95% Faster Video Production with Synthesia | Victor Riparbelli

24:20And that took us towards the direction of the product that we have today. And I think it's actually, to some extent, a fairly classical story of like new technologies, right? There's a new technology that comes up and people immediately go to like, oh, the people who are already doing something related to this technology would probably love to do it or to use this technology. But the real big market is actually democratizing, right? It's like enabling a bunch of people who couldn't make videos before, didn't really maybe care that much about video, enabling them to make video is a really powerful thing.

24:48and so it was just very simply once we launched the product it just took off like crazy I think we had that feeling of like wow this is like really what product market fit probably feels like we're just getting so many people in want to use the product they want to pay us money they want to sign up for like an enterprise contract and there was like the spark of that product market fit which then of course took and kind of molded over time what's the are there historical comps of other category creation companies that went to the existing obvious areas to augment their sets of services, like in your case, going to producers.

25:22I don't know. I'm just trying to think like, is that like PCs going to schools? Like what, what's the, what are the, what are the comps? I have a bad analogy, but it's, it's effectively like the first few years of the mobile era, when everyone was trying to retrofit a desktop UX into this like mobile device and it was janky and it was ugly and no one really understood like mobile native means. And then all of a sudden people like, wow, there's just like this GPS system and accelerometer and all these really cool things we can do. And then it comes out Uber and Instacart and a bunch of other really cool apps.

25:51That's probably the analogy I'd have. Yeah. But I'll say things like cameras and phone, for example, right? Or like the idea of like cheap cameras. It was like a lot of people thought that when we get cameras on our phones, we're all going to be taking photos. Like we're like photographers and like the kind of medium of professional photography would be translated into the mobile era, right? But what happens is when you give this to a bunch of people, have no idea what to do with photos, how to shoot photos. All of a sudden, they take a lot of photos and it's kind of like the format sort of evolves into something very different.

26:22And now we're in an era, interestingly, right? We're all kind of user-generated content, like on purpose, wants to not look polished, right? Yeah, it's like shitty selfies from teenagers. Exactly. The first iteration of Instagram was very much built on like, we help you make your photos look professional. They gave you these filters, which was the one hack, I think that unlocked it because all of a sudden people felt like they're actually like a great photographer because a filter makes everything look great. And then we had this era of like people trying to emulate professional photography. And then it sort of evolved into where we are today with TikTok, which is like big brands, almost like doing videos, like someone sitting in their car, just like talking to their phone because it feels like more authentic and real, right?

26:59So I think it's, you always see, we always see like when, when, when technologies gets democratized in this, in this sense, then people invent new media formats. And they often do things that would be very hard to predict before the moment that it happened, right? I'm pretty sure if you asked McKinsey Consultants five years before Instagram came out, what people would do if they could take great photos, they would not have come up with the world of social media that we see today. Do you have a, share whatever you can, and maybe we just released this after whatever is coming, but do you have a vision for what the what next era looks like here in this space.

27:39And you might be wrong, by the way. Like we could look back in two, three years and say that that was the wrong vision. That is the fun and scary part of continuing to be the category creator. But yeah, where do you see this going? So I think from a macro perspective, what we're interested in, I think this space is like, sure there's a category today of like avatar companies, but I really think there's still a lot of things that's up in there. Like AI video, is that a category? I think like what Runway ML is trying to do as we're trying to do is do very, very different things, right? Like I think Runway is trying to, you know, help people with no budget become Hollywood directors and make really awesome films.

28:18Our vision is actually more anchored in something like PowerPoint 2.0. We want to enable visual storytelling for everyone and we want to do that by transforming more of your existing content as in like PowerPoints, text documents, turning that into video, which I think is distinctively different from like a runway or Sora or whatever, right? So the way we think of this is like, if you think of, you know, any kind of consumer, let's take a young person just to flesh it out even more, right? When they're not at work, how much content do they consume? That is text, right? I'm not, let's leave out text messages here.

28:56Most people consume almost entirely like podcasts and videos, including myself. When I want to learn something, I will usually like kick it off with a YouTube video. I'll listen to a podcast, maybe I'll go on TikTok. And I maybe buy the book on Amazon as like the fifth step if I'm really, really interested in that topic. But in my private life, I mostly consume video and audio by now. I read very little, even though I actually love reading. For most people, that's true. But then when they go to work, it's the opposite, right? They read a lot of documents. They send a lot of emails. It's very, very text-driven today.

