The Exchanges

Q When you look back at the distribution, the channel strategy, what mistake do you think you made?

A The mistake I think we made is not hiring early enough. We had super strong PMF right from the start. We had channels working right from the start, but we didn't actually have anyone focused on certain channels. So for example, affiliates, we set up over a year and a half ago, and that now is bringing in over, you know, tens of thousands of dollars of MRR Per month. But this was set up a year and a half ago by one engineer in one week, and then no one's touched it since. And what we should have done is, once we've seen signs of life, given we already have PMF overall, staff one person who can focus on growing that one product, and those one set of KPIs, and nothing else. Keep them laser focused. So yeah, overall, I wish we had staffed up more channels with more dedicated people sooner.

AI assessment note: “The mistake I think we made is not hiring early enough.”

Q I have not heard of this. Are we reading? Carcerating Titanic? Like, why do we need motion designs?

A So video is hugely underrated as a medium for growth, and so every launch video we do, the keystone marketing piece we ship with it is the video. There's different types of videos you can do. You can do motion design, which is basically animated graphics. You could do a founder-led video where they're speaking to the camera. You could do a more like screen share style video where they're kind of recording and interacting, and it's very product focused. But why I love motion design is it's abstract, so you can really focus on what are the core value props. You can make it catchy and attention focused and mix in the abstract UI elements, and the big mistake that I see with most people's videos is that they put the founder themselves, they do a full five minute monologue themselves about the company mission, and unless you have the editing skills of MrBeast, you're just not going to keep people for the launch video. So make it sure Get the key message across in the 1:30 seconds, and motion design is fantastic for this.

AI assessment note: “why I love motion design is it's abstract, so you can really focus”

Q When you look back at the distribution, the channel strategy, what mistake do you think you made?

A The mistake I think we made is not hiring early enough. We had super strong PMF right from the start. We had channels working right from the start, but we didn't actually have anyone focused on certain channels. So for example, affiliates, we set up over a year and a half ago, and that now is bringing in over, you know, tens of thousands of dollars of MRR Per month. But this was set up a year and a half ago by one engineer in one week, and then no one's touched it since. And what we should have done is, once we've seen signs of life, given we already have PMF overall, staff one person who can focus on growing that one product, and those one set of KPIs, and nothing else. Keep them laser focused. So yeah, overall, I wish we had staffed up more channels with more dedicated people sooner.

AI assessment note: “The mistake I think we made is not hiring early enough.”

Q Will it be in two to three years? When we see the shift in terms of users and where they come from, more and more, I think, you know, Guillermo from Vercel said it was one and a half six months ago, and now it's four and a half. In two to three years, do you think we will see a much reduced SEO presence?

A With SEO, there's different types of SEO content. The first one is blog-style content. This is, you know, long words, long-form articles. That, I think, will increasingly die over time. But actually, it's not dead yet. If you look at Zapier, over 70% of their SEO traffic still goes to their blog. That's definitely not dead yet, but I think will die because you'll be able to have ChatGPT give you a really great response, it will have scraped it, ingested it, but the part which I am confident will stick around for at least five years are tool pages, and so this is where, to give you an example, if you search text-to-speech Spanish, The first result would be 11 Labs text-to-speech Spanish, which is a text box where you can type in any Spanish you want, choose a voice, and click play. And so that, like, requires some actual engineering work to make this text box. We do the same with speech-to-text. You create these mini-tools, often with proprietary data sources or some actual engineering work, and that's going to be a while before these LLMs are, like, spinning up whole dynamic pages, and by that point, maybe even bets are off because they're kind of replacing your product on the fly anyway.

AI assessment note: “stick around for at least five years are tool pages”

Q Why is inbound SDR as a role dead?

A If you think about the actual function of most inbound SDRs, so with an enterprise sales cycle, someone fills in a form, they then need to speak to an inbound SDR who basically goes through BANT screening. So they work out, do you have the budget? Do you have the authority? Do you have the need? Do you have the timing? If yes, they then book you into another call with the AE, which actually closes the deal. And so the function of an inbound SDR is I'm sure the best ones are persuasive and they're friendly, but really the function is data collection, and so instead, one, why don't you just add that to the form in the first place? Well, the reason is people don't really want to go through this back and forth and go into more detailed questions and, you know, trying to design these complex flowcharts of forms, and so instead you can switch out to a conversational AI agent which can completely replace that role and then get The people who should be booked in with AEs much quicker. You could even clone the voice of the AEs so it's like them almost pre-screening for their own deal flow.

