May 3, 2023 · 56m · news
Vince Hankes: Why We Put $300M into OpenAI; Sam Altman's Pitch; Lessons from Josh Kushner | E1009 · 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
This video features an in-depth interview with Thrive Capital partner Vince Hankes, who shares his professional journey, details Thrive's high-profile OpenAI investment, and analyzes the evolving landscape of AI value accrual, startup strategy, and venture capital culture.
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 20.6% of the talking time here. How this is scored →
Harry as informed peer 2.1 Guest teaching 2.5 Guest disagreement 1.1 Harry pushing back 1.4
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 51:31 Firm refusal during quick-fire question
Harry repeatedly tries to coax Vince into naming another multi-stage firm to invest in, suggesting Founders Fund or Sequoia, but Vince flatly rejects the host's premise, stating they only invest in themselves.
Hardest push from Harry ▶ 27:00 Host pushes back on startup value captureHarry quotes investor Alex Rampell regarding distribution vs innovation races and directly asserts his opinion that 90% of AI value will be captured by incumbents, challenging Vince to disagree.
Biggest teaching moment ▶ 46:00 Post-mortem on misapplying pattern matching to CanvaVince educates the host on investor cognitive traps by breaking down how evaluating Canva through standard enterprise SaaS metrics caused him to miss the prosumer opportunity during his Tiger tenure.
Harry holds his own ▶ 27:00 Host deploys venture framework to frame industry dynamicsHarry demonstrates high domain expertise by citing Alex Rampell's canonical framing on incumbent distribution versus startup innovation speed to challenge Vince's thesis.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Vince's Path to Thrive Capital | 1 | 2 | 0 | 0 | Harry asks warm, open-ended questions about Vince's transition from Goldman and Tiger to Thrive. Vince outlines his career trajectory and contrasts Tiger's financial rigor with Thrive's founder-empathy mindset. | |
| Personal Motivators and Learning Curves | 1 | 2 | 2 | 1 | Harry asks a deeply personal question about what Vince is 'running from.' Vince jokingly deflects the premise before offering a thoughtful response on compounding learning curves. | |
| Building Trust with Founders | 1 | 2 | 0 | 0 | Harry shares feedback from founders on Vince's even keel. Vince explains how predictability builds founder trust and how maintaining balance prevents knee-jerk reactions to market volatility. | |
| AI Hype Cycle vs. Fundamental Value | 2 | 3 | 1 | 1 | Harry introduces the question of whether AI is a hype cycle like crypto or a fundamental platform shift. Vince contrasts crypto's ideological push with AI's concrete productivity gains, citing dot-com stats. | |
| The Origin of Thrive's OpenAI Investment | 2 | 3 | 1 | 2 | Harry asks how Thrive got into OpenAI's round and interjects to ask why the deal wasn't obvious to everyone. Vince explains how traditional investors get trapped by precise TAM analysis early in shift cycles. | |
| Rethinking TAM and Market Analysis in AI | 2 | 4 | 1 | 1 | Vince educates on historical technology scaling curves like iPhone, AWS, and Google. He reframes how founders select AI models, moving away from model size to infrastructure reliability and developer friction. | |
| Addressing Model Commoditization and Ecosystem Moats | 4 | 3 | 2 | 4 | Harry presses on model commoditization and cites Tom Tunguz calling Google's AI efforts disappointing. Vince counters that big tech incumbents possess immense talent and are shipping surprisingly fast. | |
| Incumbents vs. Startups: Where Will Value Accrue? | 5 | 3 | 3 | 4 | Harry quotes Alex Rampell on incumbents acquiring innovation vs startups acquiring distribution, asserting 90% of value will accrue to incumbents. Vince rejects the 90% premise, reframing the debate around new category creation. | |
| Value Accrual: Application Layer vs. Infrastructure Layer | 3 | 3 | 1 | 1 | Harry asks whether value accrues at the infrastructure or application layer. Vince outlines why application layer value creation will far outstrip infrastructure tolls over time. | |
| The Role and Challenges of AI Regulation | 3 | 2 | 1 | 3 | Harry asks about regulation and highlights the vast knowledge gap between private tech builders and government regulators. Vince defends OpenAI's cautious safety testing prior to GPT-4's release. | |