29:28And we want to bring that experience of being able to consume every piece of content in video in order that you do in your private life, because that's actually what you prefer. That's what you do when no one is looking. We want to bring that to the enterprise. And so that's the kind of direction that we're moving towards. We think that, well, science shows that video is much more engaging, has much higher information retention, and that's a win-win both for the person consuming it and for the company who also wants the employees to be the most well-informed that they can. So that's the direction we're moving in.

29:56Now to get there, there's a bunch of things that needs to be solved. One of them is overall just the general quality and fidelity of the avatars and devices, right? We've seen the space, everything in AI, of course, is moving incredibly fast right now. I think this is going to be the year of speech and video becoming really, really good. Last year was the year of text with LLMs becoming very, very powerful. Avatar technology in general is still a little bit behind text and maybe even also speech, where we've seen speech that are now more or less indistinguishable from humans. We have that technology as well.

30:33We're just about when this podcast is being released, we're releasing our expressive avatars, which is the first avatar technology where the avatar somewhat understands what it's saying. So you put in some text, right? And if you put in something that's happy, you put in something that's sad or excited, disappointed, then the avatar will actually understand it. The voice inflection will match that. the body language, the expressions in the face. They know what they're kind of acting it out. They're performing it like a human would do, right? Whereas a lot of the avatar technology today is like quite stiff and a bit monotone.

31:08That's going to be a big shift. We're just road testing this with our clients for the last couple of months. That really unlocks a bunch of new use cases where a lot of the use case up until now has been sort of internal facing. This is going to be the unlock, I think, for much more external facing use cases, which is pretty exciting. Because the internal use case today is kind of a stone-faced avatar going through what was a very boring, I don't know, HR document and making that much more interesting just by having video as a different modality. But probably not good for sending out customer emails about a product announcement that shows excitement in somebody's, in the avatar's face.

31:46Exactly. Yeah. I think of this as like, if you use Vice as another, as an example of this, right? which of course is a big part of what we're developing as well. Then before we used very large models to build these voices, which is essentially what's happened with us, 11labs, amazing company doing voice as well. Probably a lot of people will have heard of them. Essentially this trick there is you train on like huge amounts of data and these systems like learn to sound much more natural. The era before that in text to speech was like your Siri or Alexa type of voice. Now that's like already a pretty big market, but it's very utilitarian by nature.

32:23If you're like getting directions when you're in your car, it's fine that it's like a Siri Alexa type of voice, but sounds a bit clunky and a bit robotic, right? But you wouldn't sit down and listen to your Alexa read out an entire audio book over two hours because the voice is like, you just, it's not good enough for that, right? So it's good for these like utilitarian purposes, but it's not good enough to be entertainment by themselves. I think that's what we're going to see with avatar technology this year is that today, they're like the host and they're utilitarian, and like they can tell you something.

32:53Usually there'll be some like visual backdrop to it that they're kind of going through explaining. But the avatar itself is not interesting enough that you would come home after a long day at work and watch a 15 minute avatar video. Like it's not there yet, right? But that's going to come this year. They're going to become good enough that I think people will actually like watching them just for the sake of whatever that they're saying. And that's going to be pretty big. And I think another thing of this, like it's going to break through the uncanny valley, like voice has already done. And that's going to be, that's going to be really, really impactful.

33:20And can you explain for those that don't know what the uncanny valley is? So the uncanny valley is a concept that comes mostly from film. And it's essentially when we as humans, see like depictions of digital humans, essentially. It's kind of like a spectrum. If you see something that's like very cartoony and which is very obviously like not a real video, we kind of like that. Because it's more like fun, it's interesting. It's like a Pixar film or something like that, right? And then as we go further, further up the scale of like realism, we can go quite far up in like computer games, whatever we kind of like to see that but right before it it's it's um it gets to actually break through the uncanny valley as in like you cannot tell that this is a computer generated person that you're looking at in a film for example right before that we really don't like it that's the uncanny valley it's like when it's almost there it's like five you can tell something's off right then we prefer that it's more kind of lo-fi and because then it just becomes like fun interesting thing.

34:18But if it's really transparent, but you can tell it's not there, there's something in us that just goes like, I don't like this. And so the uncanny value is when you break through that, right? And you actually can create content, which you cannot tell is synthetic. I think we're very close. I think with the technology we're coming out with now, we'll almost be there. And by the end of the year, we'll definitely be there. And that's going to be very, very impactful. And I think the TAM is going to like a hundred X, a thousand X once that happens. Hearing you talk about the way that you learn and consume different mediums.