AI assessment note: “a conversational AI agent which can completely replace that role”

Q They did kind of change their whole AI team, I think, as a result of that. What have you changed your mind on in the last 12 months?

A The biggest thing I've changed my mind on in the last 12 months is whether voice AI agents or conversational AI agents actually work. And so when I joined at Leathernavs, we had this fantastic box where you could type in text and generate lifelike speech. And Matty was saying, like, ah, soon you'll be able to have full AI agents that you can have natural conversations, and it'll sound really human-like, and you could use that for sales, and for customer support, and for relationship coaching. And I was like, that will never work. The latency will just be too high. You won't be able to have these sub 200 millisecond interactions. I was completely wrong. And so, on about the 20th customer that Asset 11 Labs had helped build these conversational AI agents, we were like, ok, Let's actually turn this into a proper platform and build conversational AI as a feature, and now it's, I actually prefer speaking to an AI customer support agent than a real person, because they're more knowledgeable, they know exactly how to solve your problems, they can escalate, they know all the policies.

AI assessment note: “The biggest thing I've changed my mind on in the last 12 months is”

Q So you've said counter positioning twice. I love seven powers, and so, like, for me, it's like, you know, incredibly exciting, but what is counter positioning first?

A Counter positioning is where you take what someone else's, like, core messaging or core positioning in the market, and you take the exact opposite, and you use that as your strength rather than your weakness, and so the example with Brex is Brex really stood for, like, hey, spend as much money on your card as possible. We give you all these fantastic Points, and it's kind of like, look how many, like, extra points and rewards and stuff we're giving. And then Ramp took the exact opposite approach, which was, hey, we're not going to give you any points. Points are a waste of money. Points you end up spending on stupid things. Instead, we're going to pump that all back into software. As soon as they made points an actual bad thing, and got the whole market to think about points being bad, let's use it to invest in great software, and the focus instead being on saving money, and almost reducing the number of these imaginary points you're getting. They were really well counter-positioned to basically every time Bricks would do something, they'd be able to riff off the other way.

AI assessment note: “Counter positioning is where you take what someone else's, like, core messaging”

Q Matty, this is for 11 Labs, and it's like, we're gonna be everything for everyone. This sounds bad. Like, I like a tight ICP where we can market to, we know which channels we're selling to, we can craft a product for them. How has a horizontal product offering worked so well when it is completely counter the narrative that we're always told?

A I a hundred percent agree. And actually, when I worked for post-hog, part of the magic was you have one ICP, which is a product engineer, and you're selling multiple products into it. So you're selling product analytics, you're selling session recording, feature flags, all in, and each new product you add, it gives you another entry point. It enables you to, like, integrate the products to do completely new use cases, which couldn't be done before. We don't have any of that in some ways. And so with 11 Labs, because we have the best AI audio models, so turning text into speech, speech into text, then making sound effects for voice cloning, voice isolator. This is like a very horizontal foundational layer, which we sell as an API. So that's kind of your first product, the developer product. And then you can sell that API at scale. So you can sell it to enterprises. You can build around workflows doing conversational AI. So that's fantastic. But then, because the team is super ambitious, you know, Matty and Peter literally want to build a bigger company than Google. They want to build something absolutely massive. And then they see like, oh, we could actually build this to build a reader app for your phone to read articles. So that's now a product. Then we're also building a creator platform. So this text box, which anyone can type in text, that's now creating whole audio books a…

AI assessment note: “I think the reason we've managed to get it working is because we've started with”

Q Will it be in two to three years? When we see the shift in terms of users and where they come from, more and more, I think, you know, Guillermo from Vercel said it was one and a half six months ago, and now it's four and a half. In two to three years, do you think we will see a much reduced SEO presence?

