How Soon Will AI Power A Billion Dollar Company Run by One Person? W/ Henry Shi & Guarav Jain | Ep11

Hi everyone, I'm Anamitra, co-founder of A4 Capital and host of the Pre-Seed Show. In today's show, we're asking a central question: how far are we from a billion-dollar revenue company with just one founder? No engineering staff, no sales staff, no customer support, no execs, nobody except one person who's doing everything with AI. Is that even possible, or is it happening already? To discuss this, I'm going to talk to two people, which is a little bit different from usual. I'm going to talk to Henry, he's the founder of Super He raised a lot of money with a lot of staff. He believes that the future AI company, AI-native companies, are going to get to a lot of revenue with very few people. To do that, he's actually published a leaderboard where he's tracking it on a weekly basis. And to add to it is Gaurav Jain, my co-founder. He's been an amazing AI investor, having invested in a company called Gamma, which is in our portfolio. You might remember Grant from Gamma who came and talked about what he's building— essentially a Microsoft PowerPoint killer and Google Slides killer, all with AI, very few staff, tens of that is in revenue. I think it might be something, a sign of the future. So to discuss this more, let's welcome Henry and Gaurav Jain.

Hey everyone, so let's get going. First, Henry, super excited to have you. Perhaps you can tell everyone a little bit about your background and your founding journey before we sort of get, get into the, the AI stuff. Just tell folks, share with them a little bit about your sort of founding experience.

Sounds great. So thanks for having me. I'm Henry, co-founder of Supra.com. It's a consumer commerce fintech company, all-in-one app that helps customers save and earn across a variety of use cases— travel, fintech, credit building, cashback, rewards, etc. Yeah, started the company 8 years ago, grew it to about 250 people, $150 million revenue a year, profitable. Just stepped back to the board at the end of last year, so now spending time helping founders doing angel investing, doing dabbling AI, do some self-research, and sharing a lot of content. So excited to be here and share kind of what I've learned and seen in the new AI era.

Very cool. I mean, and Gaurav Jain, uh, a person who needs no introduction, co-founder of Horr, share a little bit about your sort of founding journey because it's not the first time you founded something.

Yeah, well, first of all, uh, thanks for having me. This is a big day for me. I've been, uh, waiting for this invite for the Pre-Seed Show for, uh, however long you've been doing this. So, uh, Good to finally be invited on the pod. So thanks for having me. I'm, I'm Adamantur's lesser half. We started up for what, 9 years, almost 9 years ago. But prior to that, I started a company coming out of Waterloo, which is where I went for software engineering back in 2007, started a company in the mobile space that was venture-backed. That was my first exposure to venture capital as a founder, which then brought me to the Android team. I was an early product manager there, helped scale the platform from less than a million total users, about a million new users a day by the time I left. In the last 13-odd years I've been doing venture, and I've known, uh, I think for most of that, uh, 13 years that I've been, I've been in venture.

Terrific. This is gonna be, this is gonna be pretty cool. So, you know, one of the things, um, Henry, that the, uh, the reason we kind of, kind of did a show like this where we're diving deeper into a topic— normally we just talk to founders and, and, and share their experience about how they went from nothing to something. but this one is a little bit different because we're going to talk about a specific topic, uh, efficiency in the age of AI. And two things sort of happened. One is, uh, as we were, um, kind of distilling down our learnings from last year and presenting it back to our LPs at our AGM in, in November, we kind of looked at— we kind of did a cut of, uh, of our portfolio, the AI-native companies and the non-AI-native companies, and we looked at how long it took them to get to $1 million in revenue, $5 million in revenue, $10 million in revenue. And one of the things we found is they got to that kind of revenue with a fraction of the cost, fraction of the people. And at the same time, you've been publishing and researching AI-native companies via Leaderboard, and that has become quite popular. And as a result of these two things happening and sort of AI in general, it just seems an opportune time to talk about AI-native companies. And maybe with that, perhaps you can just give everyone a little bit of a background about how— what is the AI Native Leaderboard? How did you even get started with it? Uh, where did the inspiration come from?

