How Hamilton Helmer's 7 Powers Apply to AI Startups

1. 0:00 – 1:30

   The Moat Problem

   1. GTGarry Tan0:00

      This idea of moats is so pervasive and so important.

   2. HTHarj Taggar0:03

      It is interesting how moats have just become much more discussed by aspiring startup founders now than they were pre-AI.

   3. GTGarry Tan0:11

      What is going to prevent you from being basically subject to infinite competition?

   4. HTHarj Taggar0:16

      Like, a moat is inherently a defensive thing, and you have to have something to defend (laughs) otherwise, like-

   5. DHDiana Hu0:20

      If you had nothing to defend.

   6. HTHarj Taggar0:21

      Yeah. If you have nothing to defend, don't worry about your moat.

   7. GTGarry Tan0:30

      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.

   8. HTHarj Taggar0:40

      (laughs)

   9. GTGarry Tan0:40

      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.

   10. JFJared Friedman1:02

       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

2. 1:30 – 4:20

   The Seven Powers Framework

   1. JFJared Friedman1:30

      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.

   2. GTGarry Tan1:40

      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?

   3. JFJared Friedman2:26

      So, Diana, can you tell us a bit about this book?

   4. DHDiana Hu2:29

      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.

   5. JFJared Friedman3:01

      I think it's a little bit confusing the way he uses the terminology in the book. It's called The 7 Powers, but it would make a lot more sense if he just called the thing The Seven Moats, because that's really what he's talking about. He's really talking about seven categories of moats that a business can have. And I... It's true that the examples are out of date, but I think the framework is actually pretty timeless. Like, it turns out there's just only so many kinds of moats that a business can have, and they don't really change. And so, like, (laughs) e- even though the specific, like, versions of these moats are different in the AI agent world, like, the categories haven't changed.

   6. GTGarry Tan3:31

      Thankfully, we live in a world where there's markets, and there's free markets where there's lots and lots of competition. And these moats, in a lot of ways, are the only way, if you're running a business, you can sort of fight against all of the other people who might want to do exactly what you're doing. And, um, you know, famously, Peter Thiel talks about, uh, competition is for losers.

   7. JFJared Friedman3:54

      Yes.

   8. GTGarry Tan3:54

      And so the profound view there is that given infinite competition, what is going to prevent you from being basically subject to infinite competition? And then as a result, uh, you know, your margins, how much you can actually profit off of what you're selling goes down to zero. And what that means is, like, actually your business will die. And so, you know, having a moat is, uh, relatively existential eventually.

3. 4:20 – 8:40

   When to Think About Moats

   1. JFJared Friedman4:20

      You made a great point earlier, Garry, that, like, this is actually like, you kind of have to worry about this at the right time of, of a startup. Do you want to talk about, like, how, like, early stage founders should think about moats?

   2. GTGarry Tan4:31

      I mean, this is sort of why we generally tell people to go find a person with a real problem and then go solve that problem first. It's, um... What's funny about the world, uh, that's a little surprising is that you can go almost anywhere and find some pain point, some problem that could be solved with software, and especially with AI, that frankly just isn't being solved. And if that... And they're, they're so numerous and so severe that if you find that thing and solve it, you literally can mint a billion dollar or 10 billion or even 100 or hundreds of billions of dollars, uh, market cap business, and it's just lying in plain sight. That's really the first thing that people should do. Like, you should just find a problem and go solve it. And then along the way, you will probably, as you work with customers, as, as you build the product itself and engineer it and figure out what data you need for it and all of these things, like, you will stumble upon these seven powers.

   3. JFJared Friedman5:32

      Yeah, the moats come later.

   4. GTGarry Tan5:34

      Yeah.

   5. JFJared Friedman5:34

      Like, it would be, like, pretty dumb for somebody to decide not to work on a startup idea because they can't see what the long-term moats of that idea could be, right?

   6. HTHarj Taggar5:42

      It is interesting how moats have just become, um, much more discussed by aspiring startup founders now than they were pre-AI. It seems like the main reason for that, presumably, is just that big, the original ChatGPT wrapper meme and that the, the moat that most people are worried about is moat against the big model companies. And how, like, are you not going to get crushed by one of the big labs when they decide...... the product you're working on is really valuable and they want to own it too.

   7. DHDiana Hu6:09

      And I think Varun, uh, from Windsurf, who we hosted some time ago, he said it himself, "The early stages at the beginning, the only moat that startups have is really just speed." Once you pass that and build something that people want, then you figure out and go deeper into these type of moats that we're gonna discuss.

   8. JFJared Friedman6:31

      I really like Varun's point that the only moat is speed. That is not one of the seven powers in the book, but I think it probably should be.

   9. DHDiana Hu6:39

      I think it also comes with a lot of the essays from PD. Because one of the tenets really at the beginning is, yes, you're a big company, let's say OpenAI at this... OpenAI's the new Google. (laughs) It's like, sure, OpenAI or Anthropic could build all these features, let's say, with a Claude Code and then compete directly, let's say, with Cursor or et cetera. And for a startup like Cursor to really win, even in the beginning, is they had relentless execution because a larger company, like a Google or Anthropic, it, they just have a lot of, uh, more cruft that they need to do in order to ship a product. They just have all these product managers, all the operations, it needs to go through a PRD, some spec doc, and it takes mo- a lot more time to ship a feature, as opposed to Cursor. The incredible story about Cursor when we hosted Michael Twell to come talk to the batch, he was sharing how his product development cycle for shipping features and sprint cycles were one day.

