Why Old Startup Ideas Like Recruiting Now Work With LLMs

1. 0:00 – 0:41

   Intro

   1. HTHarj Taggar0:00

      There's all this, like, tooling and infrastructure still to build. There's clearly still a bunch of startups yet to be built in just the infrastructure space around deploying AI and using agents.

   2. JFJared Friedman0:12

      But if you're living at the edge of the future and you're exploring the latest technology, like, there's so many great startup ideas, you're very likely to just bump into one.

   3. GTGarry Tan0:19

      You apply the right prompts and the right dataset and a little bit of ingenuity, the right evals, a little bit of taste, and you can get, like, just magical output. Welcome back to another episode

2. 0:41 – 6:06

   What startup ideas could not work before AI?

   1. GTGarry Tan0:41

      of The Light Cone. Every other week, we're certainly realizing there's a new capability, a million-token context window in Gemini 2.5 Pro. It's just really insane right now, and the thing to take away from that, though, is that we have an incredible number of new startup ideas, some of which are actually very old, and they can only happen right now. Harj, what are some of the things you're seeing?

   2. HTHarj Taggar1:07

      Well, one thing I've been thinking a lot about recently is, what are types of startup ideas that couldn't work before AI or didn't work particularly well that are now able to work really, really well? Uh, and one idea that is very personal to me, um, would be recruiting startups, since I ran a recruiting startup, um, Triplebyte for almost five years. And I think, um, something that I've clearly seen is there was a period of time when we started Triplebyte, so around 2015, where recruiting startups were kind of like a really popular type of startup. Um, and I think a lot of the excitement around those ideas back then was this idea of applying marketplace models to recruiting, 'cause there were marketplaces for everything except how to hire great people, and specifically great engineers. And we started Triplebyte with the thesis of, you don't just want to let anyone on your marketplace. You want to build a really curated marketplace that evaluates engineers and gives people lots of data about who the best engineers are. And this was all pre-LLM, so we had to spend years essentially building our own software to do thousands of technical interviews, to squeeze out every little data point we could from a technical interview so that we'd effectively built up this labeled dataset that we could run machine learning models on. But we didn't even get to do that until, like, year three or four, um...

   3. GTGarry Tan2:21

      Yeah. And initially, it was, uh, actually a three-sided marketplace in that you needed to hire an interviewer in between to get that human signal.

   4. HTHarj Taggar2:29

      Yeah, we had- we had companies hiring engineers, we had the engineers looking for jobs, and then we had engineers we contracted to interview the engineers. (laughs) Um, so it was like, lots of things going on right now, um, and all of the evaluation piece of it at least, now with AI, is very, very possible. I mean, s- we can, specifically with the AI code gen models, you can do code evaluation, um, and I think probably one of the hot AI startups at the moment is this company called Merkle, which is essentially similar to the Triplebyte idea. I mean, it's a marketplace for hiring software engineers. Um, but I think what AI has unlocked for them is the evaluation piece of it, they could just do on day one using LLMs. They didn't need to build up this big labeled dataset. And they've been able to expand into other types of knowledge work, um, quite easily. For us to have gone from, like, engineers to analysts to all these other things would have taken years 'cause again, we had to rebuild the labeled dataset. Um, but with LLMs, you can just do that on, you know, day one effectively. And so I think this whole- this whole class of recruiting startups that are trying to evaluate humans at being good at specific tasks or not, um, is a really interesting space that's much more exciting to find good startup ideas in now than it was five years ago.

   5. GTGarry Tan3:44

      So that's a very powerful prompt for anyone listening. Uh, what are marketplaces that are three-sided or four-sided marketplaces that suddenly become, you know, two or three-sided or now there are two-sided marketplaces like Duolingo that are, you know, a little bit under fire because they're sort of starting to say, "Actually, maybe we're just going to use AI for, uh, the person that you're going to talk to in another language." That is totally a coherent thing that you could go to almost any marketplace in the world and say, "What if? What- what will, uh, LLMs do in that marketplace?"

   6. HTHarj Taggar4:19

      I think the other thing I really respect the Merkle founders for is there's also just a psychological element as a founder to when you enter into a space where there's, um, been lots of smart teams and lots of capital that's flown into it. This was definitely with recruiting startups, I mean, Triplebyte raised something like $50 million. Our main competitor, Hired, raised over $100 million. I think in aggregate, hundreds of millions of dollars went into funding recruiting marketplace companies, um, and overall as a category did not do particularly well. And so I think going in, you face a lot of skepticism if you're going to go out and pitch investors for an idea. Even when you have the, like, "Well, LLMs change everything," that pitch two years ago was still not as compelling as it is today, and so you have to be willing to sort of push through a lot of sort of cynicism and people who are burnt out or have lost lots of money on an idea to even kind of keep going to test it out and- and make it work.