A With SEO, there's different types of SEO content. The first one is blog-style content. This is, you know, long words, long-form articles. That, I think, will increasingly die over time. But actually, it's not dead yet. If you look at Zapier, over 70% of their SEO traffic still goes to their blog. That's definitely not dead yet, but I think will die because you'll be able to have chat GPC give you a really great response, it will have scraped it, ingested it, but the part which I am confident will stick around for at least five years are tool pages, and so this is where, to give you an example, if you search text-to-speech Spanish, The first result would be 11 Labs text-to-speech Spanish, which is a text box where you can type in any Spanish you want, choose a voice, and click play. And so that, like, requires some actual engineering work to make this text box. We do the same with speech-to-text. You create these mini-tools, often with proprietary data sources or some actual engineering work, and that's going to be a while before these LLMs are, like, spinning up whole dynamic pages, and by that point, maybe even bets are off because they're kind of replacing your product on the fly anyway.

AI assessment note: “That, I think, will increasingly die over time... part which I am confident will stick around”

Q Why is inbound SDR as a role dead?

A If you think about the actual function of most inbound SDRs, so with an enterprise sales cycle, someone fills in a form, they then need to speak to an inbound SDR who basically goes through BANT screening. So they work out, do you have the budget? Do you have the authority? Do you have the need? Do you have the timing? If yes, they then book you into another call with the AE, which actually closes the deal. And so the function of an inbound SDR is I'm sure the best ones are persuasive and they're friendly, but really the function is data collection, and so instead, one, why don't you just add that to the form in the first place? Well, the reason is people don't really want to go through this back and forth and go into more detailed questions and, you know, trying to design these complex flowcharts of forms, and so instead you can switch out to a conversational AI agent which can completely replace that role and then get The people who should be booked in with AEs much quicker. You could even clone the voice of the AEs, so it's like them almost pre-screening for their own deal flow.

AI assessment note: “conversational AI agent which can completely replace that role”

Q Will it be in two to three years? When we see the shift in terms of users and where they come from, more and more, I think, you know, Guillermo from Vercel said it was one and a half six months ago, and now it's four and a half. In two to three years, do you think we will see a much reduced SEO presence?

A With SEO, there's different types of SEO content. The first one is blog-style content. This is, you know, long words, long-form articles. That, I think, will increasingly die over time. But actually, it's not dead yet. If you look at Zapier, over 70% of their SEO traffic still goes to their blog. That's definitely not dead yet, but I think will die because you'll be able to have ChatGPT give you a really great response, it will have scraped it, ingested it, but the part which I am confident will stick around for at least five years are tool pages, and so this is where, to give you an example, if you search text-to-speech Spanish, The first result would be 11 Labs text-to-speech Spanish, which is a text box where you can type in any Spanish you want, choose a voice, and click play. And so that, like, requires some actual engineering work to make this text box. We do the same with speech-to-text. You create these mini-tools, often with proprietary data sources or some actual engineering work, and that's going to be a while before these LLMs are, like, spinning up whole dynamic pages, and by that point, maybe even bets are off because they're kind of replacing your product on the fly anyway.

AI assessment note: “the part which I am confident will stick around for at least five years are tool pages”

Q When you look back at the distribution, the channel strategy, what mistake do you think you made?

A The mistake I think we made is not hiring early enough. We had super strong PMF right from the start. We had channels working right from the start, but we didn't actually have anyone focused on certain channels. So for example, affiliates, we set up over a year and a half ago, and that now is bringing in over, you know, tens of thousands of dollars of MRR Per month. But this was set up a year and a half ago by one engineer in one week, and then no one's touched it since. And what we should have done is, once we've seen signs of life, given we already have PMF overall, staff one person who can focus on growing that one product, and those one set of KPIs, and nothing else. Keep them laser focused. So yeah, overall, I wish we had staffed up more channels with more dedicated people sooner.

AI assessment note: “The mistake I think we made is not hiring early enough.”

Q Number one, what's the most common or expensive mistake that you see early founders making?

A The biggest mistake I see early founders making, particularly for BtoB, is doing paid marketing too early. Because if you don't have PMF yet, you're not going to get paid marketing working, and you're going to spend all your time and all your engineering time optimizing funnels and conversion and creating these different creative and ads and caring about your rankings and your, your meta metrics. None of that actually matters, apart from just build an incredible product, and I think the core loop of do incredible launches and build an incredible product is enough for any BtoB or prosumer product to get that initial product market And then you can start testing out different page channels. The caveat though I'd say is if you are consumer, distribution, with B to B, you basically know if you build an exceptional product, you can staff a sales team to it, and they'll be able to scale it. So distribution is not as much of a risk. With consumer, it's very different, where distribution is a key part. You could create an incredible product, but if no one hears about it, then you can't scale it. And so instead you may want to consider Performance marketing earlier, but again, I would try and lead to other channels which also increase brand.