34:51Like I'm, I'm the same way. That's why I started a podcast here because I just think that's going to be the future of how people consume information. It reminds me of, um, our partner, Ilio led this series A of Loom, which got bought by Atlassian for just shy of a billion. And if you ask the Atlassian CEO, know why did he do this? What he says is that his nieces and nephews, they're like 14, 15 years old, all communicate exclusively through photo and video. That's it. It's all like Snapchat and Instagram. That's it. They don't even really text anymore. And to see where they are, they don't even ask each other.

35:31They just go and find my friends. Or there's now like new apps that are just like geolocating people. And, and he was like, look in five years, not even she's going to be in the workforce and all of her friends are going to be in the workforce. And that's going to be the expectation that they have on how they want to communicate just with their colleagues, let alone with external parties and, and, uh, and customers and that kind of thing. It reminds me of this. It's going to be table stakes in not very many years. Like today, hey, it's like, oh, wow, that's nice. I get to watch a video at work or someone took time out of the day to make a great video for the Salesforce explaining like a new product feature, right?

36:11It's like, that feels nice when you're at work today because it's like, oh, wow, that's interesting. That's great. And you remember more of it. But in not very long, it's going to be table stakes. Like, I just don't think you're going to have people reading like 25 page manuals for how to set a product, how to do something. People want it to be audio and video driven. And I think also increasingly, which is kind of very much powered by LLMs, which I think is, that's a force multiplier and everything that we do and all these kind of AI content modalities, they'll want it increasingly personalized.

36:39So for example, if you're a salesperson, well, what if before you jump on a call with someone, instead of having an entry in Salesforce with some Zoom info data and what the company is, what they do, things you need to know before you jump on that call. What if you had a four-minute video of an avatar, super high quality, that just goes through who you're about to talk to, what do they care about, what's some of the recent news about the company, here's a graph showing how much revenue they're doing in five minutes, right? That's unthinkable to do today because it'd be way too much work for someone to actually record that, but that would be super powerful, right?

37:12And you do get this to some extent sometimes in the real world, right? A salesperson might talk to, you know, an SDR has talked to that person before and give you some additional context on it. And I think that's going to be where this stuff really Excel, right? It's going to be when these new media formats really begin to utilize the fact that everything here is generated by code, which means infinite amount of possibilities. You're not tied to something that's hard made off a camera. Right now, we're just still very early. Right now, we are essentially a better way of making MP4 files because we've taken the camera out of the equation.

37:49You can do it from behind your desk, but the true power of this is going once it develops into a new type of file format where the video, the audio we consume is going to be generated just for you on demand. And that seemed very hard five years ago. Now, I think we have the intelligence layer. LLMs are obviously very, very powerful. And with the video and audio tools being developed now, you'll be able to translate that kind of whatever LLM reasons into a really nice visual that goes with it as well. This is a weird analogy, but isn't it kind of like a multimodal grammarly? When I write emails and stuff in Grammarly, which just checks my work.

38:25It then sends me an email summary at the end of every week. And it shows me what my tone was across all of these different emails. And so the technology is just now getting to the point where it can recognize without me telling it what my different tonalities are in a sets of emails. And I'm like, oh, I was in a bad mood last week, you know, or I was in a great mood, whatever it was. But now you're basically taking that text, being able to synthesize it in the same way, and then adding, layering on video on top of that with all of the technology behind that, and then giving that kind of power to the hands of enterprises, right?

39:07For sure. I think, look, up until now, text has more or less been like the only truly scalable way we have like storing communication or knowledge, right? Like if you look at, if you think of like all the world's information today, probably 99.9 % of it is stored as text somewhere, right? Because, you know, if you had to make a podcast about every single piece of information, well, like you would just never do that, right? So right today is like only the most important knowledge probably that's like available in audio and video format. And in the grand scheme of things, a lot of things in the enterprise may be like not the most valuable content that you have.

39:42And so it will never become audio or video. And what's going to happen now is that as we move into a world in which it'll be as easy to create video or audio content as it is to type something out, it's going to be very, very transformational in terms of how we communicate. There's even, there's like a thesis called the Gutenberg parentheses, which the idea here is essentially that when we look back in like thousands of years of like the history of how we've communicated, text is going to end up being just a parenthesis in the middle. And the reason for that is if you think of text as a way of communicating, is very, very low dimensional, right?

40:18Like when you text someone versus when you speak to someone in real life, you communicate so much more in real life or in a video because you have your body language, you have your tone of voice, like all these things, which is why we prefer to communicate in real life. But because we're stuck with text today as the only scalable way of communicating at scale, right? That's kind of like what we're using. But at some point, and that may be after our lifetimes, right? Someone will invent a method or maybe it will actually be in our lifetime, right? It'll be like VR, AR thing, whatever, where you can consume all information as if someone was talking to you or having a conversation with someone, either a real person or an AI.