   7. GTGarry Tan5:09

      That's something that repeats, actually, all the time. I mean, Instacart was that story exactly. Like, Webvan was sort of this rotting corpse of a startup just hanging in that doorway and most people looked at that and said, "Oh, man, I- I don't want to walk near that." Like, there's- there's going to be more, but you know, simultaneously the iPhone and, uh, Android phones were everywhere and you could have a mobile marketplace for the first time, and I guess that's why we're pretty excited about this moment because suddenly all, you know, the idea maze just moved, like, all of the walls to the idea maze have shifted around and, uh, the only way to find out is you've got to actually be in the maze.

   8. JFJared Friedman5:48

      It is very similar to Instacart and Webvan if we go back in history, right? 'Cause like the- the big new technology unlock for Instacart was the fact that everyone had a phone now enabled, like, the Webvan model to actually work for the first time and like, it's the same thing with LLMs in recruiting companies now, and- and a whole bunch of other ideas.

3. 6:06 – 7:35

   Technical screening products

   1. DHDiana Hu6:06

      I think it makes, uh, focusing even on specific parts of the marketplace...... to be great ideas to start with, even with this, uh, recruiting idea space. There's this company called Apriora that Niko, the other GP here at YC, funded back in winter '24. And their whole premise is to build AI agents that run the screening for technical interviews, where a lot of engineers spend a lot of time just doing a bunch of interviews, and the pass rate is so tiny. When I used to run engineering teams at Niantic, all that pre-screening was just so much work.

   2. GTGarry Tan6:42

      And the engineers hate doing it.

   3. DHDiana Hu6:43

      Yeah. And even that one piece, not exactly, let's say, marketplace, but what is the hardest part of it? And if you solve it right now, it works out. So Apriora actually does a pretty good job. It's being used by large companies, and it's been taking off.

   4. HTHarj Taggar6:58

      It's another example where you can actually expand the market, because I think the- there are plenty of technical screening products pre-Apriora, but you could only use them to do fairly simple evaluations to, like, weed out people who weren't engineers at all, um, effectively, um, or very, very junior. Um, but Apriora's product now with LLMs, you can do more sophisticated evaluations to kind of get more nuanced levels of screening. And so suddenly now, companies will be like, "Oh, actually I could actually give this to not just, like, my international applicants or my college students, I'll just give it to, like, senior engineers who are applying," which just opens up the opportunity.

   5. DHDiana Hu7:34

      So you were

4. 7:35 – 9:48

   Truly personalized education tools

   1. DHDiana Hu7:35

      talking a bit about education as well, Gary, about Duolingo. I think that's aspect of doing hyper-personalization is one of the holy grails where has been difficult for ed tech companies to crack, right? Because every student as they go through their learning journey, everyone is very unique and knows different things, and it sounds really cool to build, like, the awesome personal AI tutor that we did a- that Harsh did an RFS for, right?

   2. HTHarj Taggar8:00

      Yeah. The thing I'm excited about is for as long as I can remember, the internet's been around, like, one of the, like, um, dreams of it was that everyone now have access to, like, personalized learning and knowledge and, um, we'd all just, um, you know, have these, like, great intellectual tools to learn anything. And clearly the internet's made it easier to learn, but we've never had really truly personalized learning or a personalized tutor in your pocket idea, which is possible now for the first time. And I think we're definitely seeing smart teams applying to YC who are interested in building that type of product.

   3. DHDiana Hu8:32

      Couple companies that we funded that are kind of working out is, uh, this other company that also Niko funded called RevisionDojo that helps students do exam prep, and is sort of the version of, uh, flashcards but not like the janky, just, like, boring going through content, but the version that actually students like and gets tailored for their journey. And that one has, like, a lot of DAUs and a lot of power users, which is super interesting. And I think, Jared, you had worked with this other company called Edexia as well?

   4. JFJared Friedman9:05

      Yeah. Um, Edexia does tools for teachers to grade their assignments, which is another example of work that, like, is, like, not people's main job, but it's this other thing that they have to do, like engineers doing, like, recruiting, that they generally hate doing. There's, like, a lot of studies that show that, like, the biggest reason that teachers churn out of the workforce is that they hate grading assignments. It's just, like, no fun at all. And so Edexia, like, is an agent that's, like, very good at helping teachers to grade assignments.

   5. GTGarry Tan9:33

      Yeah, one of the interesting trends for some of this stuff is that, um, it's private schools who are actually much more nimble. And you know, I'd be curious what policy changes we need to make to actually support this in public schools, 'cause the public schools need it the most actually.

5. 9:48 – 14:41

   Do better products automatically get better distribution?

   1. HTHarj Taggar9:48

      I guess a question for you actually, Gary, I'm curious about this stuff is, it's clearly possible to build much better products with LLMs. If you take the ed- the learning apps, for example, they can go far beyond anything you could do for personalized learning pre-LLMs. Um, but i- i- it doesn't necessarily mean that you instantly get more distribution, especially if you're going after the consumer market. So do you- how do you think that plays out? Do better products automatically get more distribution, or will these startups have to work equally as hard to get distribution to be big companies as before?