AI assessment note: “The biggest mistake I see early founders making, particularly for BtoB, is doing paid marketing too early.”

Q I have not heard of this. Are we reading? Carcerating Titanic? Like, why do we need motion designs?

A So video is hugely underrated as a medium for growth, and so every launch video we do, the keystone marketing piece we ship with it is the video. There's different types of videos you can do. You can do motion design, which is basically animated graphics. You could do a founder-led video where they're speaking to the camera. You could do a more like screen share style video where they're kind of recording and interacting, and it's very product focused. But why I love motion design is it's abstract, so you can really focus on what are the core value props. You can make it catchy and attention focused and mix in the abstract UI elements, and the big mistake that I see with most people's videos is that they put the founder themselves, they do a full five minute monologue themselves about the company mission, and unless you have the editing skills of MrBeast, you're just not going to keep people for the launch video. So make it sure Get the key message across in the 1:30 seconds, and motion design is fantastic for this.

AI assessment note: “why I love motion design is it's abstract, so you can really focus”

Q No, they do. Absolutely. Sorry. That sounds terrible. He's so good. I like him and Stefan a lot. Um, final one. What's the best growth strategy that you've just been really impressed by in the last, like, 12 to 18 months?

A Yeah. The best growth strategy I've seen is what Brian Johnson is doing. His marketing is so good that they made a Netflix show about his marketing, and I do think Brian Johnson's actually being a Authentic. Like, he really does care about longevity, about creating great products, but he's put himself, one is like, it's just super controversial what he's doing. If you look at the way he dresses, it's like definitely not normal clothes, and if the things he's doing, he's doing like bloodletting and blood transfers with his son. One, he's just generating a lot of controversy such that, you know, we're speaking about him now on this podcast. Two is they've really dialed in the messaging. So, like, the don't die is super catchy, and they've built a brand around it, they've built a community, like, their brand building is just fantastic. And then their distribution. He's constantly on Twitter, like, anytime someone talks about doing a late night, or eating junk food, or pulling an all-nighter to ship a feature, he's there, like, telling them off. He's there.

AI assessment note: “The best growth strategy I've seen is what Brian Johnson is doing.”

Q You mentioned the push in billboards and in podcasts and all these. How do you think about when's the right time to do brand marketing?

A Ideally, you're doing it right from the start. If you're trying to build a massive company, you want people to join you for your mission. And now, how that may look like over time may switch, but right at the start, I'd really be thinking about engaging with your community. Like you said earlier, You messaging every single person on Twitter that follows you, that's you building your brand, that's you as a person being related to. So ideally you have right at the start, and it comes from the founder of them putting themselves out there, being authentic, then building it. And then yes, over time you can scale that into different brand awareness campaigns, and I think you just need to choose that ratio of spend which you're happy with. Maybe that's 20%, maybe that's 70%, I think it depends how well your other channels are working.

AI assessment note: “Ideally, you're doing it right from the start.”

Q Matty, this is for 11 Labs, and it's like, we're gonna be everything for everyone. This sounds bad. Like, I like a tight ICP where we can market to, we know which channels we're selling to, we can craft a product for them. How has a horizontal product offering worked so well when it is completely counter the narrative that we're always told?

A I a hundred percent agree. And actually, when I worked for post-hog, part of the magic was you have one ICP, which is a product engineer, and you're selling multiple products into it. So you're selling product analytics, you're selling session recording, feature flags, all in, and each new product you add, it gives you another entry point. It enables you to, like, integrate the products to do completely new use cases, which couldn't be done before. We don't have any of that in some ways. And so with 11 Labs, because we have the best AI audio models, so turning text into speech, speech into text, then making sound effects for voice cloning, voice isolator. This is like a very horizontal foundational layer, which we sell as an API. So that's kind of your first product, the developer product. And then you can sell that API at scale. So you can sell it to enterprises. You can build around workflows doing conversational AI. So that's fantastic. But then, because the team is super ambitious, you know, Matty and Peter literally want to build a bigger company than Google. They want to build something absolutely massive. And then they see like, oh, we could actually build this to build a reader app for your phone to read articles. So that's now a product. Then we're also building a creator platform. So this text box, which anyone can type in text, that's now creating whole audio books a…

AI assessment note: “the reason we've managed to get it working is because we've started with these incredible audio models”

Q Fucking nuts. Do you worry that the value of video is going down with the explosion of supply? It's something that I do worry about. I mean, we have teams of video people, but everyone is fucking doing video, and I just worry that actually with the increase in supply, it becomes harder and harder to discover.