   10. JFJared Friedman7:39

       One day. So one day sprint.

   11. DHDiana Hu7:42

       In the, at the beginning, during 2023, 2024 around era, they would start the, every day would restart the clock and try to ship things every day. I mean, that's, like, insane speed. Like, there's no big company that could ship something at that speed.

   12. JFJared Friedman7:59

       Mm-hmm.

   13. DHDiana Hu7:59

       At most, weeks, couple of weeks, and maybe the larger companies, I don't know, if you're Google, maybe, like, multiple months or sometimes years. (laughs) I mean, they had Google Bard or Gemini a long time ago. That took years to get out, right?

   14. HTHarj Taggar8:11

       I think Cursor and Windsurf are great examples of when you should start thinking about the moats because for the first few years, I don't think it really mattered that much. They just had to, like, they proved out that, hey, like, code gen is going to be a really valuable application of AI. The development environment is going to be very, very important to own. They, like, got rapid growth. And then it's only when they're at scale that, you know, like, they have to start thinking about how they're going to defend against, like, Claude Code or Codex or all the other things coming in. And sort of, like, the

4. 8:40 – 10:18

   Forward Deployed Engineering

   1. HTHarj Taggar8:40

      mental model that's really stuck with me is when we spoke to Bob McGrew a couple of weeks ago, um, and how... I think, Jared, you brought this up, actually, was that one way you could think about it is that sort of all of these startups are kind of forward deployed engineering teams, like, for (laughs) , for the labs, maybe? (laughs) And so, like, early on, actually, because this is all greenfield, we don't actually know what the valuable verticals and products to build are. So in a sense, you don't, you, step one is to figure out what that is. And it wasn't actually, even two years ago, it wasn't actually clear it was code gen or, um, the IDE. Once you figure that out and you find and you sort of struck gold, then you keep digging. That's when you have to probably see, at some point, you're going to get more competition because people are going to realize, oh, this is really valuable. There's lots of money to be made here. And then you have to start, like, defending, like, the treasure you found.

   2. GTGarry Tan9:24

      So I mean, all the things that we're about to cover, aside from speed, are sort of one to a billion, one to ten billion, one to a hundred billion, one to a trillion dollar sort of problems. And then, uh, the real stupid thing that people might do is watch this and look for this as a reason to not even get to one.

   3. HTHarj Taggar9:46

      Yes.

   4. GTGarry Tan9:46

      So that would be probably-

   5. JFJared Friedman9:47

      Or like try to use it to, like, pick between two different startup ideas because they're, like, trying to forecast five years in the future which one will have a greater moat.

   6. GTGarry Tan9:54

      Which just isn't how it works. I mean, literally you shouldn't do that.

   7. HTHarj Taggar9:58

      Like, a moat is inherently a defensive thing and you have to have something to defend. Otherwise, like... (laughs)

   8. DHDiana Hu10:03

      If you have nothing, then-

   9. HTHarj Taggar10:04

      Yeah, yeah, if you have nothing to defend, don't worry about your moat. (laughs)

   10. JFJared Friedman10:06

       Yeah. Otherwise, it's just like a puddle in a field.

   11. HTHarj Taggar10:09

       Yeah, exactly. Let's assume that someone has found something that's valuable that is worth defending. Should we talk through what some of the moats they could, they could think about are?

   12. JFJared Friedman10:17

       Yeah. So process power,

5. 10:18 – 14:34

   Process Power

   1. JFJared Friedman10:18

      again, like, the terminology is kind of funky but, like, basically it means you built something that's, like, you built a very complicated business with a lot of stuff that's just hard for people to replicate just because you, like, built all this stuff. Um, and so the example that he uses in his book is like the Toyota assembly line. And I think the AI version, the AI agent version of this, is just a really complicated AI agent that's been, like, finely honed over, like, multiple years to work really well under real world conditions. We've, we've talked about a bunch of these on this podcast, like Jay Keller with, um, Casetext is like the original example. A couple other ones I was thinking about from more recent companies, we have, like, a couple companies that sell AI agents to banks. We have Greenlight who worked with Tom. They do KYC for banks. And we have Casca which, like, does loan origination for banks, so it es- essentially tells banks, like, which loans they should give. And I think these are interesting examples because for all of these AI agents, you could build a version of Greenlight or Casca or Casetext, like a, like a demo version, in, like, a weekend hackathon. And I think when college students are thinking about these AI agents, I think what they have in their mind is, like, the weekend hackathon version of the product. And they're like, like, "I could build that in a week." Like, how could that be defensible? And, like, the reason is, like, the, the version you build in a hackathon isn't useful to anyone. It's, like, (laughs) like, like, like if Casca or, or Greenlight fail, like, the, the banks will lose millions of dollars. This is, like, mission critical infrastructure. It's, it's more like a self-driving car.

   2. GTGarry Tan11:47

      One way to look at it is way better engineering, uh, is actually of... That's, like, the most profound form of process power. Like, one example might be Plaid which-

   3. HTHarj Taggar11:57

      Yeah.