40:51And once we get to that point, what's the point of text anymore, right? We've always been on this trajectory towards communicating in ways that feels more like what we do in real life. The reason we like to look at pictures is because it reminds us more of what we do in real life, right? The reason that we like video calls is because it reminds us more of how we consume information in real life. And VR is, of course, like the ARs is maybe the frontier of that today. And I think that's a really interesting way of looking at it, right? Because I think for most people, it is true. It's maybe not in every single case you would always want to have that conversation or have being talked to someone who's like you would in real life.

41:28I think a lot of cases you would, right? And so it's interesting to see if, is techs going to be around in like a thousand, 2000 years? I think for most people, if they had a choice, they would probably prefer to watch video and listen to audio. And Josh, assuming that the underlying primitive that is accelerating all of these trends is the AI stuff that we're talking about. Like, are you, do you think we're overhyped? Do you think you're, when you listen to Victor, are you like, fuck, I gotta, I gotta go, fine. Like, like every company is going to transform. Like we're like, uh, we live in this bubble now where it's basically the only thing that possibly could be talked about.

42:08knowing that? Are you like, okay, yeah, we're in a moment of time and there's no way it could get more hype-y than this? Or listening to folks that are pulling the future forward, like Victor, are you like, no, no, no, like this is, yeah, we're not even close to there yet. My sense is that I think there are tectonic plates moving that are going to change a lot of how we think about technology and building software and really just living our lives both inside and outside the office. I think we do have a tendency as humans to overestimate the rate of that change though. So while I think the logical conclusion is everything will be personalized at scale and across every information modality, I think it's also really, really important to take a step back and be able to systematically decompose the steps that we can get there and be very cautious in first principle with every step we go, which to Victor's credit is probably one of his superpowers.

42:59And it's kind of, you didn't talk about it much, but like I think led you through this, you know, a few years of despair trying to raise money for many, many investors and not really getting there. Um, but still having that confidence in that long-term vision and the path to systematically kind of approach that in state. Yeah. Can you talk about raising, like you mentioned that you and Steph were running around with your first million dollars, trying to figure out the early use cases to Josh's point. Can you talk about the journey of actually just getting to that first million dollars? Because the part of the wilderness that we explored was just from that point on, there's still a whole nother segment to the story that predates that.

43:36Yeah. Well, I mean, it was really, it was really hard because we were just very early, right? So as I said earlier, it all kind of started with this research paper that I saw and felt like I saw magic for the first time. And I just got so like obsessed with the idea of how powerful technology was going to be. And so we kind of assembled a team and we sort of build a thesis to Josh's point, right? I think sequencing is incredibly important when you're working with like such a big vision, you can't go straight to the end goal, right? Like that, that will always fail, I think. And so when we went out to try and raise the first round, the premise was basically like, look, this technology is really, really powerful.

44:20We have a bunch of ideas about how we can get to an actual product. But if you invest in us, right? It's going to be a journey of finding that. Like I can't with hand on heart say that I know exactly how this is going to pan out next two or three years. And at that time, I was like 25, 26 years old. I didn't really have any merits for starting this company. I'm not an AI researcher. I'm not even a software developer. I had a bunch of like stuff before this working like early stage startups. By that time, I think everybody wanted me to be the PhD, right? I think if I'd been a PhD on AI research, I'd been a much more inclined to say, okay, this guy is like super smart.

44:56We're giving him some money and see what he comes up with. Worst case, it's a$30 million equity hire to Facebook and we get small Facebook shares. But I don't think that's how people like looked at me and rightly so. I think that definitely wasn't my profile. So we got turned down by I think like a hundred investors or something like that. And at one day, like - Actually a hundred, like actually that many people? Yeah. That's not the reconstructed narrative of how hard it was. We actually found the list of people we talked to in those early days. And I think it was like roughly around 100 in Europe and the US.

45:28And also we were in Europe, right? We're in London. And as much as I love Europe, like the general appetite for risk and like crazy ideas is way, way, way lower than it is on the West Coast. But after like being turned down by a lot of people, we had like one of the names on our list we want to get in touch with, Mark Cuban. Mark Cuban built broadcast.com, right? which there's actually a lot of analogies what we're trying to do here. There's like the first internet radio technology where people are like, why would anyone want radio in this? And that's like the stupidest idea ever, right? But they kind of pushed through.