A Maybe. I don't think we're there yet, and I still think it's the most effective medium in order to get your message across to keep people's attention. People don't want to read blog posts now. Instead, have a crisp video. The platforms also love it and they boost it. The way I could potentially see it moving is, like, Twitter now with their live streams. I'm not sure if you've actually seen any time someone does a live stream, their entire audience has a bright red bar on top of the Twitter screen saying blah-de-blah is live streaming. I could see the shift towards live streaming, but unsure whether that makes sense for launches. Maybe, like, post follow-up to launches is another great way to grab attention.

AI assessment note: “Maybe. I don't think we're there yet, and I still think it's the most effective medium”

Q You mentioned the push in billboards and in podcasts and all these. How do you think about when's the right time to do brand marketing?

A Ideally, you're doing it right from the start. If you're trying to build a massive company, you want people to join you for your mission. And now, how that may look like over time may switch, but right at the start, I'd really be thinking about engaging with your community. Like you said earlier, You messaging every single person on Twitter that follows you, that's you building your brand, that's you as a person being related to. So ideally you have right at the start, and it comes from the founder of them putting themselves out there, being authentic, then building it. And then yes, over time you can scale that into different brand awareness campaigns, and I think you just need to choose that ratio of spend which you're happy with. Maybe that's 20%, maybe that's 70%, I think it depends how well your other channels are working.

AI assessment note: “Ideally, you're doing it right from the start.”

Q Number one, what's the most common or expensive mistake that you see early founders making?

A The biggest mistake I see early founders making, particularly for BtoB, is doing paid marketing too early. Because if you don't have PMF yet, you're not going to get paid marketing working, and you're going to spend all your time and all your engineering time optimizing funnels and conversion and creating these different creative and ads and caring about your rankings and your, your meta metrics. None of that actually matters, apart from just build an incredible product, and I think the core loop of do incredible launches and build an incredible product is enough for any BtoB or prosumer product to get that initial product market And then you can start testing out different page channels. The caveat though I'd say is if you are consumer, distribution, with B to B, you basically know if you build an exceptional product, you can staff a sales team to it, and they'll be able to scale it. So distribution is not as much of a risk. With consumer, it's very different, where distribution is a key part. You could create an incredible product, but if no one hears about it, then you can't scale it. And so instead you may want to consider Performance marketing earlier, but again, I would try and lead to other channels which also increase brand.

AI assessment note: “The biggest mistake I see early founders making, particularly for BtoB, is doing paid marketing”

Q You mentioned the push in billboards and in podcasts and all these. How do you think about when's the right time to do brand marketing?

A Ideally, you're doing it right from the start. If you're trying to build a massive company, you want people to join you for your mission. And now, how that may look like over time may switch, but right at the start, I'd really be thinking about engaging with your community. Like you said earlier, you messaging every single person on Twitter that follows you, that's you building your brand, that's you as a person being related to. So ideally, you have right at the start, and it comes from the founder of them putting themselves out there, being authentic, then building it. And then yes, over time you can scale that into different brand awareness campaigns, and I think you just need to choose that ratio of spend which you're happy with. Maybe that's 20%, maybe that's 70%, I think it depends how well your other channels are working.

AI assessment note: “Ideally, you're doing it right from the start.”

Q Fucking nuts. Do you worry that the value of video is going down with the explosion of supply? It's something that I do worry about. We have teams of video people But everyone is fucking doing video, and I just worry that actually with the increase in supply, it becomes harder and harder to discover.

