaches $3.6 trillion
“And the total exits, the total value of all private companies right now is about 3.6 trillion, which pleasing it, by the way, if you say. Three X on invested capital. That's 1.2 trillion over 10 years.”
O'Driscoll: VC mega-funds cannot generate target returns without top outliers
“You simply can't make this kind of math work without getting those top exits. And you don't have to get in at the A, but, you know, you don't have to do the A of SpaceX, but you better show up on the cap table before they hit total, because that's a trillion d…”
Lemkin: Andreessen Horowitz has never had a fund under 3x
“And Andreessen hasn't had a fund below four, four X gross, gross. Hasn't had a fund below three X.”
O'Driscoll: a16z wins more top Series A deals than Benchmark
“Andreessen's market share is higher than benchmarks in terms of that. That worked. And I wish Rotman, the guy from DST did. It's higher in terms of the great series A's, right? As a market share, but the hit rate is much lower.”
O'Driscoll: Investment decisions degrade in rooms with over seven people
“Stock picking in general, it gets hard to be smart in a room with more than five or seven people in it.”
Lemkin: Founders do not care about multi-stage funding pitches from VCs
“I don't believe, no matter what VCs tell founders in their spiels and pitches, I don't believe founders highly value the fact that VCs can fund you through every stage.”
O'Driscoll: Growth funds allow VCs to make more Series A mistakes
“You can be promiscuous at the A if you have enough late stage stuff to cover it up.”
Stebbings: Series A is the worst stage for venture capital returns
“The worst place to be investing is Series A. You either need to be pre-seed or pre-IPO today. To make money.”
O'Driscoll: Top VCs are on 80% of Series C cap tables
“And, you know, at the A, it's 40, 50% of total deals. The interesting, it climbs steadily, and by the C, it's about 80%. In other words, they, let's call it the Rory hard to beat mental list. By the time you get to the C, 80% of the time, one of, they have one…”
O'Driscoll: First venture rounds for OpenAI and Anthropic were $23B and $4B
“The first venture round at OpenAI was twenty-three billion pre with Thrive, and the first venture round at Entropic was four billion pre with Spark and Menlo, you know, right?”
O'Driscoll: Databricks generates $4.5B-$5B at 40%+ growth and is cash-flow positive
“I'm going to pick on, in my view, one of the best companies out there, Databricks. It's doing four and a half, five billion dollars, right? It's valued at, it's got a growth rate of 40% plus. It's cashflow positive.”
O'Driscoll: 20% growth tech companies historically trade at 6x revenue
“Across the last two decades, 20% growth companies, capitalized, trade around six times.”
Lemkin: ElevenLabs generated $330M in revenue last year
“So my point is 11 Labs ended last year. They just said it. Mahdi is such a great CEO on so many levels, right? So charismatic, so good. Three hundred and thirty million in revenue from nothing.”
Lemkin: Developers will switch from ElevenLabs if cheaper alternatives emerge
“But I, if for my game, if I could do something, attempt the price or a 50th the price that was close to as good, I would have to switch.”
Lemkin: AI substitution risk will become critical in H2 2026
“This will be the year. It'll probably be the back half of the year where we have to take substitution risks seriously in AI.”
O'Driscoll: ElevenLabs needs $5B in revenue for a 3x return
“11 or twelve billion, you want to make it three X, so you got to be worth thirty billion. Right. Thirty billion at scale. You're going to trade at six or seven times. Cause that's the way human life is to get over. Right. So that's five billion of revenue.”
O'Driscoll: Distributed SMB customer bases are more defensible than enterprise concentration
“If that revenue comes from Epic games, putting voice in all their games and they're paying Epic games, half a billion dollars, Epic games is going to design them out. Are they going to grind them on cost? Are there's going to be a competitor? They have the bes…”
O'Driscoll: Cheaper compute will solve AI software gross margin problems
“I do think a lot of these non-gross margin positive things, the good thing is they're all digital products, so they'll get, yes, time will take care of a lot of that, you know, and cheaper compute.”
O'Driscoll: Wealth taxes underperform revenue projections due to capital mobility
“All wealth taxes underperform what the people project they'll raise because it tends to be very mobile and it's, it does, it's very hard to tax that and people can move.”
O'Driscoll: Founders retaining voting control in public markets is beneficial
“I would say that's something I didn't agree with 10 years ago and I've changed my mind totally. I think it's good in the public markets that founders have voting control.”
O'Driscoll: California's proposed tax will drive wealthy founders out of state
“So are you going to sit in California? If you're worth two billion dollars and say to yourself, I'm going to give a billion dollars for the privilege of living here. I don't think so. You're going to leave. So I think it's going to be fairly pernicious to what…”
Lemkin: Founders will massively exit California in 2027 if wealth tax passes
“I think founders will begin to massively exit in 20, 27 before the next one goes up because there's going to be a second if this passes and a third.”
O'Driscoll: Proposed California wealth tax referendum will ultimately be defeated
“So my guess is in the, my guess, and maybe I'm being optimistic. My guess is in the end, this loses.”
Lemkin: California wealth tax will turn Silicon Valley into Detroit
“And so, I do think if this goes as far as the folks that have Backing at want, you could have a Detroit in the Silicon Valley. Like, when it becomes a meme to do YC, or to do South Park Common, but then build, you know, get your money, build your team, and the…”
Lemkin: Miami and Austin will win tech exodus if tax passes
“The answer is the ones that almost won in 2020 and 20 21. It's that simple because there wasn't enough gravity to get people to go to Miami outside of, you know, some hedge funds and others, and there wasn't enough gravity to go to Austin because it's really n…”
Lemkin: Founders will stay in SF for YC but leave after Series B
“I worry that the, that it just makes sense to leave after the Series B. You should just leave. And I don't, I, and I think YC will get their seven percent, and the funds will still stay, but you'll just leave.”
Lemkin: B2B SaaS will normalize around $1M-$2M revenue per employee
“I think for, even for B to B grounded in what we do, I think we're going to normalize around a million dollars to two million dollars per employee.”