46:00They obviously did really well. They sold the company for like billions of dollars. And he also had this kind of media and technology, pretty good, that Venn diagram. Stefan found his email in a kind of leaked hack somewhere on the internet, pretty interesting, his Gmail. And we sent him an email just like virtually, even saying, hey, we're working with this. Here's a video of the tech kind of in action. He responded back within five minutes. He was just like, I love it. Send us some more questions. And we emailed back and forth with him for like 10 hours or something until 3.30 a.m. in the morning London time where he was basically like, I'll do the round assuming all diligence goes through.

46:34It's like a million dollars. And I think the learning there was that when we then met him after he had kind of invested, he was like, I actually built a prototype of this tech myself with a lecturer. So he knew everything about the technology. He had exactly the same vision as we had. And so he was evaluating us and the way we were thinking about the business, right? And that was very distinctly different from every other person that we spoke to who had never heard about this technology before, had a very hard time extrapolating where this could go and how important it would be for the media landscape.

47:05And I think it's one of the things, finding someone who shares the vision and is evaluating the team, it's a hell of a lot easier to close around with than someone where you have to first take them through, this is the technology, this is the vision where everything's going to go and this is how we think we're going to get there. And are you guys like huddled up over the computer just like figuring out how to draft freaking email responses and then being like, are you kidding me? You already responded and then you got to go back and do it again. Mark is like, we actually tracked this at one point.

47:33We had a spreadsheet where we would ping Mark with like anything, how quickly he responds. And I think the average was four and a half minutes. Mark is like amazing. He's like, he's so quick to respond. He's a very, very, very smart guy. And so it's been really great working with him. We'll be grateful for him taking a bet when no one else did. Well, it's interesting because if you ask, like we had Joe Lacob on stage at the CEO Summit that we were just doing. And he said the number one thing that he prides himself on and that he coaches the organization on is responsiveness. I thought it was going to be, there was a litany of things that I thought he was going to say.

48:10I did not think he'd choose responsiveness, but I completely get it. Like this idea of someone, And like, what an interesting way to show that you care by just being quick to respond. I think there's something to it. And also like to take it all back to all the video talk, right? Think about how much longer you take to phrase like important emails, as opposed to if you're in the same room with the person, right? Because there's so much less information in an email. So you really think about like, what's my tone of voice? How long should this be? Like all these kinds of things that go through your head when you're like responding to important emails.

48:43But probably if you're in the room with the person, just be easy right yeah josh when you see like um like to to victor's point like not a phd ai person um european company there's like very little precedent for any big companies being built in europe um starting with the technology and then working backwards towards a these are like all of the cardinal sins of venture capital i mean i would have passed myself yeah these are like Like, were you not asking yourself these questions of like, all right, none of this fits any pattern that is supposed to build a big company? Yeah, absolutely. It's why we passed foolishly on the C's and then the Series A.

49:25So we made a big mistake. And what we saw initially, I think when Victor came in, we thought he was exceptional as a founder for a lot of different reasons. But we felt like it was almost a toy. Like we just didn't really, we couldn't extrapolate how this could find its way into it as an enterprise software product. And he really proved us wrong. And what was also remarkable is everything he told me he would do at the seed round and at the series A round, he actually did and more. And that's just really rarefied error in startup world. So yeah, he's, he's, sometimes some of the best companies are the ones that don't meet the criteria people look for.

49:58And I think it's like, looking back now, it's easy to say, but I think it actually was, ended up being our biggest strength that I'm not like a researcher, essentially. Because me and Stefan, my other co-founder, who's also the non-technical in that sense, we just ran around three years just speaking to like thousands of people to really try and build a first principle understanding of what is video, why do people care about video, how do people want to make video. And always with the goal in mind of building a business, right? And I think part of that was also because we felt like this is an era where every professor at a great university was like starting a company, hiring three or four, five PhDs, raising some money for some VCs, building something interesting and getting bought by a big fan company for like 30, 40, 50 million dollars in an acquihire.

50:47And that was kind of a, that was a path that people were sort of planning for. And I think a lot of VCs was like, that's kind of a pretty good insurance strategy. Like if that happens and we invested at like 20 million and they could both afford it, it's like, whatever, right. That's worth the kind of the bet. But we always felt like we don't have that option. No one is going to come and buy us for$50 million because we're, I'm not the one, I'm not the guy that Meta wants, right. They want the guy who's like insanely good at the research and will, everyone will, will be like wanting to work with.

51:13So we always felt like we have to build a business. There's no other path for us to do like a quick acqui-hires, one of the big fan companies. And most of the companies got funded this era. Exactly that happened. There's lots of, I don't know if they were great companies or like these like very interesting research companies that got bought at the time. But because we're kind of forced to build a business, we're also forced to build a product and forced to find customers, which is a luxury. I think a lot of the AI companies that era didn't necessarily have to do because it was such a hunt for talent.