   4. GTGarry Tan11:57

      ... you know, the surface area of the number of-... uh, financial institutions that they have to support is so giant, it's, you know, probably thou- on the order of thousands to tens of thousands of different, different websites, different crawlers, and then all of the different, you know... Can you imagine, like, Plaid's CICD structure? And then, you know, this is pure speculation, but if I were, uh, Zach running Plaid, like, I, you know, know that I would want to be using code gen tool- the latest code gen tools to be able to, uh, you know, basically add every new financial institution on the planet quicker than anyone else. Like, that's sort of a very profound form of process power, uh, in the modern AI age.

   5. JFJared Friedman12:38

      I think this is probably the main form of defensibility for the existing SaaS companies. Like, if you look one generation before the AI agent, companies like, why is Stripe or Rippling or Gusto defensible? I think it's mostly this, right? It's just, like, they've just built a lot of software and it'd be really expensive and hard to replicate all of it. And, like, you can't just copy it from their landing page. Like, there's, like- like, the backend logic is, like, super deep.

   6. HTHarj Taggar12:59

      Also, I feel like kind of a schlep blindness aspect to this going on too, where, like, the- the hackathon version of any AI tool is, like, quicker than ever to get to. But actually, the last, like, 10% of getting it to work reliably across, like, tens of thousands of KYC requests, like, per day-

   7. JFJared Friedman13:17

      (laughs)

   8. HTHarj Taggar13:17

      ... is sort of, like, a particular type of painstaking drudgery work in a way (laughs) -

   9. JFJared Friedman13:23

      Yeah.

   10. HTHarj Taggar13:23

       ... um, that I think, like, lots of engineers are just not excited to do. And then that is also kind of, like, the teams at OpenAI are gonna experience this too, right? Like, if you're- if you're working in one of the big model labs and there's teams of people trying to invent AGI, um, it's going to be hard to get jazzed about nailing the, like, final 5% consistency on your, like, KYC tool.

   11. JFJared Friedman13:42

       KYC.

   12. HTHarj Taggar13:42

       Yeah. (laughs)

   13. JFJared Friedman13:42

       Yeah. And so I- I think this is especially true for, like, verticals, like KYC, that are- s- require specialized knowledge to even know what to- even to know what the edge cases are. Like, if we had to pick from the seven powers, like, I think speed and this, these are probably, like, the two dominant ones that come up the most often.

   14. GTGarry Tan14:00

       And those are most, uh-

   15. JFJared Friedman14:01

       Right.

   16. GTGarry Tan14:01

       ... related to execution.

   17. DHDiana Hu14:02

       Is where, uh, the hardcore builders win, having really good product taste and building the best product really matters. And I think it comes to a lot of the point, maybe the- the misconception is, I think a lot of these products, you can probably build the 80% solution with 20% of the effort. But for these solutions and products to work, you need the 99% accuracy one, which then takes, like, 10 times or even sometimes 100 times the amount of effort, right? Sort of that Pareto principle type of thing. What about, uh, the- the other power for

6. 14:34 – 19:30

   Cornered Resources

   1. DHDiana Hu14:34

      corner resources?

   2. GTGarry Tan14:35

      I think the classic view is they're just coveted assets or things that, uh, you know, they're not arbitrable, um, they must be independently valuable. And then, uh, sometimes they offer preferential access with b- you know, rates that are way lower. So, uh, the classic example that, you know, you could look at is, you know, pharma companies have these patents that are very hard to get. Um, they have to come up with them and then prove them and get through regulatory approval. And the sheer fact that they have a patent plus, you know, uh, getting through FDA approval is something that can be very durable. And it's, uh, you know, so powerful that patents have a, uh, limited lifespan because, you know, you don't want people to have that forever. A more modern example, I think, you know, on the regulatory side might be, you know, Scale AI is doing a ton of work with the DoD, um, you know, Palantir as well. Uh, in order to even get there, it's, you know, a painstaking process. You've h- you've got to hire the right people, you've got to spend a lot of time in DC and Langley or wherever, you- you know, you're trying to sell to. And, uh, you've got to literally build, um, uh, skiffs, like these, like, sort of, you know, special data, uh, centers where, you know, it- it's at great pain and expense, um, you have to get embedded with the government. But then when you do, like, well, you've got a... You know, the corner of resource, in some senses, even the brain space in people who work in the government. Like, you- th- right now, if you're working with AI, like, you've got to go through a Palantir or a Scale. And that's, like, literally written into their, uh, public documents around, like, how they're thinking about the nature of warfare and the nature of, uh, you know, everything that they want to do having to do with AI moving forward. So, you know, the corner of resource doesn't have to be a diamond mine, it could be the diamond mine in your customers' heads. Those examples are sort of, uh, closer to, like, being way up in the sky, having this, like, insane decacorn, like, worth hundreds of billions of dollars sort of situation. But what's relevant for startups that I think all of us, uh, sort of see every day is sort of what you were mentioning with, uh, this forward-deployed engineer, unit FDE, forward-deployed engineer model, that, uh, that is what a lot of startups that are extremely successful today are literally doing. Like, they're going out and getting a cornered resource in the form of real data and real workflows, um, literally sitting with a customer who normally would never get access to good software and then spotting, okay, uh, this is sort of the tailored time and motion. You know, first the, uh, you know, a request comes in by email. Then we take this and we enrich it in this way. Uh, sometimes we have to have a call center, call this person. Like, you know, actually understanding what, um, might be a very boring process, um, and then translating that into your own prompts, your own evals, eventually your own, uh, datasets to tune your own models. Like, those are all things that are incredibly valuable. And then, uh, clearly there are examples, you know, earlier, uh, we were saying, like, Character.AI, for instance, um, you know, took LLMs, you know, o- obviously built some of the first LLMs, then took many of them and then fine-tuned them in a way so that they could bring down the cost of, uh, serving those models by 10X. And so, you know, that itself is also a form of a cornered resource.