A Maybe. I don't think we're there yet, and I still think it's the most effective medium in order to get your message across to keep people's attention. People don't want to read blog posts now. Instead, have a crisp video. The platforms also love it, and they boost it. The way I could potentially see it moving is like Twitter now with their live streams. I'm not sure if you've actually seen any time someone does a live stream, their entire audience has a bright red bar on top of the Twitter screen saying blah-de-blah is live streaming. I could see the shift towards live streaming, but unsure whether that makes sense for launches. Maybe, like, post follow-up to launches is another great way to grab attention.

AI assessment note: “Maybe. I don't think we're there yet, and I still think it's the most effective”

Q You mentioned the push in billboards and in podcasts and all these. How do you think about when's the right time to do brand marketing?

A Ideally, you're doing it right from the start. If you're trying to build a massive company, you want people to join you for your mission. And now, how that may look like over time may switch, but right at the start, I'd really be thinking about engaging with your community. Like you said earlier, you messaging every single person on Twitter that follows you, that's you building your brand, that's you as a person being related to. So ideally, you have right at the start, and it comes from the founder of them putting themselves out there, being authentic, then building it. And then yes, over time you can scale that into different brand awareness campaigns, and I think you just need to choose that ratio of spend which you're happy with. Maybe that's 20%, maybe that's 70%, I think it depends how well your other channels are working.

AI assessment note: “Ideally, you're doing it right from the start.”

Q Fucking nuts. Do you worry that the value of video is going down with the explosion of supply? It's something that I do worry about. I mean, we have teams of video people, but everyone is fucking doing video, and I just worry that actually with the increase in supply, it becomes harder and harder to discover.

A Maybe. I don't think we're there yet, and I still think it's the most effective medium in order to get your message across to keep people's attention. People don't want to read blog posts now. Instead, have a crisp video. The platforms also love it and they boost it. The way I could potentially see it moving is, like, Twitter now with their live streams. I'm not sure if you've actually seen any time someone does a live stream, their entire audience has a bright red bar on top of the Twitter screen saying blah-de-blah is live streaming. I could see the shift towards live streaming, but unsure whether that makes sense for launches. Maybe, like, post follow-up to launches is another great way to grab attention.

AI assessment note: “Maybe. I don't think we're there yet”

Q the enterprise discussions that you had earlier moving into enterprise. The big concern with a lot of essentially these Gen AI companies today is that they're sugar high revenues. They're unsustainable. We've seen them both. They hit fifty million an hour. Wow, wow, wow. But it's like, is it real revenue? How do you think about, do we have a generation of companies where it's pretty synthetic revenue, or not?

A You know, it probably varies by product to product, but if they're solving a real problem, and the retention is strong, I see no reason why these aren't real revenues, and the thing, I've spoken to a few VC friends, I don't understand why they're not upping the rate of their investing. They're seeing companies which are now, they're like, oh, it's already on six million dollars MRR, In, like, one year. And I'm like, why aren't you investing on it? And they're like, well, there's this other one which is now on, like, ten million dollars of MRR. And you're like, surely now is the time. Like, the wrong time to be the tiger global play of investing quickly was back in the COVID boom. The exact right time is now, when I think you have this unlock of capabilities with the AI models, which has meant new products which just couldn't exist before.

AI assessment note: “I see no reason why these aren't real revenues”

Q Fucking nuts. Do you worry that the value of video is going down with the explosion of supply? It's something that I do worry about. We have teams of video people But everyone is fucking doing video, and I just worry that actually with the increase in supply, it becomes harder and harder to discover.

A Maybe. I don't think we're there yet, and I still think it's the most effective medium in order to get your message across to keep people's attention. People don't want to read blog posts now. Instead, have a crisp video. The platforms also love it, and they boost it. The way I could potentially see it moving is like Twitter now with their live streams. I'm not sure if you've actually seen any time someone does a live stream, their entire audience has a bright red bar on top of the Twitter screen saying blah-de-blah is live streaming. I could see the shift towards live streaming, but unsure whether that makes sense for launches. Maybe, like, post follow-up to launches is another great way to grab attention.

AI assessment note: “Maybe. I don't think we're there yet”

Q the enterprise discussions that you had earlier moving into enterprise. The big concern with a lot of essentially these Gen AI companies today is that they're sugar high revenues. They're unsustainable. We've seen them both. They hit fifty million an hour. Wow, wow, wow. But it's like, is it real revenue? How do you think about, do we have a generation of companies where it's pretty synthetic revenue, or not?