51:42Yeah. We, it's insane. I mean, it's like absolutely insane. Do you feel like those, how instructive do you think those battle scars are to the culture of the company today? Meaning, I'm just thinking about this now, but it's an interesting tension of scarcity throughout the entire origin of the company for three plus years, right? Just like never having enough. And now you're kind of in an era of abundance, right? Like you just raised over a billion dollars. You raised another a hundred million or so from Excel. Like you're the cool, sexy kid on the block now. You're one of like seven European unicorns and AI.

52:24There's plenty of AI technology to go around. There's now all of a sudden people that can actually purpose built focus on moving the platform and the technology forward. There's a name for all these things. How do you live between those two worlds of like, all right, now you're getting pushed and you're in the spotlight, but like your whole origin story is one of, yeah, scarcity, like I said. Yeah. I think it has shaped a lot of the culture. We talked about the first funding round now, right? But the second one was almost even more catastrophic because like the first funding round was like, you know, is either we get some money and we try and do this.

53:00The Cuban round. The Cuban round. Yeah. Like we get some money and we try this out or we don't get some money. We've wasted like, I don't know, six, seven months. The second seed round took me like nine months to raise, like an absolute desert walk. And we were very close to going bankrupt at the time. and I think that definitely shapes the company, right? Because we spent years like really truly in the grind where there was like no one handing out checks for us and saying, ah, just do some cool research and like we'll figure out the business bottom down the line, right? This was like, well, you have no business models.

53:30We're not going to invest and you're not an AI researcher. So, you know, you're not going to get bought by Facebook. So better figure it out if you want to raise some money. And I think we've definitely brought that with us today. One of the mantras we have internally into these years is what we call utility over novelty, which is essentially being obsessed not about the technology in itself but driving value for our customers, right? And it's one of those things that kind of sound dumb if you speak to someone who doesn't work in tech because like obviously if you run a business you would want to like make a product that people want to buy.

53:57But when you work with these like exciting technologies, right? It can be easy to get lost in driving attention and interest and wow moments or PR moments that's exciting or gets people riled up and gets them going. because they don't always create building a great product that people want to use over the long term, not because it's interesting and sexy, but because it actually solves a problem for them. And I think we've always defaulted back to like, of course, you want to utilize the fact that it's a jaw-dropping technology, has like magic, you know, magic thing to it. That's very good for like a go-to-market perspective and driving interest, but it's very important we convert that interest and awareness into actual business use cases, right?

54:39So, you know, what I get the most excited about is like talking to a customer who's like an Italian cement mixing company with like 7 ,000 employees. You've never heard of them before. And the person I'm talking to could not give less of a F about AI or LLMs or GPT style technologies, right? They're just trying to solve, they're trying to do their job better. And we're providing a solution for them to do it. They don't use it because it's AI, because it's cool or interesting. They're just like, I need to get this job done and this tool helps me do it better. and obsessing about those types of customers and those types of users is very important for us.

55:14And that's important because they're also the ones who will help us succeed, right? Doing big, flashy PR campaigns and moments is great in many ways, but I think it can be very easy to get kind of lured in by that. And I think that can be like pretty fatal potholes if you don't obsess over like actually driving utility, and not just doing things for the sake of being kind of novel. So I think that's a very important part of this Enthesa DNA. Yeah, Josh, isn't that like an interesting thing that we see all the time here now, especially exacerbated in this AI world where you have one group of founders that come in that are all technology-based.

55:54They're building foundational models. They're raising$100 million from NVIDIA. They're constantly focused on the research and the PhD-level science of pushing these models to their boundaries, but they're not, they have no commercial orientation at all, right? It's just not in their DNA. Whereas on the flip side, you might have someone like you, which was the knock of you, which there's very few of you actually in this world, which would be like, oh, you're not technical in air quotes. However, the, the, the, I guess the feature or not bug in this case about your background is that, well, you're figuring out customer problems and then working backwards to figure out how to retrofit that technology into it.

56:40Exactly. I think there's something very powerful about deep tech combined with customer obsession. And I often find that in like founding teams of companies, it's like, it's rare to find that combination of the two things. I think to some extent, I'd argue that a lot of the autonomous driving stuff within the last five, 10 years have a bit of the same flavor of that problem, which is like you raise a lot of money, you get very upset about technology, but it's not always driven by like solving customer problems or customer obsession. And so I think there's also, of course, many stories of the opposite being true, right?