   3. HTHarj Taggar18:03

      The best cornered resource to have is your own model that can, like, do...

   4. JFJared Friedman18:07

      ... the specific work-

   5. GTGarry Tan18:08

      Yeah.

   6. JFJared Friedman18:08

      ... better, right?

   7. GTGarry Tan18:09

      And for a while, people thought that was the only moat that you could have in this space. If you didn't have your own model, like, you were totally hosed. Turns out, that's not true. Turns out, they're just one of the possible moats.

   8. JFJared Friedman18:18

      Partly, that is a threat people are worried about in, in the big picture, the 10,000-foot scary thing is if the labs at some point decide to treat their models as a cornered resource and they restrict access.

   9. GTGarry Tan18:29

      I guess the interesting thing right now is, like, it may well be true that the, you know, platonic ideal perfect manifestation of an AI system will require a lot of both, you know, maybe, uh, pre-training, post-training, RLHF. Like, just so many different things that you have to throw at it to get it to, like, ChatGPT level. But we're also so early in the revolution that, um, you know, even if just context engineering gets you 80 or 90% of the way there, that's plenty.

   10. JFJared Friedman19:02

       Yeah.

   11. GTGarry Tan19:02

       That's actually all people need to do for, like, the first two years of their startup almost always. You know, Cursor didn't start out by doing, you know, full-parameter fine tunes of GPT-5, which they probably have access to now. Um, they started just by making something people want. Y- you know, earlier we were saying, like, don't use these frameworks to count yourself out prematurely. (laughs) And this is a very profound version of that.

   12. DHDiana Hu19:28

       So, the third

7. 19:30 – 24:54

   Switching Costs

   1. DHDiana Hu19:30

      power we're gonna discuss is switching costs. That is, uh, the concept where you get a moat when your customers are kinda trapped because it becomes very expensive for them to find a other solution, even if the other solution might be, like, a little bit better. It's just very painful for them to switch financially or in terms of the operations, times our effort because they just have so much of it in the current solution. And examples that are given in the book are, um, like, databases like Oracle. When you have all of your system or record and ev- all your data in Oracle, it becomes incredibly hard to migrate. Like, database migration is something that people don't do. Other example given is, uh, Salesforce and... Because once you have all your customer records in Salesforce, you've built all these workflows, the UI, and it's just a lot to retrain a lot of your sales team to use, like, a new software. You need to, like, migrate all the data. And then at that point, for the company to switch to a new CRM is probably gonna take, I don't know, lose, like, a whole year of productivity or something, even if the new solution is a little bit better. I think how AI companies are building moat with this has to do with a version of what Gary mentioned with the forward-deploy engineer. We give an examples of this with Happy Robot or Salient, where they start with specific workflows that are very customized per company and they work, uh, with large enterprises. And part of it is actually with the forward-deploy engineer, they ha- may have actually very long pilot, pilot periods, which might last, like, six months to a year. But if they succeed, these convert into seven-figure contracts. And the reason why these pilots are so long is because they're very much building custom software for the specific operations in these companies. And the examples for, uh, Happy Robot, they got customers like DHL, where they went deep into integrating into a lot of the workflows for how all their logistic operations are done, which is very custom to the DHL operation. Or the example for Salient, who's building AI voice agents for the financial industry, they integrate with banks and a lot of the banks have very different workflows on how they do a lot of the loan conciliation, how do you do the debt recovery, how they do a lot of the fraud monitoring and risk and compliance. And it's all a little bit different because all these companies have built kinda internal tools and the whole part of, uh, bringing an AI company that builds these workflows, they build custom workflows then that work with them. But as a result, the trade-off is you do have very long pilot cycles, but the pot of gold is worth it because you end up with this big contract. And once you're in, you're kinda minted and the big enterprise is not gonna do another bake-off because it's gonna, it's gonna be a huge waste of time for them to, "Let's try the other, whatever, cool AI voice agent company." At that point, it's like, "We just wanna get the benefits." So, that's how these AI companies are winning.