   2. GTGarry Tan10:21

      I guess one of the more awkward things that's still true is that, um, you know, intelligence is much cheaper. It's quite a bit cheaper than it was last year, but it's still enough that you have to charge for it probably. Um, but that's something I would probably track. I mean, it seems clear that, um, you know, distillation from bigger models to smaller models is working. It seems clear that the mega giant models are teaching even the production model size of today to be smarter. The cost of intelligence is coming down quite significantly, so, you know, I know that we tease this sort of, uh, almost every other episode, but, like, consumer AI, it finally might be here soon. Uh, and I think the thing to track is, well, how smart is it such that, like, any given user incrementally only costs, I don't know, pennies or, like, 10 or 15 cents? Like, then it becomes so cheap that you will just have intelligence for free. Maybe it'll be a return to the freemium model that we got used to during Web 2.0, this idea that you could basically give away your product and then for, um, you know, 5 or 10% of those users, there are things that they so want that, you know, y- you're going to sell them a $5 or $10 or $20 a month subscription.

   3. HTHarj Taggar11:41

      That's basically what OpenAI is doing, right?

   4. GTGarry Tan11:43

      Yeah.

   5. HTHarj Taggar11:43

      Like, that is, like, their business model.

   6. GTGarry Tan11:44

      Perplexity does it, OpenAI, uh, you know, going back to education, Studee with two Ds, they're doing it, and they're seeing, uh, a lot of success. I mean, on average, the kids who use that actually get on grade level or, you know, kind of go up, uh, even a couple grade levels. Those are real outcomes for students. So, you know, right now you still got to pay for it, but, uh, maybe not for a while. And that's actually a really big unlock, you know. That- that's the moment where you could have 100 million or a billion people using it. OpenAI, uh, might be furthest ahead with it, but the hope is that, you know, really thousands of apps like this start in- start coming out, um, across all the different things you'll need. And that's something that-I know we keep saying it. Like, it's going to happen.

   7. DHDiana Hu12:32

      I mean, it's kind of happening already for edtech. Speak, it's this company that got started a couple of years ago before LLMs were a thing at all. It was a team of researchers that really believed that you could personali- personalize language learning, which might have been a bit contrarian back then because Duolingo seemed to be the game in town that was winning, and they really focused on really personalizing that whole language learning, and they got, they started taking off in Korea for a lot of, uh, learners that were trying to learn English. And when GPT-3 and 3.5, they were early adopters of it, started coming out, they saw that, "Wow, this is gonna be the moment." They doubled down and they've been on, on this trajectory now with lots of MAUs, DAUs that's really working out.

   8. HTHarj Taggar13:21

      I think one thing, going back to the consumer thing that we haven't talked as much about, um, we've seen a lot with the startups that are selling to enterprises or companies about how the budgets become so much bigger when companies are willing... When companies stop thinking about you as software as a service, but they start thinking about you as replacing their customer support team or their analytics team or something like that, they'll just pay way, way more. So the same thing will apply in consumer, right? Like if you think about a personalized learning app, uh, often edtech companies struggle with who's actually the buyer and who's going to pay. And if you have like younger children, for example, it's like you've got to get the parents to pay. But the parents aren't gonna pay that much for an app that their kids, like, don't retain or complete, like some sort of online course that they're disengaged with. But we know that parents will definitely pay for, like, human tutors and, like, you know, that's, like, actually probably quite a big market. And so if your app goes from being like a self-study course that doesn't get any completion to actually being on par with the best human math tutor for your 12-year-old, parents will pay a lot more for that. And so those, like, it's possible that, like, the product now just become, has a business model that you didn't have before, and that alone means you don't necessarily need millions of parents using it, but even 100,000 parents using it, paying you a significant amount means you now have like a much bigger business than was possible before.

   9. GTGarry Tan14:39

      Yeah. I feel like we have to talk

6. 14:41 – 16:08

   Moats

   1. GTGarry Tan14:41

      about moats a little bit. I mean, it's pretty clear a company like Speak or almost any of these other companies that could have durable revenue streams, like what you need is brand, you need switching costs, sometimes it's integration with other, uh, technologies that are sort of surrounding that experience. Like in, uh, a school, it'd probably be being connected to Clever, for instance, like login is au- authentication is pretty obvious. So yeah, I feel like Sam Altman has talked about this a bunch. You know, it's, uh, it's not enough to drop AI in it. You know, you still have to actually build a business. I don't think OpenAI is necessarily, uh, you know, out to get all the startups. Like I actually think that on the API side, they very much hope that a lot of them do really, really well, and certainly we want that too.

   2. HTHarj Taggar15:26

      They did just hire, like, the Instacart CEO as their CEO of application.

