
The 7 Most Powerful Moats For AI Startups
Garry Tan (host), Harj Taggar (host), Diana Hu (host), Jared Friedman (host)
In this episode of Y Combinator, featuring Garry Tan and Harj Taggar, The 7 Most Powerful Moats For AI Startups explores seven Enduring Moats Every Ambitious AI Startup Must Master Fast The episode reframes Hamilton Helmer’s “7 Powers” as seven types of moats that still apply in the AI era, despite originating in a pre-AI, Web 2.0 world.
Seven Enduring Moats Every Ambitious AI Startup Must Master Fast
The episode reframes Hamilton Helmer’s “7 Powers” as seven types of moats that still apply in the AI era, despite originating in a pre-AI, Web 2.0 world.
The hosts argue that early-stage founders should obsess over speed and solving painful, concrete problems first, then worry about moats only after they’ve built something valuable to defend.
They walk through each moat—process power, cornered resources, switching costs, counter-positioning, branding, network effects, and scale economies—and illustrate how AI startups are already using them today.
Throughout, they emphasize that AI doesn’t eliminate moats; it changes where they appear (e.g., in workflows, data, evals, and pricing models) and often amplifies advantages for fast-moving, deeply embedded application-layer startups.
Key Takeaways
Solve a painful problem first; design moats later.
Founders shouldn’t reject ideas because they can’t see a long-term moat on day one. ...
Get the full analysis with uListen AI
Speed is the only real moat at the very beginning.
Early-stage AI startups like Cursor won by shipping useful features on one-day sprints, outpacing large companies bogged down by process. ...
Get the full analysis with uListen AI
Process power comes from hard, unglamorous engineering that’s tough to copy.
Mission-critical AI agents for banks, legal workflows, or KYC require years of edge-case handling, evals, and reliability work—far beyond a weekend demo. ...
Get the full analysis with uListen AI
Own unique data, workflows, or access to create cornered resources.
Deep relationships with regulated customers, private workflow data, custom evals, and even proprietary fine-tuned models become assets others can’t easily replicate. ...
Get the full analysis with uListen AI
Exploit and create switching costs through deep integration and memory.
AI startups that embed themselves into a customer’s bespoke workflows—or accumulate rich user memories and personalization—make it extremely painful to rip them out, even if an alternative is slightly better or cheaper.
Get the full analysis with uListen AI
Use counter-positioning and pricing to attack incumbents’ weak spots.
Legacy SaaS vendors tied to per-seat pricing and old engineering cultures struggle to ship agents that truly do the work. ...
Get the full analysis with uListen AI
Leverage data and eval flywheels and, where applicable, economies of scale.
More usage yields more data and eval feedback, improving AI agents and creating network effects (e. ...
Get the full analysis with uListen AI
Notable Quotes
“A moat is inherently a defensive thing, and you have to have something to defend.”
— Harj Taggar
“The early stages at the beginning, the only moat that startups have is really just speed.”
— Varun Mohan (quoted, Windsurf)
“The version you build in a hackathon isn’t useful to anyone.”
— Jared
“The cornered resource doesn’t have to be a diamond mine. It could be the diamond mine in your customers’ heads.”
— Garry Tan
“Don’t use these frameworks to count yourself out prematurely.”
— Jared
Questions Answered in This Episode
At what concrete milestone (revenue, users, depth of integration) should an AI startup formally start designing for moats rather than just shipping fast?
The episode reframes Hamilton Helmer’s “7 Powers” as seven types of moats that still apply in the AI era, despite originating in a pre-AI, Web 2. ...
Get the full analysis with uListen AI
How can a small AI startup realistically build defensible cornered resources when large model labs control the core frontier models and infrastructure?
The hosts argue that early-stage founders should obsess over speed and solving painful, concrete problems first, then worry about moats only after they’ve built something valuable to defend.
Get the full analysis with uListen AI
In which verticals are AI-native startups most likely to beat SaaS incumbents because of counter-positioning and new pricing models?
They walk through each moat—process power, cornered resources, switching costs, counter-positioning, branding, network effects, and scale economies—and illustrate how AI startups are already using them today.
Get the full analysis with uListen AI
How should founders balance collecting data for network effects with increasing privacy, security, and regulatory constraints around AI usage?
Throughout, they emphasize that AI doesn’t eliminate moats; it changes where they appear (e. ...
Get the full analysis with uListen AI
What specific signals show that your product is starting to develop real switching costs or process power, rather than being just another ‘ChatGPT wrapper’?
Get the full analysis with uListen AI
Transcript Preview
This idea of moats is so pervasive and so important.
It is interesting how moats have just become much more discussed by aspiring startup founders now than they were pre-AI.
What is going to prevent you from being basically subject to infinite competition?
Like, a moat is inherently a defensive thing, and you have to have something to defend (laughs) otherwise, like-
If you had nothing to defend.
Yeah. If you have nothing to defend, don't worry about your moat.
Welcome back to another episode of The Light Cone. Today, we're going to talk about moats. So, in your head, you might be thinking about barbarians storming your gate.
(laughs)
You've got this little startup, and you've got every other company out there who wants to come and eat your lunch. Uh, and you know, right outside your castle is a moat that keeps them away. Jared, when you were going to college campuses, this isn't sort of this trivial thing that people are thinking about. It's actually, uh, something that keeps them from starting companies right now.
Yeah, this is a question that we got from a lot of very smart college students on, on our, on our recent call ships. And basically, their question is like, they don't see how these new AI agent companies, like a lot of the ones that we've talked about on, on this podcast, could have moats. Um, it plays into this meme of, like, the ChatGPT wrapper that, like, all of these companies could be easily cloned. And so they can see how you could build a business that makes some amount of revenue, but they don't really see how you can build a long-enduring business. And so, I think it's actually not true. I actually think these businesses do have quite deep and interesting moats, but they're not totally obvious what they would be. So, I think this is an interesting topic for us to, to explore.
At our recent AI startup school, backstage, I had this exchange with Sam Altman that I thought was kind of funny. You know, we spend a lot of time thinking about, you know, make something people want, very simple maxims that are sort of anti-business school. And yet, this idea of moats is so pervasive and so important. We sort of remarked how funny it is that, uh, one of the more important books to read these days is actually business school fodder, um, this book called The 7 Powers. So, today, we thought that we would actually go through those seven powers. What are they? What are some concrete examples and ways that a startup founder who's just starting out, uh, could or should be thinking about these things from real-world examples that we've seen?
So, Diana, can you tell us a bit about this book?
This book was written by Hamilton Helmer, who taught at Stanford Economic School, and was published in 2016. And the book title was The 7 Powers: The Foundations of Business Strategies. And a lot of the examples are more with the era of, uh, internet companies from the 2000s. So, a lot of the examples are like Oracle, Facebook, Netflix, which is a older generation. So, we want to do a bit of a reboot right now, how it applies now, 2025, with AI.
Install uListen to search the full transcript and get AI-powered insights
Get Full TranscriptGet more from every podcast
AI summaries, searchable transcripts, and fact-checking. Free forever.
Add to Chrome