   2. GTGarry Tan22:46

      I think it's, like, uh, once a moat and it's also, uh, in- it's interesting in the age of AI that, uh, simultaneously, you could how- see how AI brings down the cost of switching by a lot. And that's, you know, sort of another lever that a startup could use. Like, if you can write, um, use CodeGen to basically extract data out of old ossified systems or your competitors, then you, you know, there are things that might have really relied on switching costs that you could potentially bring it down to zero. (laughs)

   3. JFJared Friedman23:18

      Yeah, there's actually two different flavors of switching costs, right? There's the, the old school ones from the SaaS era, all the system of records like Salesforce, but also ATSs like, like Lever and Ashby, where, where the switching cost was the painfulness of migrating data from one system to another. And I agree with Gary, LLMs might significantly reduce those switching costs 'cause the LLMs can figure out how to, like, morph the data from the old schema into the new schema. You'd use browder au- like, use browser automation on both sides to, like, solve issues where, like, people don't let you export the data. But then there's this new form of switching costs that I think is pretty native to the AI era, like you're talking about, T- T- Tayana, which is, like, these, these, these lengthy onboarding processes that lead to, like, deep customizations of the logic of the agent, not just the data that didn't really exist in the SaaS era. Like, I guess you'd, like, customize your s- like, your Zendesk implementation a little bit, but, like, not that much.

   4. GTGarry Tan24:16

      Yeah. I mean, and then for AI companies on the consumer side, I mean, this is all very nascent but, like, I think memory is already becoming a bit of a switching cost for me.

   5. JFJared Friedman24:27

      Absolutely.

   6. GTGarry Tan24:27

      Like, it actually blew me away that Claude was so behind on memory, and then, you know, uh, my relationship with ChatGPT I feel like has evolved very significantly in the last year where I'm like, "Oh, I actually just generally... It seems to know, you know, what I'm into and what I care about." So, you know, that switching cost, I think, over time will only become greater and greater. And so personalization for consumer is actually a huge piece of that.

   7. DHDiana Hu24:53

      What about counter-positioning,

8. 24:54 – 31:24

   Counter Positioning

   1. DHDiana Hu24:54

      the other moat on the book?

   2. JFJared Friedman24:56

      The definition of counter-positioning is doing something that is difficult for the incumbent that you are competing with to copy because it would cannibalize th- their business. I think there's a couple of ways that this plays out. In every category, there is a Darwinian competition between the existing SaaS incumbents b- building their own AI agents and the new AI native companies building AI agents on top of the existing SaaS companies. So, like, for customer support, the existing SaaS incumbents, like Zendesk and, uh, Intercom and Front are all building their own AI agents. But then we have, like, a new wave of companies that grew up in the last couple years that are building AI agents that interface with, with the system. I think it's like... I don't know, this could be a topic of a whole, like, Light Cone episode, which, like, who will win in, in, in each of these fights? I think it's really interesting. Um-

   3. GTGarry Tan25:51

      Unstoppable force meets immovable object.

   4. JFJared Friedman25:54

      Yeah. (laughs) One way where this is playing out in the counter-positioning is that all, almost all these SaaS incumbents, their pricing model is they charge per seat, i.e. per employee. And this is, I think, a very big Achilles' heel that they have strategically, which is that if their AI agents do a good job and actually work, those companies will need fewer employees doing this work because they're... Like, the work will be automated by AI agents. And in a s- and in a simplistic way, that will just actually reduce... The more successful they are, the more they will reduce their revenue. My guess is, like, some of them will be able to navigate this, like, especially if they're still founder-controlled. I think, like, Intercom, for example, like, the l- I think the founder-controlled versions of these companies are smart enough to recognize that this is existential and they may be able to cannibalize themselves. I think the ones that are not founder-controlled, I don't have a lot of hope for. It's super hard to cannibalize your own revenue.

   5. HTHarj Taggar26:45

      The alternative, as we're seeing, is so much of the startups, um, pricing models are around sort of, like, work delivered or tasks completed. I think it's, it's exactly what you said, but it's also... That then switches the product towards having to actually be able to complete the work.

   6. GTGarry Tan26:59

      Yes.

   7. HTHarj Taggar26:59

      And, um, something I actually repeated at the last YC batch, um, at the end as closing advice is that I wish the founders in a batch could just somehow go spend a month at some of the late-stage companies.

   8. JFJared Friedman27:11

      (laughs)

   9. HTHarj Taggar27:12

      Um, uh, 'cause the top thing we hear from the founders running those companies is how hard a time they're having sort of re-setting engineering culture in their org to actually embrace AI, to use the tools, to want to do, like, context engineering and prompt engineering. And, and the, the net result of these teams not actually being able to be AI native, for want of a better term, uh, is that they just can't deliver the products that work, right?

   10. JFJared Friedman27:36

       (laughs)

   11. HTHarj Taggar27:36

       And so, like, they both don't want to switch from per-seat pricing because, like, that's what they're used to, um, uh, and in a world of AI being able to do the work, there's gonna be less seats to sell to. But they also just cannot deliver on products that can do the work. And so they, like, they wouldn't... That, that pricing model's not gonna make any sense for them either.

   12. JFJared Friedman27:54

       Yeah, it's, it's like the process engineering part. They're not good at the process engineering part for this new kind of engineering, yeah.

   13. GTGarry Tan27:59

       I mean, something sort of, uh, emerging that's very interesting in a bunch of YC startups like, uh, Avoka, for instance, they're doing customer support software kinda like ServiceTitan, but for, um, HVAC. So, uh, literally, like, people who help you with heating and, uh, air conditioning. And, uh, you know, I think ServiceTitan has something like 1% wallet share, 1% of the gross transaction value of, like, a given HVAC company, um, which is very small, right?

   14. JFJared Friedman28:27

       Very small.