   3. GTGarry Tan15:30

      That's right.

   4. HTHarj Taggar15:30

      So it does kind of seem like they are definitely paying more attention to the application layer.

   5. GTGarry Tan15:35

      That's right. Yeah, I mean, you'd be crazy not to, right? Like, by all accounts, OpenAI is highly likely to be a trillion dollar company at some point and, uh, you know, as powerful as a Google or an Apple or, um, any of them. The interesting thing right now is like they're still on the come up and then if anything, um, the big tech platforms are actually still holding back a lot of the AI labs. And the most profound example of this is uh, why is Siri still so dumb?

   6. HTHarj Taggar16:04

      (laughs)

   7. GTGarry Tan16:05

      It makes no sense, right?

   8. HTHarj Taggar16:07

      Totally.

   9. GTGarry Tan16:07

      Uh,

7. 16:08 – 17:40

   The need for platform neutrality

   1. GTGarry Tan16:08

      I mean, I think that points to something that we actually really need in, uh, tech today. We actually really need platform neutrality. So in the same way, you know, 20, 30 years ago there were all these fights about net neutrality, this idea that there should be one internet, that ISPs or big companies should not self-preference, uh, their own content or the content of their partners. Uh, you know, that's what sort of unleashed this giant wave of really a free market on the internet. The other profound example of that is actually Windows. If, uh, you know, if you open up Windows, you actually have to choose your browser and then you also need to be able to choose which search engine you use. And these are things that, you know, the government did get involved in and said, "Hey, you, you know, you cannot self-preference in this way." And you know, if you remember the moment where, where Internet Explorer had a majority of web users, like that could have been a moment where Google couldn't have become what it became. So we actually have a history of the government coming in and saying, "This should be a free market," and for that free market to create, uh, choice and then therefore prosperity and abundance. And so I would argue like, you know, why doesn't this exist for voice on, uh, phones? Like you should be able to pick, not... You shouldn't be forced to use Google Assistant, you shouldn't be forced to use Siri. You should be allowed to pick and, you know, it's been many, many years of having to use a very, very dumb

8. 17:40 – 23:24

   Big Tech and AI

   1. GTGarry Tan17:40

      Siri.

   2. HTHarj Taggar17:40

      On the moat topic, something I just find fascinating is I saw some numbers recently about how, um, Gemini Pro models, like just their usage, particularly from consumers, is just an insignificant fraction of ChatGPT's. And I think at- at YC, we've been doing our own internal work building agents and, um, actually being at the cutting edge of a lot of the AI tools, and we found that Gemini 2.5 Pro is like as good and in some cases a better model than 03 for various tasks. That hasn't trickled down into...... public awareness yet, right?

   3. JFJared Friedman18:12

      Which is fascinating since Google already has all the users-

   4. HTHarj Taggar18:14

      Right. (laughs)

   5. JFJared Friedman18:15

      ... with their, with their phones.

   6. HTHarj Taggar18:16

      And I don't think anyone would say OpenAI is not a startup anymore, but relative to Google, it-

   7. JFJared Friedman18:21

      Yeah.

   8. HTHarj Taggar18:21

      ... essentially is. So there is clearly some sort of intangible moat around being the first in a space and sort of staking your claim as, like, the best product for a specific use case. And I feel like some-

   9. GTGarry Tan18:32

      And actually making it good.

   10. HTHarj Taggar18:34

       Yeah. Yep, yep, yep. But at some point, maybe it doesn't even necessarily need to be, like, objectively the best, it just needs to be good enough.

   11. GTGarry Tan18:41

       I mean, that's the bet that I think a lot of the big tech companies are trying and failing at. I mean, there's... Microsoft has a copilot built into Windows now that is still quite inferior to anything OpenAI puts out. Gemini itself is very, very good and I use it quite a lot. Um, it's probably, I don't know, 40% of my agent, you know, sort of if I need to especially summarize YouTube videos, it's very-

   12. HTHarj Taggar19:06

       Mmm.

   13. GTGarry Tan19:06

       ... very good at that.

   14. DHDiana Hu19:07

       For multimodal, it's really good.

   15. GTGarry Tan19:09

       Yeah. A lot of the Gemini integrations into Gmail or Goo- you know, Google Drive are not good.

   16. JFJared Friedman19:15

       They're totally useless.

   17. HTHarj Taggar19:16

       Very bad.

   18. JFJared Friedman19:16

       Yeah.

   19. GTGarry Tan19:17

       Yeah.

   20. HTHarj Taggar19:17

       (laughs)

   21. JFJared Friedman19:17

       (laughs)

   22. GTGarry Tan19:17

       It's like, "Is there someone at the wheel over there? I don't get it," you know?

   23. DHDiana Hu19:20

       I mean, I think that's even confusing for us, is even using it as a developer, there's actually two different products.