Uh, here you can see. So I built this leaderboard, uh, about a month or so ago, um, because I started noticing this trend on social media where all these companies were getting to 5, 10, 50, 100 million revenue a year within a short amount of time, 1 or 2 years, with very few people, 10, 20, 30 people. You probably heard some of them like Cursor, Record, 11Labs, Lovable, Bolt, etc. But this information was floating around different threads, different sort of tweets, and different places. So I thought, oh, why not collect it all and build a leaderboard to gamify? I come from a consumer background, so, and there's this new, uh, there's always a bit of applying user psychology and gamification. So what if you collect this data, put it in one spot, and gamify a little bit, and then highlighted companies who maybe you haven't heard about before, but should be, or are on the leaderboard. So in that process, I built this leaderboard and I started researching, and there were a lot more companies than I thought, or than people thought, that were doing this. And a lot of companies were reaching, started reaching out to me to say, hey, you know, we're 10 people, doing $40 million AR or $30 million AR and we should be featured. So it's created this whole movement where founders are now coming to me to be featured. And I think it's like the new flex in some ways, whereas before it was about how much money could you raise? Oh, you know, we raised $100 million. But people were very, I think, nervous to share the AR because in the prior pre-AI era, it was not that easy to get to $20, $30, $50 million AR. But now that it's much easier and possible, I think people are flexing based on actual revenue, which is what a good business should be measured on anyways, or profits. But now you're seeing all these companies who are getting to very high revenues very quickly, very leanly, and they're very much open to sharing. I think it's like a bit of the new flex, and I think that's better for the ecosystem because it's much better to showcase actual revenue and results versus just talking about how much money I raised or what valuation, even though numbers may not justify that.

Such a, such a, such a cool concept.

Such a—

so many things to talk about here. I think the number 1 company there is— looks like it's Telegram. And then, and then you have the, you have the who's who. Mercore is up there, Midjourney, ElevenLabs. How do you— so, so what is the— how does someone get on, get in on this list? And how do you verify the two key things here, which is like their revenue and the number of employees?

That's a good question. So a lot of the companies on this list have shared their information publicly in some thread or some post or some newswire, and you'll see every row has a source column. So we want to make sure that it's verified, or at least public, so that way if the founder is making a public statement about their revenue, you know, you're typically assuming they're not making stuff up. So that's that. But a lot of them are founders reaching out to me. So for example, Develop Health is a good one. The founders reached out to me and said, hey, you know, we're at whatever, $6+ million revenue with 5 people. And I was like, great, can you share some proof or data points or public announcement? And they were kind of— they didn't share this publicly before, but the founder actually took some screenshots from his internal dashboard to verify with me his revenue, right? So you see a lot of private verification with founders, and that's kind of how I verify that these are actual kind of real numbers and people aren't just making them up. So yeah, a lot of it is external, but also a lot of it is just private verification through like a Stripe dashboard or internal billing dashboard.

Right.

It's really good. I mean, I think the meta point here is it's less about the specific numbers, but more about a pattern. That is emerging, which is look at the number of employees, look at the number, look at the amount of revenue, and that ratio is crazy, right? It's a, it's what, what does that tell you? What does that tell you, Henry?

Uh, I think it just tells you that you, uh, two things. One is you can get to revenue much quicker in the AI-native era, uh, because you can bill based on outcomes, you can charge higher ACVs, and it also tells you you can get there much more efficiently with way fewer employees, right? And you see a lot of them actually sub-10 people, right? Which is the crazy part. Whereas before, I think your entire engineering team is probably more than 10. Now the entire company could be sub-10.

Right. And then, you know, what's also interesting here, Henry, is I see more than half of the companies are profitable on this list, right? That is shocking. I've been doing this business for 13 years. I never had any company get to any meaningful level of profitability. —because these companies, there's a reason venture capital exists, is you burn a lot of money. But it's fascinating, these companies are profitable and while still growing. Yeah, yeah, exactly, right?

That's the, like, trifecta: growth, profitable, and high revenue, and efficient. That's right.

Maybe we can take a deeper look at— let me pick on a company or so. Let's say Gamma is a good one, Gaurav, because you've been sort of intimately familiar with that team right from their very early days, very first check. So maybe Maybe you can sort of explain those things, like first of all, what is Gamma? And for the audience, we've spoken with Grant before at a prior episode, but from a different lens, from how he started a company, not the AI aspect as much, but the revenue number, the number of people, and the profitability kind of speaks to something else that's going on.