A You know, it probably varies by product to product, but if they're solving a real problem, and the retention is strong, I see no reason why these aren't real revenues, and the thing, I've spoken to a few VC friends, I don't understand why they're not upping the rate of their investing. They're seeing companies which are now, they're like, oh, it's already on six million dollars MRR, In, like, one year. And I'm like, why aren't you investing on it? And they're like, well, there's this other one which is now on, like, ten million dollars of MRR. And you're like, surely now is the time. Like, the wrong time to be the tiger global play of investing quickly was back in the COVID boom. The exact right time is now, when I think you have this unlock of capabilities with the AI models, which has meant new products which just couldn't exist before.

AI assessment note: “if they're solving a real problem, and the retention is strong, I see no reason”

Q Don't let Matty see this email. Don't let Matty see this email. What do you actually think happens there? Does Lovable become a ten billion dollar company and change the way that we create websites? Does it do Squarespace, Wix, you know, adopt similar processes? What happens?

A I think there's a few ways they could go. Overall though, I'm just very bullish. I don't see a reason that the level of intelligence in these AI models will slow down. And so, given that it will keep rising, I think you're going to see more and more engineering. We're already at, say, 60% of all our code being AI generated. Then you have, like, the broader life cycle of all the testing, deployment, infrastructure. There's no reason why any of that couldn't be all done by AI as well. And so I think Lovable's in a fantastic place to own that core stack. The tricky bit, I think, at the start is, how do you make it super secure? How do you get enterprises to build it? How do you do full apps? And I think Retool is kind of dead at the wheel at the moment. I don't hear the recent things they've been launching, and so I could imagine Loveball going in that direction. You're cringing.

AI assessment note: “I think Lovable's in a fantastic place to own that core stack.”

Q Back to it. No one leaves the office. No one leaves. Don't let Matty see this email. Don't let Matty see this email. What do you actually think happens there? Does Lovable become a ten billion dollar company and change the way that we create websites? Does it do Squarespace, Wix, you know, adopt similar processes? What happens?

A I think there's a few ways they could go. Overall though, I'm just very bullish. I don't see a reason that the level of intelligence in these AI models will slow down. And so, given that it will keep rising, I think you're going to see more and more engineering. We're already at, say, 60% of all our code being AI generated. Then you have, like, the broader life cycle of all the testing deployment infrastructure. There's no reason why any of that couldn't be all done by AI as well. And so I think Lovable's in a fantastic place to own that core stack. The tricky bit, I think, at the start is, how do you make it super secure? How do you get enterprises to build it? How do you do full apps? And I think Retool is kind of dead at the wheel at the moment. I don't hear the recent things they've been launching, and so I could imagine Loveball going in that direction. You're cringing.

AI assessment note: “Lovable's in a fantastic place to own that core stack.”

Q Stanford second year and third year and every class that they did in between. I do just want to touch on one thing that Matty also told me to touch on, which is Fellas, which is the company you started before joining 11 Labs. What are some of the big takeaways from that when you reflect on it and how it shaped how you think about growth and product building?

A I mean, Fela was a wild journey. It's me and my friend Richie, and we just want to do a startup. So we applied to YC, we got into YC actually with an AI developer tool, and we quickly realized we were trying to do, like, browser-based automation, where you automate tasks on people's behalf. So, you know, you're sending an email and it, like, auto-completes it. But this was back Back in 2019, before LLMs were really a thing, the transformer paper had just came out, but like, GPT-Tree didn't even exist. So we got into YC, we tried building a prototype, and we're like, ah, this just isn't gonna work. And a brief detour is, as we realized this wasn't gonna work, Covid hit, and we were there trying to brainstorm the next billion dollar idea, whilst also seeing this, like, impending wave of what everyone thought was a killer virus crossing across the world. So we set up a COVID testing clinic in San Francisco. We were like one of the first people to realize COVID was going to hit. We're following Bellagi on Twitter. And so we're like, okay, how do we actually set up all this testing? So we set up a drive through COVID testing. We were very European. We were like, of course, we're going to do it not for profit. And this is a pandemic. We found a company, which also NYC, which was doing testing of, I think, sepsis, They decided to pivot into COVID testing, we distributed all their firs…

AI assessment note: “one key lesson, which, um, is probably a bit niche, is like, don't do beta tests”