   24. GTGarry Tan19:27

       Yeah.

   25. DHDiana Hu19:27

       There's Gemini, where you can consume Gemini, and Vertex Gemini, and I think they're like different orgs.

   26. GTGarry Tan19:34

       Yeah.

   27. DHDiana Hu19:34

       I think it's suffering a little bit from being too big of a company, and essentially shipping the org. There's, like, these two APIs you can consume to use Gemini, and we're like, "Why two?" One is from DeepMind and the other one is from GCP.

   28. GTGarry Tan19:48

       I think that comes from the culture of Google, though. I mean, there's definitely this sense that, uh, if two orgs are competing and fighting, normally, in a normal org, you go up... And in a functioning, uh, startup for instance, you know, it goes up to some level, and then ultimately the CEO or founders, and then they just say, "Okay, well, I see the points over here, I see the points over there, we're going this way." But, you know, having lots of friends from Google, it doesn't seem like that's the culture there. Like, there's a layer of VP and sort of management that is actually like, "You guys just fight it out," and so then you ship the org.

   29. DHDiana Hu20:21

       I think the crazy thing about Google, they probably should've won a lot of the experience of the best model. There's almost like... I don't know where all this Game of Throne analogy could be used, they might be a little bit like Daenerys Targaryen-

   30. GTGarry Tan20:37

       Mmm.
9. 23:24 – 25:14

   AI horseless carriages

   1. JFJared Friedman23:24

      Our partner, Pete Kumhin, wrote this really great essay where he talked about, um, the Gemini integration with Gmail, and he really, like, broke down in great detail why Google built this integration all wrong, and how they should have built it. Um, it's almost like he was a PM at Google. Oh wait, he-

   2. HTHarj Taggar23:42

      Yeah, he was a PM. (laughs)

   3. GTGarry Tan23:43

      Yeah (laughs) .

   4. JFJared Friedman23:43

      He was a PM at Google. Um... (laughs)

   5. GTGarry Tan23:46

      It was very profound, um, in that one of the things he pointed out was that, you know, you have a system prompt and a user prompt, and if you are actually going to empower your users, you actually allow your user to change the system prompt, which is the part that normally is like above... You know, to use the Venkatesh Rao's idea of like sort of the, the API line. It's sort of like the system prompt is actually what is exerted upon... It's like sort of imposed upon the user.And so, you know, Gemini follows this very specific thing. Uh, I think the example is, uh, actually an, an email saying that, uh, Pete's gonna be sick, to me. (laughs) And he's like, uh, "Sorry, I'm not going to be able to come in," and, uh, he asks the agent to write this letter, and it's very formal, and of course it is because there's no way to change the tone. It's actually one of the best blog posts in that I think he had to vibe code the blog post itself 'cause you can actually try the prompts yourself on that web page.

   6. JFJared Friedman24:50

      Yeah, it's just s- super cool. It's like a, it's in this, like, interactive-

   7. GTGarry Tan24:54

      Yeah.

   8. JFJared Friedman24:55

      ... templating language.

   9. GTGarry Tan24:56

      Which made me think, it's time to start an AI first vibe coding, uh, blog platform.

   10. JFJared Friedman25:02

       Oh, like a-

   11. GTGarry Tan25:02

       Ooh.

   12. JFJared Friedman25:02

       ... AI, like a, like a AI Posterous?

   13. GTGarry Tan25:04

       Yeah, basically.

   14. HTHarj Taggar25:05

       (laughs)

   15. JFJared Friedman25:05

       This is, is the time for Posterous 2.0?

   16. GTGarry Tan25:06

       Yeah, it might be, it might be. Yeah, with all my extra time, that's what I'm gonna work on, but that's a free idea for anyone who's watching. (laughs) We'll fund it.

10. 25:14 – 30:03

    Gross margins

    1. JFJared Friedman25:14

       There's another class of startup ideas that I'm, I'm particularly e- excited about that I think are, like, perhaps the time is now, which is, um, do you guys remember the tech-enabled services wave?

    2. HTHarj Taggar25:26

       Yep.

    3. JFJared Friedman25:27

       Yeah, so for folks who, who, who, uh, didn't follow this, in the, in the 2010s, there was this huge boom in companies called tech-enabled services. Um, Triplebyte was one, actually.

    4. HTHarj Taggar25:38

       Yep, yeah, it very much.

    5. JFJared Friedman25:38

       Yeah, that was, like, tech-enabled services for recruiting.

    6. HTHarj Taggar25:41

       Yep.

    7. JFJared Friedman25:41

       Right? Um, we also had Atrium, which was tech-enabled services for law firms.

    8. GTGarry Tan25:46

       Law.