Yeah, so Gamma is a company that's about 4 to 5 years old. We were lucky to be part of the first round of funding. The idea there is to reimagine essentially what PowerPoint could be. You know, PowerPoint was built probably in the '80s or '90s. The world has changed a lot, you know, not to, not to mention, of course, AI and the internet broadly and how we do work, how we communicate ideas, how we present our ideas has changed a lot, how we collaborate on these on these ideas. So Grant and the team, they wanted to reimagine what PowerPoint could be, started Gamma, had built a very product-centric, I think that's one theme you'll probably see across all of these companies. Part of the reason they've been able to stay small is they're very product-centric and leveraging AI to be able to scale rapidly. And for Gamma, they were at the right place, right time when the ChatGPT moment happened. They already had a head start in terms of thinking about what this new way of presentations would look like. And then of course leveraged AI and OpenAI very quickly to build out what an AI-native PowerPoint or AI-native Google Slides would look like. So here instead, you know, today when you go to PowerPoint and Slides, you start off with a blank canvas, right? And the challenge for me and for a lot of people is that sort of writer's block, right? How do you start in communicating your ideas? Of course, once things start to flow, it becomes a lot easier. But getting started is the hardest part. What Gamma does is you come in and you essentially free text, you know, share what you're trying to accomplish, what your ideas are, what you want to communicate, and essentially gets you started, might get you 60, 70% of the way there. And then it's like you're talking to somebody as you're developing the presentation, right? So I might say, hey, you know, this headline is not punchy enough. Can you change it? It'll come up with a few different ideas, or I want to flip the text to the left and the images to the right, or I want to change this from a pie chart to a histogram. All of these things that I could do today in PowerPoint, but it's much slower. And of course, this is the perfect example of AI, and not just AI, but also AI in the consumer and prosumer space, which we haven't seen a lot of innovation there, but I think it's coming. And I think Gamma's certainly at the tip of the spear in terms of what AI in the consumer world could look like. And because it is a consumer prosumer product, the growth has been, you know, meteoric, as you can see there. They don't have an enterprise sales team. It's all sort of PLG, product-led growth, where, you know, one person inside the company might bring it in and it sort of spreads virally inside of the, inside of the organization. As you can see here, they have sub-30 employees and doing tens of millions in revenue. And prior to, you know, prior to AI, just for them to build a product like this where you need feature parity with a PowerPoint or Google Slides would take tens if not hundreds of engineers, right? Because you can imagine these products that, that the incumbents have been built over, over many, many years. But because of AI, their engineers are so much more productive. One engineer can do the job of 2, 3 engineers, and they've been able to accomplish all of this with a very small team.

That's right. Well, I'm curious, uh, what do you think Gamma did right versus Tome, uh, I think which is also an early sort of player in the presentation space, but I think they ended up pivoting.

I think one of the things that the Gamma team did well, which I think is a pattern you'll see across all of these companies on your leaderboard, is stay small, stay very focused, right? I think the companies, you know, we tell, remind our companies that raising too much capital is just as bad, if not worse, than raising too little capital. And some of the Gamma competitors ended up raising way, way more capital than Gamma did. The problem is, you know, a lot of times there you end up building the wrong product, right? Or for the wrong ICP, and you're farther away from the destination. In Gamma's case, you know, we were part of that first round, which was probably $2 million plus minus. And then they raised a small round from Accel, a seed round, I think it was about $4 million. And then they raised a bigger round, about $12 million or so. But that wasn't until very recently when the company was already, I'd say, post product market fit. So they stayed very lean for the longest time, which just allows them to iterate on that product and really get it right. Because you can't just copy paste what PowerPoint does and then, you know, apply that to AI. You have to think about what an AI media product looks like, which just, it just takes a lot of iterations, especially a horizontal product. Another team that I was involved with is Airtable, and I think this is similarly what the Airtable team did well is, you know, for the first few years, really focus on getting the product right, and then you look to scale, which is what I think the Gamma team is really good at. Right.

Um, hey, Andrew, maybe we can, uh, um, go back to the original screen, um, because we can, we can discuss a few other items. Outside of the— outside of this. So, uh, so, Henry, question for you. Um, on the one hand, um, we've historically heard, uh, and seen, um, uh, AI labs, uh, raise vast amounts of capital. Just, I think, just, uh, in the past week, we heard about a $40 billion round that OpenAI raised at some crazy valuation, um, obviously leading the AI revolution. At the same time, so they're sucking up a lot of capital. A lot of the AI labs are investing significant amounts of money. I think Microsoft and Google have kind of reaffirmed, despite what's going on with tariffs, their intention of investing tens of billions, $50, $70 billion into their foundational models and infrastructure. On the one hand, we're seeing significant amounts of capital go into these companies, and on the other hand, we're seeing these application AI companies being very, very nimble raising very little and going quite far. How do these both two things make sense? And if so, like, how do you align the two? What's going on?

