May 31, 2025 · 38m · a16z
Where Crypto Meets AI with Chris Dixon & David George
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 featured talk from the a16z LP Summit 2025, General Partners Chris Dixon and David George explore the intersecting evolutions of cryptocurrency and generative AI. They analyze key technical breakthroughs, changing web business models, market defensibility frameworks, and what it takes for early-stage founders to build category-defining companies.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 6.0 Guest teaching 3.0 Guest disagreement 1.2 The host pushing back 1.5
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 18:17 Rhetorical Push on Google's Search Revenue
Chris Dixon forcefully challenges the realistic options open to Google, asking if they should simply abandon over $100 billion in search revenue to pursue AI search.
Hardest push from the host ▶ 25:01 Pushback on the Obsolescence of Ad ModelsDavid George refuses Chris Dixon's implicit framing that ad-based models may not exist in the future, arguing that in consumer internet everything eventually converts to an ad model.
Biggest teaching moment ▶ 12:30 Coordination vs. Intelligence FramingChris Dixon clearly delineates the core difference between AI and crypto, educating the listener that AI creates intelligence while crypto solves collective action and coordination problems.
The host holds their own ▶ 13:46 Detailed AI Cost Curves and Capability MetricsDavid George demonstrates deep domain knowledge by citing specific empirical data, including a 99% cost reduction in AI over two years and capabilities doubling every seven months.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| The Current State of Crypto and Infrastructure Progress | 5 | 3 | 1 | 1 | David George active-lists and contributes solid contextual knowledge, citing Stripe's stablecoin acquisition and monthly Visa volume comparisons. Chris Dixon explains the regulatory environment and infrastructure progress without friction, keeping the tone entirely collaborative. | |
| Generative AI, Second-Order Effects, and Coordination vs. Intelligence | 4 | 3 | 1 | 1 | Chris Dixon delivers a broad conceptual overview of AI history, second-order technological effects, and the distinction between crypto coordination and AI intelligence. David George joins in naturally, contributing observations on AI impacting the laptop class and agreeing with Chris's historical analogies. | |
| AI Cost Curves, Barbell Value Accrual, and Intersecting Megatrends | 7 | 2 | 1 | 1 | David George demonstrates strong expertise by citing quantitative data on AI cost reductions and model capability doubling rates. Chris Dixon defers to David and other firm verticals on AI specifics, agreeing on hardware intersections. | |
| Disruption of Web Search and Google's Innovator Dilemma | 7 | 3 | 2 | 2 | David George presents detailed analysis on search volume shifts away from Google and frames Google's innovator's dilemma. Chris Dixon offers a mild rhetorical challenge, questioning how Google could ever abandon $100B+ in search revenue, leading to a thoughtful peer exchange. | |
| The Broken Covenant of the Web and AI's Impact on Content Creators | 6 | 4 | 1 | 3 | Chris Dixon educates on his concept of the web's broken covenant caused by AI answer engines. David George responds with industry examples like Chegg and pushes back on the notion that ad-based models are obsolete, predicting native ad evolution. | |
| Market Structures, Moats, and the Glengarry Glen Ross Framework | 7 | 3 | 1 | 1 | David George introduces the Glengarry Glen Ross market structure framework alongside consumer AI growth stats. Chris Dixon elaborates on venture capital category selection rules, showing deep mutual alignment between interviewer and guest. |
Statements from this episode (20)
Dixon: Sending crypto for under a penny and second is achievable
“You know, one of the big goals was to have, to be able to send money for under one penny and under one second, And that is now achievable if you download, like, the Coinbase wallet, for example, you can do that.”
Dixon: Monthly stablecoin volume has reached trillions, surpassing Visa
“For those tracking it, like, stablecoins have grown dramatically in usage. It's now in the trillions of dollars, you know, higher than Visa per month.”
Dixon: Stablecoin legislation will unlock bank adoption and accelerate network growth
“We're trying to get legislation, and that will, I think, further accelerate it, because you've got, essentially, You know, 99, think of it as a network effect, and you've got 90, probably five, 90, or high percent, or some percentage of the nodes on the networ…”
George: Stripe's stablecoin company purchase was its largest acquisition ever
“Stripe had their big, you know, product conference, and one of the big things they talked about, they did their biggest acquisition ever, they bought a stablecoin company”
Dixon: SpaceX's Starlink uses stablecoins for cross-border money transfers
“SpaceX, like Starlink is like, uses it to move money from one country to another”
Dixon: Cryptocurrency is a second-order effect of social media
“I think of crypto as being a second order effect of social media.”
Dixon: People will make AI-generated feature films within two years
“In two years, you'll have people making feature-length films.”
Dixon: Crypto solves coordination problems, whereas AI solves intelligence
“What crypto is really about is not about intelligence, it's about coordination, right? It's about collective, it's solving collective action problems and coordination problems, which is, I think of as an orthogonal set, a different set of kind of problems to s…”
George: AI inference costs have dropped 99% over the past two years
“On the AI side, there's been a 99% reduction in cost over the last two years.”
George: AI model capabilities on higher-order tasks double every seven months
“I think if you give it a set of tasks that are slightly higher order, I think the models are currently doubling their capabilities every seven months.”
George: Web transactions will soon occur without navigating through Google
“We're getting pretty close to the point where you can actually conduct business activity on the Internet without having to go through Google.”
George: Google's Gemini models are currently state-of-the-art across multiple domains
“They actually have the best models right now. Like, they have the leading, you know, the state-of-the-art models in Gemini across a bunch of different domains.”
Dixon: Google's business model relies on delivering links over direct answers
“The whole business model is predicated on not actually getting the, it's on getting the link and not the actual product, right?”
Dixon: ChatGPT effectively one-boxes the entire internet, ending referral traffic
“In that framework, you can think of AI, like ChatGPT, as sort of one-boxing the whole internet, right? It's just like, it gives you the answer. You don't need to click through, right?”
George: Generative AI obviated Chegg's core business virtually overnight
“Chegg is the best example, right, which is basically, it was a high-flying public company, it's a place where, you know, kids would go to get homework help and, you know, study guides and things like that, and then, like, overnight, like, got hit by a ton of b…”
George: AI business models will adopt freemium over simple subscriptions
“And so, you know, I think there's a question of how it will play out. Will it just be a simple subscription or is there gonna be something more You know, freemium that comes along. I would bet on the latter. I would bet on the latter.”
George: Consumer AI reached one billion users faster than any predecessor
“We've now gotten to the point where there's about a billion monthly active users of consumer AI applications, and this has grown significantly faster than predecessor companies, Google, Facebook, TikTok, and in the case of TikTok, they actually grew through, t…”
George: AI market value will accrue primarily to applications and chips
“We're very high conviction that at the application layer you will have tremendous value and at the chip layer you will have tremendous value.”