    9. HTHarj Taggar25:46

       Um, it started with Balaji's blog post about full stack startups, if you remember. Like, the concept was just that, um, software eats the world, means software just kind of goes into the real world. And so this is not a, the success example, but an example of it was, hey, like, instead of just having an app to deliver food, you should also, like, have a kitchen that cooks the food and software to optimize the kitchen, and you just do everything, um, and that, like, the full stack startups in theory would be more valuable than just the software startups because they would do everything.

    10. JFJared Friedman26:17

        Yeah, 'cause instead of just selling, like, software to, like, the restaurants and capturing, like, 10%, you could just own the restaurant and you could capture 100%.

    11. HTHarj Taggar26:24

        This is exactly what Triplebyte was 'cause we, we were like, "We're gonna be a recruiting agency, effectively. We're not selling software to a recruiting agency. We're actually just doing the whole thing." Like, we're gonna... We also had recruiters on staff that were just there to help people negotiate salaries and match them to the right companies, and, um, yeah, it was very much in that wave of do everything.

    12. JFJared Friedman26:41

        Yeah.

    13. GTGarry Tan26:41

        But the-

    14. JFJared Friedman26:42

        Yeah.

    15. GTGarry Tan26:42

        ... that wave of startups generally forgot-

    16. JFJared Friedman26:44

        Yeah.

    17. GTGarry Tan26:44

        ... that you need gross margins.

    18. JFJared Friedman26:45

        Yeah, what happened, like, I mean, like, fast-forward, basically the short version is, like, it didn't really work, and the full stack startups actually were not more valuable than the SaaS companies, and the SaaS companies sort of, like, won that round of the, like, Darwinian competition of different business models.

    19. HTHarj Taggar26:59

        I think fundamentally it's just what Garry says, is just they were actually not great gross margin businesses, but it was actually, I think what it, like, it was just hard to scale them. At least in Triplebyte's-

    20. JFJared Friedman27:11

        Yeah.

    21. HTHarj Taggar27:11

        ... situation, we actually got to a $20 million annual run rate, $24 million run rate within a few years for, like, if you compared us to, like, a regular recruiting agency, it was, like, super fast, but, um, if you were c- uh, compared us to, like, the top software startups, not that, like, um, impressive.

    22. JFJared Friedman27:29

        And it became harder and harder to scale-

    23. HTHarj Taggar27:30

        Yeah.

    24. JFJared Friedman27:30

        ... because you had more and more people.

    25. HTHarj Taggar27:32

        Yep. Yeah, basically, like, the margins didn't work out particularly well, and so they needed to keep raising more capital. And so if you were, like, a fearsomely good fundraiser, you could sort of do it and kind of push yourself, but even in those cases, I think most of those businesses, at some point, it just caught up with them. Like, at some point, like, actually we have to figure out a way to scale the business and have good margins and make this, like, profitable and not just rely on the next fundraising round, is what I felt hurt a lot of the...

    26. GTGarry Tan27:57

        I mean, you could argue Zenefits was one of those for, um-

    27. HTHarj Taggar28:00

        Yeah.

    28. GTGarry Tan28:00

        ... insurance and a bunch of different HR-related things. It was actually, um, they basically relied too much on hiring more salespeople and more customer success people instead of actually building software that then would create gross margin. And so Parker Conrad said, "Well, I'm not gonna do that again, and I'm also going to force all the engineers to do the customer support so that they go on to build software that doesn't require so much support." And thus there is gross margin, and that, you know, was a whole lesson that, uh, I feel like the whole tech community learned collectively through the 2010s. If we learned one thing, it's gross margin matters a lot. Like-

    29. JFJared Friedman28:39

        Totally.

    30. GTGarry Tan28:40

        ... you can, you cannot and should not sell $20 bills for $10 'cause you're gonna lose everything.
11. 30:03 – 32:30

    Full stack companies

    1. JFJared Friedman30:03

       What I've been excited about recently is, like, I think you could make a bull case that, like, now is the time to build these full stack-

    2. HTHarj Taggar30:10

       Yeah.

    3. JFJared Friedman30:10

       ... companies because, like, you know, like you were saying, like, the Triplebyte 2.0s won't have to hire this huge ops team and have bad gross margins. They'll just have agents that do all the work. And so, like, now actually, like, full stack companies can look like software companies under the hood for the first time.

    4. HTHarj Taggar30:24

       And you gave a great example. So, Atrium, f- started by Justin Tan, full stack law firm, didn't work out for all, I think, a lot of these same reasons. Um, but now YC-

    5. GTGarry Tan30:32

       I've heard him say that before. It's like, "Look, we went in trying to use AI to automate large parts of it, and it wasn't... The AI was not good enough at that moment."

    6. HTHarj Taggar30:41

       Sure.

    7. GTGarry Tan30:41

       "But it's good enough now."