Yeah, that's a good question. I think both can be true at the same time, and both probably could also play out in a different type of multiverse, right? You have Deepseek, which is no venture funding. I mean, they have money from their trading, but it's a lean team and they're still able to build frontier models. So I think it just shows you that capital is helpful, but it's not the only way to play the game. And I think a lot of founders, especially repeat founders who've been through the venture cycle, realize that there are many ways to play the game and be a founder. I talk about various funding models, seedstrapping, bootscaling, venture bootstrap, and I think it's becoming a lot more common for founders, especially repeat founders and smart founders, to realize that there are many ways to play the game and there are many, many paths you can take to be successful. And it's not just one path like maybe it was 5, 7, 10 years ago with hyperscaling, blitzscaling venture capital. So I think that's kind of some of my perspective is there are many paths to take and kind of like in the Tome example, right, you could also raise a lot of venture funding to do AI for PowerPoint. Doesn't mean you're gonna win in the end, but certainly And more founders are realizing there's ways to maintain control, but still work hard. And these are not lifestyle founders. They still work hard, but maintain control and maintain nimble and work with great people and not have to deal with overhead of hiring hundreds of people, managing hundreds of people, HR issues, and whatnot.

And if I can jump in, you know, if you can look at sort of the, yeah, the cloud shift as some analogous to what's going on right now. If you look at the hyperscalers, AWS, GCP, Azure, those are very expensive, you know, buildouts, right? It's a CapEx heavy, takes billions of dollars to get to the scale that Amazon or Google are at. But then what that allowed the SaaS startups built on top of it to be hyper capital efficient because you didn't need to invest in the data centers because Amazon and Google had done that for you. So you were able to essentially flip that CapEx into OpEx, right? You only paid for what you used on AWS versus having to build the data centers. I think what AI is doing is that on steroids, right? So now, because you're able to either automate parts of the job or at least like replace human labor in some ways, now you're able to take that fixed cost of talent and flip that into a variable cost, right? Because if you look at startups, 70% or 80% of the money that you and I would give to these companies would go to talent, right? Would go to engineers, right? And then sales over time. If you can now flip that into a variable cost, that makes these companies hyper-efficient because the variable cost only goes up as your revenue goes up. So I think you'll see a similar pattern where companies like OpenAI will be very capital intensive because they're essentially building what AWS did for cloud, but for AI, and then stuff that's built on top of it is going to be hyper-efficient.

So, um, uh, so on, on that note, uh, uh, we know what bootstrapping means. Bootstrapping is a founder, um, kind of not raising any outside money, using their— using up their savings to build a product, uh, find customers, and then eventually raise money down the line. You introduced a few other concepts, Henry, uh, seedstrapping, uh, and a few others. Can you tell our audience what it means? What is seedstrapping? Um, And what are some of the other terms that are opposite of blitzscaling and so on that are, that should enter the venture lexicon? Sounds good.

Well, I have a chart that illustrates that, so hopefully makes it a little bit easier, right? But seedstrapping, typically there's 4 models actually. So there's bootstrapping, which is not raising any money, so you start with zero and you just kind of grow steadily. Then there's venture capital, which is you raise money, grow, raise money, grow, raise money, grow, Series A, B, C, D, E, F. Gee, you keep climbing the ladder. Then there's seedstrapping, which is you raise initial seed money and then you just start to compound growth from there and you don't necessarily have to raise again. You could raise, but you don't have to. And then there's bootscaling, which is you don't raise any money, you bootstrap for the first 5 years, then all of a sudden you get a massive round like GitHub, $100 million, and then you start scaling from there. Typically it's like a PE type of, or growth equity type of round. So these are the 4 different models. And I think there were people doing bootstrapping and seed funding before, but now I think you're going to see seed funding a lot, a lot more common, like many of the companies on the leaderboard. I'll have you dive into some specifics and how it compares when it comes to dilution, ownership, revenue growth, liquidity, and whatnot. But essentially, these are the sort of 4 different models.

That's really cool.

By the way, I'm going to add one thing. Sorry, if I can add one thing real quick. I think what we find is for founders to quit their jobs and run a set of experiments for the first 6 months, 12 months to find the right product, the right market, they need a little bit of funding. I think this is where we're going with seedstrapping, right? Where you raise a little bit of funding just to run a set of experiments. And then once it starts to work, perhaps you don't need as much capital, but we still find there's a need for capital upfront just so you can do this full-time. Right, exactly.

And here you'll see if you— so if you bootstrap, right, you're in the negative for a very long time because you're using your own credit card or savings or something. So the good part about raising some money is that the founder, they're not starting in the negative and they don't have to wait many years before they're in the positive. So that's a nice sort of way for the founders to overcome the friction or hurdle of being an entrepreneur.

That's right. What are you seeing on the seedstrapping side? Because that's, that's interesting. That basically says you can, um, with a— with— you don't need a lot of money, you maybe a few million dollars to get to profitability and growth. Traditionally, that's been really hard. You can't get to profitability or growth without a lot of capital. That's a big difference here, right? Mm-hmm. Yes, that's right. What are you— in your sort of the companies that you are tracking, which one is playing out more, do you see? Do you see seedstrapping to be the most popular one? Do you see bootscaling the more popular one? Which— are there any patterns or trends emerging for you?