57:13You could argue to some extent that OpenAI, which of course is like the fantastical success of the last, well, I think they started seven or eight years ago as well. And they very much started from like a pure research driven thing, but they were very good at then figuring out, okay, now we've built something that's interesting and valuable. Let's figure out how people can use it. So it's, of course, every rule has exceptions. But even when I do angel investing myself today, I generally prefer investing in more kind of deep tech companies. I think it's putting that customer obsession and helping with that is something I love to do with founders, especially those who are technical, that can very easily get get caught up in the, in the excitement about building the technology, right.

57:54Or the vanity metrics around, um, you know, doing something that lots of people find very interesting and very cool and they're very supportive. But when you ask them to actually pay for it, not just once, but like over time and renew their contracts and those sorts of things, um, maybe less inclined to do so. Yeah. Our partner, Ilya says that the one rule of venture capital is that there's no rules. And I think this is exactly why, because there's, um, there are so many ways to go carve out this path. The only rule that I have found is people's ability to overcome hardship more effectively than others, hence the name of the show.

58:28And I think you're kind of the ultimate embodiment of that. So it's, it's incredible to see. Do you feel pressure now? Do you feel pressure now, Josh? Like, do you guys feel like, oh shit, like this thing, like, oh shit, this is like, this is working. Like this is, this is real. Like the opportunity, at least before there was no pressure of expectation. It was the inverse of it. Right now. It's like, you know, if someone's investing, if we're investing at the series B and someone's investing, it's like, there's multiple expectations, you know, there's now expectations of you to capture this opportunity in front of you.

59:04I guess I'd love to hear from both of you. Like, how do you think about that pressure now? Yeah. Of course there's more pressure, right? Like that kind of comes with raising more money, increasing valuation. and a market that's heating up. But I think ultimately, like, sure, there's this pressure and that can be, I think it can be healthy, it can be unhealthy if it's too much pressure, of course. But I think in very practical terms, what it means for me and what I spend a lot of my time thinking about is like, now we actually have a big business, right? A very big, very real business, but lots of customers that are happy they're using our product, which is awesome.

59:41But now we have to kind of balance that with also being very nimble and quick and doing new things and being fast. When you're pre-product market fit and trying to product market fit, you have nothing to lose, right? So you can be very extreme in the way you do things, very extreme in the way you test things, very extreme in the way you feel like MVPs and kind of research the market and figure out like where is your space in it. But once you become a real company, right, you actually have something to lose. Like I can't just steer the ship like 180 degrees the other way and be like, hey guys, you know what?

1:00:13we're just going to do this thing now, right? I think so. That's one of the things that as you grow up, you have the kind of pressure of actually having something that's working in a real business. That's something I spend a lot of my time thinking about. How do we strike the right balance from an orchestrated perspective? How do we make sure we set ourselves up in a way where we get the best of both worlds? And I think that's a very interesting part of that pressure. So I can see the pressures being ultimately more from customers than from investors. Of course, those two follow each other very closely if all of our customers.

1:00:43the customer pressure becomes very high because they don't like what we're doing and they start churning, then the investor's not going to be happy. But ultimately it all comes back to the customer. Yeah. I think that's well, do you feel pressure? Uh, like not like Victor. I think, I think it's lonely at the top, like honestly for CEOs, like it's, it's on you and there's kind of no backstop in some ways. So my job as a VC is really to push when I see potential and that just keeps going up. So I have to kind of keep the pressure on Victor a little bit to make can perform as the best and, you know, support when, when we hit the lows as we inevitably will.

1:01:15But, um, no, it's kind of game on right now. But you don't like go to sleep at night being like, oh, this, this is, this has the potential to be a generational end of one company. Yeah. It's ours to lose. Yeah, exactly. Like you feel that. Oh yeah. Oh yeah. You feel that. But, but isn't it even worse when like, at least when Victor feels it, he can like go put his hands on the steering wheel and like actually affect change. Totally. Isn't it a different type of feeling when all you can really do is feel the pressure. You can apply a healthy amount of pressure, but ultimately there's no way we're going to be able to put more pressure on him than he and the customers already putting on himself, you know?

1:01:54So doesn't that kind of make it a more frustrating feeling when you're like, fuck, there's only so much. There's only so much. Yeah. Yeah. Especially for me, I grew up as kind of a, I just really liked being in control. And And I think one of the realities of venture capital is like, that's just not the case. Like you have to lead by influence and really at any one time, there's probably only one thing that's actually important. So like, can I make sure that that one thing we make progress on either by own or our own efforts through our network or resources, or really just through, um, starting Victor with the right people and, uh, and that kind of anecdotes.