    8. HTHarj Taggar30:43

       If you look at wh- within YC, we have Legora, which is, like, this, like-... one of the fastest growing companies we've ever funded. Um, and it's not building a law firm, but they're essentially, um, you know, building AI tools for lawyers. But you can see where that's going to extend out to, right? (laughs) Like, eventually their agents are just going to do all of the legal work and it'll, they'll be the biggest law firm on the planet. Um, and yeah, I think that's a kind of full stack startup that just wasn't possible pre-LLM.

    9. GTGarry Tan31:08

       I think this started right when Uber and Lyft and Instacart and all of these companies were happening. And the thing is now, I mean, you can actually have LLMs do a lot of the knowledge work and then, I mean, increasingly it, it could actually have memory. I mean, this is one of the RFSs. It's literally you can have virtual assistants, um, but they become less and less virtual if they can also, uh, hire real people to do things for you. (laughs)

    10. JFJared Friedman31:35

        Virtual assistant marketplaces was definitely like a whole category of companies for like 15 years in- including exact where you build like a marketplace of like people in the Philippines and like other, other countries and then you like get exposed to sort of like Airbnb UI. I don't think any of them ever like really got... really became like amazing businesses though.

    11. HTHarj Taggar31:54

        Going back to Pete's post, I think the other thing that's interesting about the, um, the points he made around sort of the system prompt and the user prompt and maybe we want to expose the system prompt to users a little bit more, um, it's an example of just how we're still so early in just using AIs and building agents. There's all this like tooling and infrastructure, like, still to build. You have to do evals, you have to run the models, like a whole bunch of stuff to build still. And so there's clearly still a bunch of startups yet to be built in just the infrastructure space around, you know, deploying AI and using agents.

12. 32:30 – 37:14

    ML ops

    1. HTHarj Taggar32:30

       And Jared, you know, it's interesting, something that struck me about when I first came back to YC in 2020 is I remember a class of idea we weren't interested in funding was anything in the world of like ML, machine learning operations or...

    2. JFJared Friedman32:44

       Yeah.

    3. HTHarj Taggar32:44

       ML tools.

    4. JFJared Friedman32:45

       Yeah.

    5. HTHarj Taggar32:45

       And I remember reading some applications and people were like, "Oh, like another ML ops like team, like these sort of never go anywhere." Um, clearly if you were working on ML ops in 2020 and you just stuck it out for a few years- (laughs)

    6. JFJared Friedman32:58

       (laughs)

    7. HTHarj Taggar32:58

       ... you're in the right spot. Uh, h- h- any context you can share from that period?

    8. JFJared Friedman33:02

       Yeah. I remember I got so frustrated after years and years of funding these ML ops companies with really smart, really like optimistic founders that just like didn't go anywhere that I ran a query to Qount. And I remember finding that, I think this was around 2019, we had more applications in 2019 for companies building ML tooling than we had applications for like the customers of those companies.

    9. HTHarj Taggar33:30

       (laughs)

    10. JFJared Friedman33:30

        Like, like anyone who's like applying ML to like any sort of product at all. And like, I think that was the core problem is that like these people were building ML tooling, but there was no one to sell it to because like the ML didn't actually work. So there just wasn't anything useful that you could build with, with, with all this ML tooling.

    11. GTGarry Tan33:46

        People didn't want it yet.

    12. JFJared Friedman33:47

        Yeah.

    13. GTGarry Tan33:47

        I mean, directionally, it was absolutely correct.

    14. JFJared Friedman33:49

        100%.

    15. GTGarry Tan33:49

        Like from a sci-fi level on a ten-year basis, it was beyond correct.

    16. JFJared Friedman33:53

        Yes.

    17. DHDiana Hu33:55

        I mean, you-

    18. GTGarry Tan33:55

        It's just wrong for that moment.

    19. JFJared Friedman33:56

        Yeah.

    20. DHDiana Hu33:56

        You actually have a team that stuck it out. I mean, part of the lesson is sometimes it will take a bit of time for technology to catch up. And this company called Replicate that you worked with stuck it out. It was from that era.

    21. JFJared Friedman34:08

        Yeah. Replicate was from winter '20, and they started the company right before COVID. And during the pandemic, it was going so poorly that they actually stopped working on it for several months and just like didn't work on it because like it wasn't clear that the thing like had a future at all. And then they picked it back up and just started like working on it quietly. But it basically was just like they were just building this thing in obscurity for two years until the image diffusion models came out and then it just like exploded like overnight.

    22. DHDiana Hu34:36

        Ollama is another good example that-

    23. JFJared Friedman34:38

        Oh, yeah.

    24. DHDiana Hu34:38

        ... you worked with.

    25. JFJared Friedman34:38

        Do you want to talk about O- Ollama?

    26. DHDiana Hu34:40

        So the Ollama folks were also from that pandemic era and similar story to, uh, to Replicate. They were kind of trying to do different things around here too, and they were trying to work it out to make open source models deploy a lot better. And they were also quietly working on it for a while. Things weren't really taking off. And then suddenly, I think the moment for them was when LLaMA got released. That was like the easiest way for any developer to run open source models locally and it took off because suddenly the interest to run models locally just took off when things started to work. But not before that, because there were all these other open source models, um, that were in Hugging Face. And especially the ones from like BERT models, those were like the more used deep learning models. They were like just okay, but not many people were using them because they weren't quite working.