Yeah, I think, uh, personally I think seedstrapping is, uh, superior to boot scaling, uh, because you could always seedstrap to boot scale, uh, right? So it doesn't close any doors. You could seedstrap or raise a lot of venture if you want to, which just becomes traditional venture capital, but you don't have to. Or you could seedstrap and then wait many years and then get a mega PE growth equity round. But I think it's strictly superior boot scaling because you don't need to be in debt for the first X number of years. And there's a lot of risk to bootstrapping because with bootstrapping or bootscaling, you don't really have any buffer, right? Like what you eat is what you kill, and if you have an off month, you know, if you're living off your profits and revenue, it could be really scary. Whereas if you have a good sizable seed round, it gives you that buffer, it gives you a cushion, and it lets you sort of make investments. So I still think the first round is sort of important, but it just gives you the best optionality of all four. Right.

I want to come back to what are the conditions under which companies can do either of these things behind, like, under the surface. But before that, and since we're talking about venture capital, Gaurav Jain, I want to get your reaction on what the implication of all of this is to venture capital in general, especially as it relates to maybe seed stage or Series A stage. Are beyond, what, what do you think is going to happen?

Yeah, I think there's several implications, and there's different implications depending on where in the company's lifecycle you invest. For somebody like us that invested pre-seed, obviously seedstrapping is very attractive, right, for several reasons. One, our potential dilution, right, from the time we invest to the time there's an exit would be very different. right, than the pre-AI world. In the pre-AI world, we used to model out at least 50% dilution from where we start to where we end up in our best companies. So if we start off with 10% ownership, we're thinking by the time this company exits, whether it's an M&A or IPO, we're going to own 5% or less because, you know, we can only invest in so many rounds of a company. At some point, it's just too expensive for a fund of our size to be participating in those rounds. So we start to take on dilution. Now, if those rounds are not happening or happening much later in the company's life cycle, when the company is raising at a much higher valuation, our dilution curve may look very different. Now that may have several implications, right? One, perhaps we'll have to pay a higher price, 'cause there's like lower dilution baked in, right, to get in, or we might be okay with a slightly smaller outcome, 'cause ultimately what matters to the fund is the outcome size times our ownership. So if our ownership is higher, you can live with a smaller outcome and still have a similar outcome for the fund. So I think it's too early to tell exactly how this will pan out, but we're already starting to see some early signs of that. I think if you're a later-stage investor, obviously there's some serious implications if these companies are not raising capital or raising capital much later in the lifecycle when they're more de-risked and hence Hence their, their, their price expectations are going to be much higher. I don't know exactly how they'll react. I think we may have to innovate on new products, right, where maybe the money is not going on the balance sheet of the companies. It's not a primary investment as it used to be. Maybe it's more secondaries, right? Maybe the founders or early employees are looking for liquidity, or early investors like us are looking for liquidity. And hence the later-stage investors are still getting exposure to the company, but just getting exposure in a different way. And then there is the third implication of, well, what hap— if these companies can be profitable and can technically run on forever, like how do we eventually exit, right? Ultimately, our business is not about dividends and being in these businesses forever, but it's about, you know, capital gains. It's about getting in at a certain price and then selling that stake, which the companies used to be forced to sell because if you're cash flow negative at some point, you need to either get to cash flow positive, which is very hard for high-growth companies, or you exit through an IPO or by a strategic M&A. But now with this new world, I don't know, maybe the founders want to keep running this company forever. So exactly how we exit these positions, I think again TBD, or even later-stage investors that come in, how did they exit the positions, right? Perhaps then you might see new products like a debt recap, right, where the later-stage investors are able to raise debt for the company and basically buy back the shares, right? Because these companies are profitable, they can take on debt. Anyway, it's too early to speculate exactly what the exit profile is going to look like, but we can tell it's going to evolve. And another thing I want to say is I think it gives a lot of power to the founder, right? Because they own the destiny of their company, right? Because in the pre-AI world, when you're losing cash every day, you are beholden to the capital markets, you are beholden to investors to keep your company alive, especially because for high growth, you need a lot of capital. In a world where you can keep growing and be profitable, I mean, that's the perfect, perfect place to be as a founder. So what that means is if, for example, you hit a couple of rough quarters, you can now toggle down your variable cost, right? Because your revenue goes up and your team is fairly small, your fixed costs are fairly small. So you're not, you know, you're able to withstand the volatility of your customer base or the markets broadly. And I think that— I think what that may mean is we might see fewer, fewer complete zeros in the portfolio as well, because historically the reason companies go to zero is because they run out of cash. They run out of cash because they're not able to attract more capital. But in a world where they're profitable, perhaps not growing as fast, they can still survive, right? In the pre-AI world, if you were not growing fast enough, it's game over because you're competing with companies that are growing really fast and they're the ones attracting venture capital. So you had to find an exit. Now there may be more options available to you as a founder. So I think ultimately this is very empowering for the founder, which we think is an incredible development. That's a really good point.