   15. GTGarry Tan28:27

       I mean, people don't spend that much money on software because these are relatively low-margin service businesses. But the wild thing that Avoka discovered is that, you know, they can come in as software but then, over time, they're actually getting a bigger and bigger chunk of the wallet share because they can get the HVAC people to pay them, uh, actually for the customer support piece, which is not 1% of their spend but 4 to 10% of their spend. So, what you may well find is that, uh, this new breed of AI startup will actually have more growth, uh, and a higher wallet share. So, you know, actually, we may well be all, uh, undervaluing how powerful and how big the vertical SaaS, uh, AI companies will actually be because you're not, like, 1% of wallet share. You can get to 10.

   16. DHDiana Hu29:20

       That's what we talked in that episode where vertical AI SaaS agents will be 10 times, at least 10 times bigger than SaaS because this relate to your point, Garry, tapping into a whole different part of the spend of the companies is not the wallet of software where you're kinda... Uh, at this point, I suppose, is a bit of a, a finite budget, but is really new space with, with things that were not done possible and it was mostly workflows around... From people.

   17. GTGarry Tan29:49

       And I, you know, I know that people are, like, pretty sensitive about, uh, workforce displacement, but, you know, customer support for an HVAC services company is not a fun job.

   18. HTHarj Taggar30:00

       (laughs)

   19. GTGarry Tan30:00

       And you can tell because all of these customer support jobs actually have, like, 50, 80% annual attrition rates.

   20. JFJared Friedman30:07

       Mm-hmm.

   21. GTGarry Tan30:07

       Like, they're just such torturous, not fun jobs that, uh, the companies themselves and the call centers themselves spend almost all of their time trying to vet and bring in more people to work on these terrible jobs. And so when you have better software, what's sort of happening is that instead of, like f- y- people aren't losing their jobs, these people are quitting their jobs anyway because it's a terrible job. And then, if anything, uh, what Avoka has told me is that many of the people who were in those customer support, uh-... you know, sort of roles. Uh, now they're actually having more fun jobs-

   22. JFJared Friedman30:42

       Mm-hmm.

   23. GTGarry Tan30:42

       ... because instead of, like, managing a whole set of people who don't want to be there, uh, they're actually managing AI agents and then handling the interesting weird cases. The coolest part of it is, like, they actually can go in and sometimes alter the prompts and sometimes, you know, actually have an impa- a direct impact on, uh, both the experience of the customer but then also their own day-to-day. And that immediately is like a ten times more interesting job, like wrangling a bunch of AI agents and making, uh, the support process better and better over time. Like that's, you know, as knowledge work goes, like way more interesting than follow this script and read what the computer says.

   24. JFJared Friedman31:23

       So Harj, you

9. 31:24 – 34:00

   The Workforce Displacement Reality

   1. JFJared Friedman31:24

      y- y- you had a really interesting point about a second form of counter-positioning.

   2. HTHarj Taggar31:28

      The space has moved so quickly that in every vertical, um, or many verticals, there's sort of early on emerged one company that's seen as the early winner in the space. And often it's actually like the second movers, at least within the YC context. We have seen over and over again that like there's advantage to being the second mover in a space like Stripe came after, uh, Braintree and Authorize.net bunch of things and was able to like actually win by just building a better product. DoorDash came after GrubHub, Postmates, various other delivery services, and eventually went on to win. And so I think it's interesting to sort of just consider about if you're entering a vertical where it's already feels competitive or there are already, there's already team to be like a, a early winner in the space. How do you counter-position against them? One thing I think is really interesting here is Legora versus Harvey. Legora is obviously, uh, both in the legal AI space. Harvey is the early winner. The counter-positioning that I see from Legora is Harvey came in early and maybe got early sales, um, but focused a lot on fine-tuning, uh, sort of like their product differentiation when over time it's seen that that was actually probably not the right move. You wanted to actually focus in on the application layer and actually just sort of building a better product. And, and Legora has focused on that. That's what their branding and positioning is, and it seems to be working really well for them as a second mover into the space. A company that I've worked more closely with, GigaML, enter the customer service space and they're competing with Sierra and Decacon, like really well-known customer support companies and from having seen their sales motion, how they've been able to sign up some big customers, their, I think their counter-positioning is their product fundamentally just works better out of the box and as a result they can have a much faster sales and onboarding process. So it's like their counter-positioning is if you want to sort of get your customer support working as quickly as possible, um, you should go through like the GigaML onboarding process versus like the Decacon. And I think that's actually worked quite well for them.

   3. JFJared Friedman33:26

      Yeah, GigaML is an interesting example of how, to your point about like hyper displacement, it's clear that an AI agent can do this job not just as well as a human, but actually much better than a human. Like the DoorDashers that the GigaML agents are talking to, a lot of them don't speak very good English. They speak all kinds of languages. You can't hire a customer support person who's fluent in 200 languages. Um, but the-

   4. GTGarry Tan33:49

      But LMs are actually-

   5. JFJared Friedman33:50

      Yes. Yes.

   6. GTGarry Tan33:51

      ... out of box, out of the box.

   7. JFJared Friedman33:52

      Um, and they're infinitely patient if like there's a bad connection or so that's pretty interesting.

