Dec 3, 2025 · 50m · a16z
Why AI Moats Still Matter (And How They've Changed)
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 The a16z Show, General Partners David Haber and Alex Rampell join host Erik Torenberg to discuss how AI is reshaping software defensibility, labor economics, and competitive moats. They examine why AI enables software to perform actual work rather than acting as a seat-based IT utility, and explore how vertical startups can build enduring value alongside foundation model platforms and enterprise incumbents.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The host holds 5.3% of the talking time here. How this is scored →
The host as informed peer 2.0 Guest teaching 3.0 Guest disagreement 1.3 The host pushing back 0.4
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 48:00 Dismissing anti-AI political commentary
Alex Rampell aggressively mocks political efforts to restrict AI, sarcastically suggesting elected representatives want to turn Silicon Valley back into tangerine farmers.
Hardest push from the host ▶ 39:15 Steve Jobs Dropbox counterexampleThe host directly challenges Rampell's feature-versus-product framework by recalling Steve Jobs's famous dismissal of Dropbox as merely a feature.
Biggest teaching moment ▶ 29:00 The gold bricks theory of platform prioritizationAlex Rampell educates the audience and host on why foundation model platforms leave massive vertical markets open using Dan Rose's 'gold bricks' framework.
The host holds their own ▶ 16:24 Demanding a steel-man of the opponent's viewThe host shows strong interview command by refusing to accept consensus and explicitly forcing the guests to articulate the strongest counterargument for brand and momentum as moats.
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 |
|---|---|---|---|---|---|---|
| Do Moats Still Matter in the AI Era? | 1 | 2 | 0 | 0 | The host opens with a broad question about whether moats still matter. David Haber and Alex Rampell establish their thesis on differentiation versus defensibility, using the analogy of scale acting like gravity for data network effects. | |
| Defensibility, Enterprise Software Pricing, and Incumbent Vulnerability | 1 | 3 | 1 | 0 | The host asks if AI companies are more or less defensible than Web2 incumbents. Rampell reframes enterprise pricing models and cites Clay Christensen's overshoot theory to explain incumbent vulnerabilities. | |
| The Goldilocks Zone vs. Greenfield Opportunities | 2 | 4 | 1 | 0 | The host prompts for examples of the Goldilocks zone versus Greenfield opportunities. Rampell explains how stickiness in payroll software like ADP differs from rationalized SaaS spend and EHR systems. | |
| Steel-Manning Momentum, Brand, and Scale Effects | 3 | 2 | 1 | 1 | The host demonstrates conceptual understanding by asking guests to steel-man the counter-thesis that brand and shipping momentum are the real moats in AI. Haber and Rampell unpack economies of scale and founder context. | |
| Foundation Model Platforms vs. Vertical AI Applications | 2 | 4 | 1 | 0 | The host asks how founders should evaluate competition from foundation model owners. Rampell outlines feature versus product versus company dynamics, sharing a anecdote about Facebook's 'gold bricks' prioritization. | |
| Platform Prioritization and Market Shakeout Dynamics | 2 | 3 | 1 | 0 | The host asks what OpenAI should build first and whether market consolidation is inevitable. Rampell details historical platform strategies like Microsoft versus Borland and market shakeout mechanics. | |
| Model Provider Cutthroats, Feature Evolution, and Incumbent Consensus | 4 | 3 | 2 | 2 | The host pushes back on the 'feature vs product' distinction by citing Steve Jobs famously dismissing Dropbox as a feature. Rampell accepts the premise but explains how great founders backfill products and criticizes incumbent software slowness like Apple's Screen Time. | |
| Value Capture, Disruption of BPOs, and the Economics of AI Labor Abundance | 1 | 3 | 3 | 0 | The host asks how value capture will split between startups and incumbents. Rampell passionately reframes the AI job disruption debate, mocking political pushback and arguing cheap AI labor creates net new demand. |
Statements from this episode (15)
Haber: AI capabilities offer product differentiation, not long-term defensibility
“But the source, the AI-ness of that capability, in my opinion, is not a source of defensibility. It, it's largely differentiation. The defensibility of a software product resides, in my opinion, you know, from owning the end-to-end workflow, you know, from the…”
Rampell: AI lowers software creation costs, increasing early-stage competition
“Nobody can because it's so easy to actually produce software, and that's kind of the, that's the double-edged sword of AI, is that it's very, very easy to produce software. Everybody can go do something that is a very obvious idea, because it's obvious everybo…”
Rampell: Adobe and Zendesk will sell fewer software seats due to AI
“Is Adobe gonna sell as many seats if now you don't have to hire as many graphics designers? Or is Zendesk going to sell as many seats if the, Software just answers all the queries. Like, the answer is no.”
Haber: The marginal cost of software production is approaching zero
“The marginal cost of producing software is, you know, declining asymptotically towards zero”
Rampell: Back-office utility software incumbents are almost impossible to displace
“And there are some companies that operate in this Goldilocks zone of irrelevance, like these janitorial services, where even if you have nine million competitors, Like, they're just not gonna go anywhere, which is why, like, a lot of the strategy that we talk …”
Rampell: Greenfield software success requires patient founders avoiding locked-in buyers
“When you look at Greenfield opportunities, you need two things to be true. You need the entrepreneur to be very, very patient and say, I'm not going to try to sell to everybody who's, if I'm starting a net new payroll company, I'm not going to try to sell to G…”
Haber: AI founders are younger, more technical, and less domain-native
“Founders today are often younger and more technical than we've seen in, in prior generations. You know, and so they're less often native to the particular industry, but they're fluent in the tool set.”
Haber: Contingency law firms have an unlimited appetite for AI adoption
“In lots of areas of legal if you make your employee 50 times more efficient, you're eroding your billable hour. In their business, they operate on a contingency basis. Meaning, You know, they only get paid if they make, if they win. So there's no sort of limit…”
Rampell: Momentum is the only route to scale for copied AI startups
“Momentum has the highest chance of getting you to gravitational scale where you do have a moat. And if you don't do that, by contrast, you're just gonna get eaten alive.”
Rampell: Meta earns more quarterly profit today than 2010 annual revenue
“They have more profit every quarter today than they had revenue per year in 2010.”
Rampell: AI startups face forced consolidation due to unsustainable loss-leader pricing
“So I think that will probably play out the same way here, because you just can't have a market where you have everybody lost leading and nobody's big enough to get any kind of scale effects.”
Rampell: Non-state-of-the-art foundation model companies cannot build viable businesses
“The model company is the most cutthroat, because, like, unless you're, if you're state of the art minus, minus, minus, And you're trying to earn a living. It's just like that. That's just not going to work. So that game is super cutthroat.”
Rampell: Platform incumbents will copy successful startup features within five years
“These companies, if they get their act together, they will marshal a lot of resources to go compete with you. It might take them five years, but they will 100% do it. You have to backfill your feature with a product, and you have to have a moat for that produc…”
Rampell: AI is consensus, unlike past cloud and mobile shifts
“It is just so consensus, like cloud was not consensus, mobile was not consensus, and that's why the incumbents kind of screwed up.”
Rampell: Near-zero AI labor costs will unlock previously impractical enterprise tasks
“Now you're going to start hiring AI in all of these different areas that you just would never bother hiring a human for, because it's just like, you can't train the human, you can't find the human, and the human's too expensive.”