Henry, one question for you is like, as normally VC investors think of themselves as not just distributors of capital, but also of ideas, of mentorship, of advice, of networks, and their brand. One of the reasons many founders take VC money from some of the best funds in the business is because it allows them not just the capital to grow, capital to maybe acquire companies, but also it's a stamp of approval, it's a validation. for employees, customers, etc., that this company has been vetted, it's on an IPO track, it's going to happen. Do you think that gets diminished, or do you think there is founder appeal for something like that still, despite, despite them not needing the capital to grow?

Yeah, absolutely. I think there's still a big part of founders who want or need or like to raise money because the cultural zeitgeist or the social norm is to raise money, right? And whether that's good or bad or right or wrong is TBD. But certainly I've talked to many founders who didn't really need to raise money, but for their employees and for some of the hiring, they had to get some logos. So it's kind of like, well, in some sense a toll or tax. And there's obviously real benefits too. But I think for many founders, especially those who don't touch the money, right, I think one thing someone said is, you know, there's founders who brag, "Oh, I raised $100 million and I didn't even touch a single dollar." Well, actually, that's not very smart because you just gave up a lot of your equity for no reason because you don't even need the money, right? And that can be very expensive. So I do think that hopefully can change, and I think we're starting to see some of it with the leaderboard. Like, the leaderboard is one way for people to showcase themselves and their credibility, legitimacy, or awareness without having to raise money. So there are a lot of founders who reach out to me who want to share because I think it's their way to showcase, to hire, to be featured, and to be known, and to break out of the noise based on actual metrics and execution and not just, you know, a logo of some firm. So I think we'll see some of that change. And to Gaurav Jain's point, I think with some of the secondary, perhaps there can be ways where Well, if you don't need the money, but you still want the network and the benefits and the community and the mentorship and advice, maybe the secondary route is more win-win for the founder, where at least, you know, they get some money in their own pockets versus raising a lot of money for— and who actually don't need the money or never touch the money.

Yeah, one thing if I can add real quick, um, while the venture economics might be changing, right, and how much money you need to raise and how quickly you can scale, What's not changing is how software is sold, at least today, right? What's not changing is humans are not evolving at the same rate as technology is evolving. What that means is humans are still, you know, lazy, afraid, risk-averse, all of those things that make us human. That's still the case. What that means is, especially for companies that are in the B2B world, they still need to figure out How to run an enterprise sales motion, how to win a bake-off, how to convert pilots into real contracts, how to win the renewal. All of those things have not changed. And I think in some ways gotten even more complicated because there's way more competition now coming from other startups, not to mention a lot of build versus buy going on. So, and remember at the same time seeing founders getting younger and younger, partially because AI is just so accessible that we're seeing the average age of founders come down, but which is great from a technical perspective, but those folks that have the experience of having sold products before, built large go-to-market machines, and I think that's where there will still be a need for mentorship around how do you build a company around this product.

All right, so Henry, just pulling on that, that thread a little bit more in terms of, uh, of a number of people you need. So many of the companies, um, that we are seeing, and I guess you're seeing, whether it's 5, 10, 20 Um, which are the roles that are most conducive to AI first? Are you seeing— are you seeing any patterns there? Is it— does it always start with engineering? Does it always start with sales? And— and where is the breaking point? When do companies realize, no, I actually need to hire someone in this position even though I have an AI assistant or AI copilot? I can't go further.