   8. DHDiana Hu33:58

      Think you have other example where

10. 34:00 – 37:30

    Brand & Speed as Moats

    1. DHDiana Hu34:00

       to your point of a superhuman abilities is where the AI version of the product actually works. I think Harji that you had the example of a Duolingo versus Speak.

    2. HTHarj Taggar34:11

       Duolingo is obviously the biggest language learning app, I think, um, most consumers know. The, the emerging criticism of it, I would say is that, um, what l- it's actually just sort of like a gaming app versus a language learning app that like, the way the app works is orthogonal to learning a true language. And then you have Speak, um, which is a, uses LLM, like uses voice to actually like help you practice and actually learn the language. Um, and that counter-positioning is working really well for them, right? And sort of Speak has got explosive growth and it's not trying to compete with Duolingo on the we're, we've got like lots of gamification and points and sort of like a great game mechanic, it's competing on, hey, we're actually just the place you should come if you want to learn the language by speaking it. I, I think the counter-positioning moat is very, um, sort of close and overlaps with the branding moat idea. I think in the book he talks about, you know, like brand is, it's essentially a moat when you become so well-known that even if you have an equivalent product, um, consumers will still choose you, um, because I think the brand effects. And I think the, the example used is like Coca-Cola. In the AI context, I think it's probably harder to apply brand as a moat directly to startups because it just takes time to acquire brand. Um, but you can certainly see its effects. Like the thing that still stuns me is OpenAI ChatGPT has more consumers using it per day than Google's Gemini. I think anyone who understands the models and uses them, um, daily would say that Gemini Pro 2.5 and Gemini Flash 2.5 are like equivalent models.

    3. JFJared Friedman35:42

       And Google also had all the users.

    4. HTHarj Taggar35:44

       Yeah. Yes. Yeah. (laughs)

    5. JFJared Friedman35:44

       Like basically everyone in the world is a user of Google, OpenAI had no users initially.

    6. HTHarj Taggar35:49

       Gu- Google was already one of the biggest consumer brands on the planet. It was almost certainly the biggest consumer brand on the internet. (laughs) Uh, and yet somebody else came along and built the brand as the consumer AI app. And G- Google is like playing catch up.

    7. JFJared Friedman36:04

       If someone had tried, had told me in 2022 that that's how it would play out, I would've been fairly incredulous.

    8. GTGarry Tan36:10

       It's also a perfect example of counter-positioning again. I mean this is, Google had a, uh, a business model that required it to continue to support ads and an organization that, uh, they shipped. And so you have the greatest cash cow in the history of man, so why would you disrupt it, um, even at the cost of setting back, uh, human access to knowledge by a few years? Even if that's like the core stated goal of Google itself to organize the world's information.

    9. DHDiana Hu36:41

       There's also the untold story of how, uh, the origin story of ChatGPT, how it came to be, which is really the original mode for startups with speed. It shipped very quickly in a matter of months with a very small team of a couple engineers.

    10. GTGarry Tan36:56

        I mean, it required, uh, you know, Sam Altman and YC Research and Greg Brockman to go, uh, hire Ilya Sutskever out of DeepMind (laughs) because he was there and you know, he, all the people, a lot of the people who went on to help create OpenAI, uh, they came from DeepMind, like it was already in the right place. It's just that that place didn't nurture exactly the thing that society really needed.

    11. DHDiana Hu37:22

        For speed.

    12. GTGarry Tan37:23

        So th- there's that mode again. Speed, number one.

    13. DHDiana Hu37:25

        Mm-hmm.

    14. JFJared Friedman37:26

        Do you want to talk about network economies, Diana?

    15. DHDiana Hu37:28

        Yeah. In the book, uh, network economy

11. 37:30 – 41:00

    Network Economies

    1. DHDiana Hu37:30

       s- is described as, uh, where the value of the product increases as more users or customer get and use the product. And everyone derives more value as an effect of more people using it. And examples that were given in the book are, uh, Facebook, where as you use it and your friends use it, it's more fun for me to use Facebook because all my friends are in there. As more users come in, then is- the social network becomes more valuable. And this was very much the era of, uh, the internet, where people talked about, uh, network effects that came to be. And the other example it gives is, like, Visa, the Visa network, where the more merchants are using Visa, the more value the consumer gets because you can swipe the Visa card more places. Then that becomes the- the moat because it's harder to then acquire and amass this number, a large number of, uh, users or merchants in order to- to win. So, that becomes very defensible. In the current era for AI, the shape of, uh, network effects is different. It really comes into the shape of data. I think a lot of, uh, the data that a lot of AI companies get access to becomes the moat, where the more data they get, the custom models they build become better. And the better models, it becomes a better product for users. And there's lots of examples of these. And, um, besides, like, the big foundation lab companies where they probably use some of the data. I don't- I mean, they probably use some of the data from the users. They probably do.

    2. GTGarry Tan39:04

       And ChatGPT almost certainly, like, feeds a lot of that back because you have a certain reward function for-

    3. DHDiana Hu39:11

       Right.

    4. GTGarry Tan39:11

       ... each training run, right?