| Valuation, Pricing, and Team Decision-Making | 3 | 3 | 1 | 3 | Harry brings up the reported $29B valuation and presses Vince on how Thrive justified the high entry price. Vince explains how exponential technological adoption breaks standard financial spreadsheets. | |
| The Role of Financial Models in Venture Capital | 2 | 3 | 2 | 2 | Harry asks when to throw financial models out the window. Vince clarifies that models aren't thrown out, but rather weighted as one tool among many when evaluating exponential outliers. | |
| Thrive's Structural Agility and Psychological Safety | 1 | 2 | 0 | 0 | Harry inquires about Thrive's internal fund structure. Vince details how psychological safety and spirited debate allow their team to make large, non-consensus investment calls without fear. | |
| Vince's Evolution as an Investor: Deep Customer Empathy | 1 | 2 | 0 | 0 | Harry asks how Vince's perspective on investing has evolved. Vince reflects on shifting from purely financial analysis to understanding organizational design and deep customer empathy. | |
| Learning from Mistakes: The Canva Case Study | 1 | 4 | 0 | 0 | Harry asks for Vince's biggest investment mistake. Vince gives a candid post-mortem on passing on Canva at Tiger due to rigid enterprise SaaS pattern matching. | |
| Defining Investment Success through Authenticity | 2 | 2 | 1 | 1 | Vince explains why Thrive focuses forward rather than celebrating past wins, noting that authenticity wins deals. Harry praises Vince's unusual candor regarding past mistakes. | |
| Quick Fire: Other Funds, Seed Investors, and Changing Minds | 2 | 1 | 3 | 3 | During quick fire, Harry tries to force Vince to select a competing multi-stage fund like Sequoia or Founders Fund. Vince firmly refuses to pick any firm other than Thrive. | |
| Lessons from Josh Kushner and Future Ambitions | 1 | 1 | 0 | 0 | Harry asks for lessons learned from Josh Kushner and Vince's 5-year outlook. Vince shares personal leadership lessons regarding family priorities and closing pleasantries. |
Statements from this episode (27)
Hankes: Only Customer and Product Questions Matter for Any Company
“There's only two questions that matter for a company. Who is the customer and what is the product? And you'd be shocked at how many folks can't really answer that question with clarity.”
Hankes: Asking Operator-Level Questions Fosters Deeper Founder Empathy
“The way we'd ask that same question is, if I'm a SDR on your sales team and I'm trying to qualify a prospect, you know, what am I looking for? At the end of the day, it's really the same question. You could even argue that the way we do it at Thrive is less ef…”
Hankes: Building Trust With Founders Requires Extreme Predictability
“Trust with founders is actually just being very predictable. People want to not feel like they're getting surprised by how you're thinking. They want to understand how you think, how you're going to react and feel like, you know, they understand you.”
Hankes: Investors Over-Extrapolate Good Times and Under-Extrapolate Bad Times
“In good times, people over extrapolate in bad times, people under extrapolate.”
Hankes: Early AI Startups Were Thin Wrappers Around Underlying Models
“Many of these companies and we spent a bunch of time with them and we kind of kept coming back to the fact that they were thin user experiences on top of this cool thing, which was this model”
Hankes: Sam Altman Pitched OpenAI Round With Closed GPT-4 Demos
“The way we kicked off the round was he did almost like a closed demo with lots of investors on a couple of calls of the technology they're working on and ultimately in GPT-IV.”
Hankes: Many Venture Capital Firms Passed on Investing in OpenAI
“Well, you know, I think there's a lot of people that said no. So at the end of the day, I, you know, there, as much as I like to think we want a really competitive deal, I do think It was not obvious to everybody, and even now I don't think it's still obvious …”
Hankes: Disrupting Google Search Is a Trillion-Dollar Opportunity
“We're not talking about building the next unicorn or decacorn. We're talking about disrupting search or Google. I mean, that's a trillion dollar opportunity.”
Hankes: AI's World-Transforming Impact Equals iPhone, AWS, and Google
“But the technology, if you really think about it and the zeitgeist it's captured, I mean, it is of that elk, in my opinion, of the transformation it can have of the world.”