Yeah, that's a good question. So I think, uh, the primary and the breakthrough use case right now is in coding. Whether it's in agentic systems or Copilots or low-code tools like Lovable, I think you're seeing a lot of sort of automation and augmentation there. So definitely if you were to start a company today, right, you would do AI-first, AI-native, especially on the coding side, on the product side, engineering side. So that's that. On the non-engineering side, I don't know if there's been a breakthrough winner or emergence of a tool like Cursor, but you're starting to see a lot of workflows being automated. And I'll just share my screen. I just made a post about this. There's a founder of Arcads who just went from, uh, went to $5 million ARR with 5 people in under a year, and he actually shared his playbook of how he operates his lean AI startup under the hood. So all the different use cases outside of coding, right? So he has an AI spy agent that looks at competitor ad libraries, feeds the best ones to O1 and auto-regenerates viral clips using his own product, which is Arcads. It's a ghostwriter, scripts top-performing tweets, suggests new ideas, and then drafts them for review. Uh, there's an AI intern who pulls Zoom transcripts, extracts ideas, generates top ideas. There's a Google Ads agent which looks at keywords and then creates landing pages for each keyword using Claude and then auto-adjusts the bids daily. It's a custom support agent Sales Manager for Lead Enrichment, and AI Accountant. So a lot of this is home-built based on some tools or scaffolding, but you can get a sense of the types of tools and agents outside of coding that are starting to form. And maybe there will be a cursor for marketing or for sales, something like that, but I don't think that exists today in the sense of this breakout winner. So maybe there's an opportunity there. And in terms of the last question, last part of the question is when do you hit a breaking point? When do you actually hire people? I don't know if there's a hard science at the moment, but what I've found is for me, the tools come to a point where it's good enough, but if I want someone to take this completely end-to-end, it's nice to have a human who can just manage these tools. So even for Arcad, they actually do have a human customer support person who manages 20% of escalations. But also, I'm assuming that person probably helps manage the support bot as well to improve it or to tweak it. And you still need a lot of cycles or mental cycles to tweak these tools. So having a person who's able to just own it end to end, including refining, optimizing, tweaking the AI agents I found to be helpful as well, just so that it gets sort of this task off of your mental plate and so you can focus on other things.

Yeah, I remember this post because he kind of breaks it down to the various subcomponents. And for everyone in the audience, the LinkedIn inbound messages that Henry receives on the right are all folks who want to be in the leaderboard and maybe at some point that will be AI'd also. Gaurav Jain, I have a question for you, and it is one of the things that founders, founders in the beginning, like a beginning of, for tech, they needed to build their own servers, they needed to build their own website infrastructure. Now, some of those things have been sort of abstracted away, they don't need to worry about servers anymore, they don't need to worry about building their own websites anymore, it's easy. Even then they needed to to hire a few people to build out an interesting product, because it's not just a website, it's a feature set. But nowadays, you don't need to hire anyone to just get an MVP or a prototype or something up and running. As you and I, we're looking at sort of pre-seed stage companies, where is the bar now to a company? Is it still the famous saying that I raised with a pitch deck, or is that no longer true because you should be able to build something? With all the tooling that's out there.

Yeah, and what's interesting now with vibe coding is you don't even need technical skills, right, to build a prototype. I mean, you and I are investing at the prototype stage, which pre-AI might mean a bunch of Figma mocks, right, wireframes, right? That's what we were investing off of, or sometimes they don't even have that and it's just an idea. But now increasingly we're seeing companies with a working prototype, perhaps some early design partners. Now it might not be a lot of revenue yet, but you can now accomplish a lot more with no capital, but it's not just about capital, but barely any time spent, right? Evenings and weekends hacking away while you're still employed somewhere else as you're thinking about what's next. You can run a bunch of experiments, you can build a prototype. So the bar's definitely gone higher, but it's not just higher for, pre-seeds, we're seeing bars just going up higher for like every round of funding because these companies are growing so much faster and so much more efficient that if you're not, you know, scaling revenue like this, it's just becoming harder to attract talent, harder to raise money if and when you want to. So the bar's definitely gotten higher. I mean, you look at the stack that this founder's put out, he's able to 10x the output of the number of people that this person probably has in the company. And now everybody's going to be compared to those benchmarks. So if you're not adopting AI in a very deep and intimate way in your company, not just to write code, but across the board, you're going to get left behind.

Yeah, you know, to that point, even larger companies, tech-forward companies, are kind of figuring that out and asking their employees to act accordingly. The Shopify CEO sent out a note on hiring, right? Did you hear about that one, Henry? That's right, yeah.

Yeah, the Shopify team and Harley is one of our angel investors. They had a note about being AI-first, AI-native, and sort of adopting or adapting, or maybe this is not the place for folks who don't want to adapt. So in some sense it's inevitable, and I think the letter ended with, or had one of the points, which is for every hire there has to be a justifiable reason why AI can't do this task in order to open a headcount. Right, right.

And Henry, it's not going to stop there, right? The next thing they're going to do is not just for new hires, but you need to justify your job on why you should not be fired, right? And why can't AI just replace you? And they're going to start off with the bottom 10% first, right, that are going to get fired because AI will replace their job, and it's going to go up the stack. Now, I personally don't believe that AI will be able to automate everything that humans do. I think what we're headed to is AI will do a lot of stuff that humans don't want to do, or perhaps can do, but there are better things they can be doing with their time, right? Like the thinking work, like judgment, like taking risks, like, you know, pushing the boundaries on what's possible and let the AI do a lot of the grunt work, which I think is is great, and we all have to evolve what we do day to day to meet that new normal. Right, right.