1:02:23For sure. For sure. But it's also what makes it fun, right? I think, uh, if it was easy, then I don't think it wouldn't be as interesting. No, definitely not. Well, I, uh, I appreciate you both doing this. Can I end with, do you have like two more minutes? Is that okay? When we were sitting down at dinner talking, we were exploring this idea of like the founder just having a different fundamental set of motivations than basically everybody else in the world. And by the way, we can strike this from the record if you want. But what you were saying was, because like I was lamenting about, I was talking about this idea about like during COVID, how I sold everything and just traveled out of a suitcase.

1:03:02and I felt so liberated. I just felt so free having so little. And you were talking about how

1:03:12you made a bunch of money in crypto randomly, which I had no idea. Did you know that? I had no idea. And bought the nice watch and all these things. But today, and now you're the fancy company and all these things, but you still... There's the satisfaction that you were describing to me that you get about getting on a Lyft bike in the morning and just biking to the office and seeing all the like nice cars in London that are driving by, just kind of knowing that none of those things are tethering you to what you want to go build. That was a really cool conversation. Yeah. I mean, I think I just always, I'm a very obsessive person.

1:03:50I'm especially obsessed about like fringe ideas. And that can be expressed as like, you know, weird AI research in 2017 that eventually turned into dysia. Or it can be like, creating techno records from the eighties that like no one has heard before. Like I love, I love this kind of like weird, strange idea. That's always been my motivation for building. But I think of course there's also the element of like, you know, money and success is a part of all of our lives and identities. And I think you just go through this thing where like in my early twenties, like definitely like want to make like money, want to be successful and do all these different things.

1:04:23Even though, you know, like a lot of people tell you that that's not the end be of all. and I think for me it was just like it's just kind of been it's just kind of interesting when you then like you attain like at least some level of that then you have to try it for yourself right like you have to figure out does it actually make you happy to have a really nice watch of course the answer is for a week maybe it's fun to walk around with your new watch but then it doesn't really matter and what matters is doing something you're truly deeply passionate about and yeah the example with like driving around the bike is like often when you then get all these things right?

1:04:57I think people get very stressed about it. Like you have a really nice car, someone's going to scratch it, you know, it's expensive. You have to pay insurance for it. And so I think it's just, you know, figure out what you read. I think the best things happened for me the last like five, six, seven years is really being very, very, very focused on what I know I like to do and what interests me and then just not doing anything else to the extent that it's possible, which is also my weakness, right? And I know exactly what I'm interested in, exactly what I like to do and how I want to spend my time.

1:05:31And maybe for some people that's raising expensive cars, like definitely not saying that's anything wrong with that. It's just not the thing that makes me happy. Yeah. Well, dude, it's amazing to see what you all are doing. And I think I speak for both of us. Like, can't wait to see what's next. Are you hiring? Always hiring. I know you're hiring. What are you hiring for? Everything? We're hiring more as every team right now. where? New York, London, Europe, West Coast for some roles. But yeah, I think go to market. We're growing really, really quickly right now. We want great people who can help us build a generational company, get in front of customers, help us sell it.

1:06:15The tech side, we're doing a lot of very, very interesting AI research. We've been hard at work for the last one and a half years, and we're going to drop some really crazy stuff by the end of the year. I think we're at an interesting point in time where we're big enough to do interesting things because we have the capital. We're also small enough that anyone who joins the team now can have very true, deep impact on where we go. So anyone who's excited about that mission, just ping me, find me, Victor at Synthesia.io or LinkedIn, or look at our jobs page and I'll be very happy to speak to you.

1:06:51When you hear the word grit, what do you think of? being relentless, not giving up. And I think more than anything, it's like you have to be an optimist, right? I think me and Steph, my co-founder, and I think Josh knows this from how we run board meetings and everything else is that we laugh a lot. And then I do the worst things are the more we laugh. So I remember when we got turned down by like some VC, which we thought was going to go through and we're like two weeks of runway left to the bank or something like that, right and we're just like i could get off this call and just start laughing both of us just like this is how fucking absurd is that we spent two years of our life and this shit's gonna go into the absolute shitter and that's that's something that's very important for us like and i think grit is i think everyone just finds their own expression of grit maybe for some people it's like running a marathon every weekend to stay focused for us it's like laughing a lot laughing a lot staying optimistic and enjoying the journey but of course being relentless while doing so it's amazing thank you thanks fellas anything else nothing it was great ending yeah thanks man appreciate it guys thanks for having me thanks juvin that's it for now if you like the episode please leave us a review or go back into the archives where we've done more than 200 episodes with some fantastic folks this podcast is a client of perkins production and i'm juvin thanks for listening