Hankes: Resource-Constrained Startups Shouldn't Build Open-Source AI Infrastructure
“And some companies will choose that path, but by and large for the vast majority of companies, particularly ones that have scarce resources, which are many of the ones I work with, you know, I don't want you spending time building infrastructure and scaling op…”
Hankes: AI Moats Will Be Ecosystems, Not Model Outputs, Within Five Years
“I do think, again, the dimension by which people are going to think about this in five years from now, looking back, isn't going to be about The commoditization of the model and the raw output, it's going to be, oh, the ecosystem around the model has become mu…”
Hankes: Best AI Talent Remains Heavily Clustered in Big Tech
“If you think about where the best talent in AI is right now, I think it's OpenAI, And then I think you'd, everyone would tell you it is Google and Facebook and Microsoft and Amazon. And maybe there are folks that have kind of dripped their way into the startup…”
Hankes: AI Startups Will Struggle to Compete With Fast-Shipping Incumbents
“And so if you're a startup, I get that you have the advantage of speed, but you need a multi-year execution window to be able to take advantage of all that product you can ship quickly. If these big companies are able to ship this quickly, they've got so much …”
Stebbings: Incumbents Will Capture 90% of AI Value Creation
“And I more lean towards 90% of it being captured by incumbents, which isn't great for us as venture investors”
Hankes: Bet on Startups Creating Novel User Experiences Over Incumbents
“Yes, if you're going toe to toe right now with an incumbent on their home turf and their shipping, I think it's a hard bet to take the opposite side of, you know, in our business. But if you're creating something that's totally a different user experience that…”
Hankes: Most AI Value Will Accrue to Applications, Not Infrastructure
“I have to believe it's gonna accrue mostly to the application layer. I think if you think about infrastructure companies as platforms, you know, if you think about the software market as the relatives, the value of AWS and Microsoft and Google cloud versus the…”
Hankes: VCs Should Withhold Capital Until AI Product Value Becomes Clear
“If you're really answering that question, you're struggling because you think in two weeks from now, there's a chance that the customer doesn't care or the product won't have any value. Well, that's hard to build a company around. And so in any of those enviro…”
Hankes: OpenAI Delayed GPT-4 Release by Five Months for Safety Testing
“On GPT four, we saw the demo in the fall. They didn't just release the product then they took, I think four or five months To test and learn about safety in the edges of the model and then ultimately released it to the world.”
Hankes: Regulators Have Successfully Brought Themselves Up to Speed on Crypto
“Regulators, I've been impressed with how up to speed they've gotten on crypto, and they've really taken time to learn about it.”
Hankes: Thrive Separates From Quantitative Rigor for High-Upside Bets
“We're able to kind of separate ourselves from the kind of quantitative rigor that, you know, we rely on for a lot of investments we do and hold the tension of what could go right.”
Hankes: Waiting for Clear Proof in Tech Means Getting Priced Out
“The scalability properties of what they have built are so good that, you know, we'd rather bet on the compounding upside of that scalability Than sit on the sidelines because once it's clear, it'll get priced up very quickly.”
Hankes: Iconic Companies Always Get Priced Ahead and Look Cheap Later
“No iconic company, you know, at least that we've been a part of, was it clear and obvious in the early days of that technology, they always get priced well ahead. And they look cheap in hindsight because they defy the gravity of the model.”
Hankes: Many VC Firms Lack Structure for Massive, High-Risk Investments
“And frankly, I think there are a lot of firms that Could not have done this because their organizations are not set up to make these kinds of big decisions.”
Hankes: Best Companies Have Simple Strategies Understood at All Levels
“The best companies have very, very simple explanations that, you know, it trickles down on the company. The CEO can articulate in one sentence, but the manager four rungs down also can articulate it. And that means that that manager knows how to go left when t…”
Hankes: Tiger Global Passed on Canva by Misapplying Enterprise Software Metrics
“When I was at Tiger, I flew out to Sydney and spent a bunch of time with the Canva team in person. And, you know, it's obviously now people know about it. It's a remarkable company at the time. It was much smaller than it is now. And At the same time, we were …”
Hankes: Rigid Metric Pattern-Matching Causes Major Venture Capital Mistakes
“You can't walk in biased about what you're looking for out of the metrics of a company or what the product should result in in metrics. You should walk in and try to have a very open mind and say, do these metrics explain the qualitative of the business that I…”
Hankes: Investors Over-Attribute Startup Success to Execution During Market Booms
“This is probably more, this is not for the rapid fire, but you know, one thing that's become clear as we've kind of come out of or gone into this more difficult environment to operate in is in the good times, we probably over attribute to teams and products, h…”