Now, ending the podcast on where we started, which is the notion of the concept of one day there will be a billion-dollar revenue company with just one person. So Henry, how far away are we from there? How far until in your leaderboard you have a billion-dollar revenue with one person that's profitable and growing? Yeah, that's a great question.

I think, you know, we're still some ways away. Telegram right now is about 30 people doing a billion in revenue and profitable. I think last year they made like $500 million in profit or something crazy. So there's, you know, that's some notion of that. But to get truly to a one-person billion-dollar company, I think we're still a bit ways away. However, one question I'd like to pose to the audience is, as you see more and more AI in the workforce, do you think it's going to be one person, like one, or sort of a human at the top as a CEO and AI workers, or is it actually the AI as a CEO and you have humans doing some tasks, right? I think that's an open question because you might argue that AI is really good at synthesizing all this information across all this domain with millions of tens of millions of context windows and running, you know, hundreds of simulations in code and outputting the best sort of strategic plan, and then humans just go execute it? Or is it the other way around, where humans design the plan, the AI take sort of some of the grunt work or the IC work? And it's not really clear, um, necessarily which model is going to be the prevailing model. Um, so I look forward to seeing how that evolves. How that evolves. And I think I just saw on social media where one founder just made a post where she resigned as a CEO and made her AI CEO. So, so let's see how this all evolves.

Yeah, I mean, I mean, it's in the real world, for example, in, in automotive, the driver's seat has already shifted. It's AI that's in the driver's seat, uh, and then humans are in the passenger seat, which, which might happen in the digital world too. It might happen in this world where and it won't appear to be AI. It'll appear to be a human being, um, and on a call just like this, but, uh, instead of it being real Henry, it's, it's Henry's avatar, and then we're working for it.

Henry, I used to tell founders that your job is safe, uh, you know, and, and there won't be an AI founder, but you're making me think twice now. I mean, if an AI can be CEO, it's not going to take a lot for a reasoning agent to also become a founder. Maybe, right?

Well, I mean, even today, like, if, you know, if the founder is consulting an AI oracle, right, and the founder just relays the message, right, who's really the CEO, right? And a lot of times you don't know. Like, you know, it's kind of like who's really running the government. So obviously not AI today, but you could see a world where maybe the AI has just better synthesis of information. Maybe it's not really good at browser use, but it's really good at information synthesis, and that actually might be better for a CEO. Just like in real life, the CEO is not the one writing the code or the one using the browser. The CEO is the one making the strategic decisions.

And then maybe one final question for you, Gaurav Jain. In this new world where Henry already has a company that's doing a billion with only 30 people, so not not far, where it's just one person doing a billion, they don't need any money, they're profitable. What does it mean for VCs? What do we do in that new world?

I think there's a lot, frankly, that we don't know yet. I still feel that the day of AI founder is still far away, right? I think humans are really good at spotting gaps, taking risk, right? Because starting a company is is risky, right? Running a bunch of experiments. And I definitely think there will be a place for that initial capital to like let you run the experiments and then some mentorship and help to figure out how to put a company around it. But as these companies start to scale, there may not be as much of a need for growth capital, right? And expensive dilutive capital down the road. And I think maybe the capital markets will have to evolve to meet that new normal. But net-net, I think if you zoom out, I feel very excited about the future. I think we'll see many more companies getting started. Those companies will be a lot more effective and efficient in solving the customer's problems. So overall, the productivity of our society will go up. Many more problems that today are not getting solved are going to get solved. And people, you know, these things are all compounding, right? You build on top of each other, right? So that compounding increases, give it 10, 20, 30 years, and we're going to be so much farther ahead than than we would've been. And I think just like, you know, workers we were talking about at Shopify, I think all of us have to evolve, including venture capital, right? And we have to adapt as these things are changing. I think it's too early to tell exactly what 10 years down the road looks like for venture capital, but I think one thing I can guarantee is it's gonna look very different. Yeah, yeah.

You know, on that note, I think we are all technology accelerants here. We are all optimistic about the future and the accelerating power of AI. So with that note, we'll, maybe we'll do this again in a year or so and see where we stand and how the leaderboard changes. Henry, wonderful having you. Gaurav Jain, wonderful having you join. Let's see where all of this goes. And thanks for everyone who joined.

Thanks for having us. Azad, thanks so much.