    5. DHDiana Hu39:12

       So, all of the history of every chat from ChatGPT-1, 2, 3, 4, 5, now goes fed into GPT-6, and then so on and so forth. It helps create the- the next model version. And there's, uh, even smaller versions of this. For example, Cursor, they have probably one of the best, uh, tap-tap autocomplete because one, the- the free version of Cursor, they actually say it when you sign up that they- they will use the data, and they use that to train it. And the more users they get, they-

    6. JFJared Friedman39:42

       I think it's, like, all the data. Like, I think it's, like, quite literally, like, every mouse click and every keystroke that you u- that you emit when you're using Cursor, like, is fed into the model, which is, like, kind of crazy.

    7. DHDiana Hu39:55

       Which then, the more developers use Cursor, the better the product gets. And then they compound a lot of the- a lot of the wins with that. And the version where this applies to AI startup is when they go work with enterprises and large companies, they get access to private data. I mentioned earlier Seliom or Happy Robot. When the employees of the companies, where they become customers, as they use their product, they have a lot of that private data that makes a lot of the workflows better. And the way they improve that, which is the second way of having moats with networks, is really evals. We- we talked a lot about evals being the key moat for AI startups, is evals is where you get a lot of the... This workflow worked or didn't work, and then take that back and iterate and improve your context engineering. And that is a flywheel that you can only achieve when you get more and more usage of your product, whether it be in a consumer or a- or a AI vertical SaaS agent. So now, the last moat in the book is,

12. 41:00 – 43:44

    Scale Economies

    1. DHDiana Hu41:00

       uh, scale economies. Jared, do you want to tell us about it?

    2. JFJared Friedman41:03

       Scale economies or economies of scale, you've invested a lot of money to build something that's really big, and as a result, you have economies of scale, and you can offer the service cheaper than anybody else. So, like, the- the classic example would be, like, UPS or FedEx or the Amazon delivery network. They built, like, massive, like, physical infrastructure, and as a result, they have like a lower cost per unit, um, compared to a smaller competitor. Um, I think the way this has played out in the AI world, I don't think it's actually played out that much at the application layer. It's really played out at the model layer, right? Like, training a state-of-the-art LLM is very capital-intensive. Only a few companies can afford to do it. Once you've done it, you can afford to, like, let people do inference on that model very inexpensively. Th- this is why the DeepSeek announcement was so, um, was so earth-shattering last year because it seemed like it might be a lot cheaper than people previously thought to train a frontier LLM, which would greatly diminish the power of this, like, economies of scale moat that people thought the- the AI labs had.

    3. DHDiana Hu42:07

       The key thing about DeepSeek was they figure out and made public this new unlock for models, which is, uh, how to do RL. They still build on top of, uh, one of the large foundation models, so it's still expensive. The RL part is cheaper, but you still need the very expensive big foundation model. So, that's one of the things that media got wrong.

    4. JFJared Friedman42:25

       There's a separate question that people talk about, which is like, how will the foundation model companies be defensible against each other? And, like, this is certainly one way, right? It's just like, it's- it's very hard to be a new entrant into that game now because of this economies of scale. And we were- we were thinking earlier about, like, how this had played out with startups. And there's not that many examples, but I think a couple of good ones... Well, one- one good one is- is a company of yours, Exa. Harj, do you want to explain what- what Exa does?

    5. HTHarj Taggar42:50

       Yeah. Exa is essentially search for AI agents. Um, it provides an API for anyone building AI applications at once to search the web.

    6. JFJared Friedman42:58

       And the way I- I think this is playing out for Exa is in order to provide that service, they need to crawl the web, not the whole web like Google does, but a big chunk of it. And that's very expensive to do. It requires, like, a large, like, fixed capital, uh, investment. But then once- once you crawl a big chunk of the web, you can reuse that same crawl for- for many different customers.

    7. HTHarj Taggar43:17

       I think what's interesting about Exa, the- the parallel to their model companies is that they- they had invested in that, like, sort of before agents had really taken off. Like, they were fairly early to this. I think they were working on this actually even pre-ChatGPT launching. So, they made the investment early on, took a bet, same way that the lab companies took a bet on, like, transformers and, um, uh, and scaling laws.

    8. JFJared Friedman43:38

       Yeah. And there are two companies in just the most recent batch, uh, Channel 3 and Orange Slice, that are both doing

13. 43:44 – 45:05

    Final Advice

    1. JFJared Friedman43:44

       Exa.AI-like plays where they crawl a big chunk of the web, have a big, like, static crawl on their own servers, and then have agents that run on top of those, of- of that crawl. So, I think we're gonna see more and more of this, especially as the web agents work better.

    2. GTGarry Tan43:57

       You need to mainly focus on, uh, the first moat that isn't even in the book, which is speed. Like, you know, if you're really breaking your brain about like, "Oh, well, are we gonna be a cornered resource or not?" You're just thinking about it in the wrong way. Like, you should not start there. You should start with, "Do I have a specific person who has some sort of pain point?" And it's pretty painful. It's not like a, "Oh, it'd be nice if I could do this." It's a, "Oh, I am not going to get promoted this year. Maybe I will get fired." Like, this is so painful that I don't want to go to work today. Like, that's sort of the type of pain that you're looking for. And if you can write software or build things that actually alleviate that pain, like existential pain, like the business is going to go out of business, or, "Oh my God, we could totally take over everything next year." Like, that's sort of the feeling that you want in your customer. Uh, if you can find things like that, go- go zero... You know, go find that and go zero to one on that first. With that, see you guys next time.

Episode duration: 45:05