    27. GTGarry Tan35:34

        What's the moral of the story? I mean, some of it is like, uh, be on top of the oil well before the oil s- starts shooting out of the ground. But is that actionable?

    28. HTHarj Taggar35:44

        It's kind of the classic startup advice of follow your own curiosity.

    29. GTGarry Tan35:47

        Mm-hmm.

    30. HTHarj Taggar35:47

        Like most of these teams, or almost all of these teams, were working on it because they were just interested in ML. They wanted to deploy models. They were frustrated with the tooling, probably weren't necessarily commercially minded and trying to pick the best startup idea they could possibly work on. But I know sometimes you get lucky, sometimes-
13. 37:14 – 40:19

    Updated startup advice for the AI age

    1. JFJared Friedman37:14

       is maybe a meta point on this whole conversation. So, um, we were at colleges. Uh, Diana and I w- went on this college tour, um, and we spent several weeks speaking to college students. And I realized that there's this piece of startup advice that became canon that I think is outdated. Back in the pre-AI era, it was really hard to come up with good new startup ideas because the be- like, the idea space had been picked over for, like, 20 years. And so a lot of the s- the startup advice that people would hear would be, like, you, you really need to, like, sell before you build. You have to do, like, detailed customer discovery and make sure that you've, like, found a real, like-

    2. HTHarj Taggar37:54

       Yeah.

    3. JFJared Friedman37:54

       ... new customer need.

    4. GTGarry Tan37:55

       It was like the lean startups and-

    5. JFJared Friedman37:57

       It's that lean startup.

    6. GTGarry Tan37:58

       Yeah.

    7. JFJared Friedman37:58

       Yeah, exactly.

    8. GTGarry Tan37:59

       Fail fast.

    9. JFJared Friedman37:59

       Fail fast, all this stuff. And that is still the advice that college students, I think, are receiving for the most part because it became so dominant. But I would argue that in this new AI era, that the right mental model is closer to what Harj set, which is just, like, use interesting technology, follow your own curiosity, figure out what's possible. And like, if you're, if you're doing that, if you're living at the edge of the future, like PG said, and you're exploring the latest technology, like, there's so many great startup ideas. You're very likely to just bump into one.

    10. GTGarry Tan38:30

        I guess the reason why it could work extra well today is that you apply the right prompts and the right dataset and a little bit of ingenuity, the right evals, a little bit of taste, and you can get, like, just magical output. And then that's still a secret. I think... Yeah, I mean, you can tell it's still a secret because you can look at l- there are like hundreds of unicorns out there that still exist and that are doing great. You know, like, growing year on year, have plenty of cash, all of that. But the number of them that are actually doing any sort of, like, transformation internally, it's not that many. Like, a shocking few number of companies that are, you know, 100 to 1,000 person startups that, you know, they're going to be great businesses. But that class of startup, like, by and large, they are not entirely aware.

    11. JFJared Friedman39:24

        Totally.

    12. GTGarry Tan39:24

        Like, there isn't a skunk works project in those things yet.

    13. JFJared Friedman39:27

        Yeah.

    14. GTGarry Tan39:27

        Like, you know, the extent of it is, um, maybe the CEO is playing around with it. Like, maybe some of the engineers who are really forward-thinking are doing things in their spare time with it. Maybe they're using Windsurfer cursor for the first time. And it's like, you look down and you're like, "What year is it?"

    15. JFJared Friedman39:45

        (laughs)

    16. GTGarry Tan39:45

        Like, it's a little bit like, "Hey, you know, get on this." Like, I think Bob McGrew, uh, came on our channel, and he was just shocked. Like, he was one of the guys, as chief research officer, like, building, you know, building what became O1 and O3 and all these things. And then he releases it and, like, who's using it? Like, he expected this, you know, crazy-

    17. HTHarj Taggar40:05

        Mm-hmm.

    18. GTGarry Tan40:06

        ... you know, outpouring of like, "Intelligence is too cheap to meter. This is amazing." And it's like, actually, like, people are mainly just, "We're just still on our quarterly roadmap unchanged from, you know, even a year ago."

    19. JFJared Friedman40:18

        Yeah.

    20. GTGarry Tan40:18

        Pretty wild.

14. 40:19 – 40:46

    Outro

    1. HTHarj Taggar40:19

       All right, cool. I think that's all we have time for today. My main takeaway from this has been there's never a better time to build. So many ideas are possible today that weren't even possible a year ago. Um, and the best way to find them is to just follow your own curiosity and keep building. Thanks for watching. See you on the next show. (upbeat music)

Episode duration: 40:46