Meet the Founder of the Billion Dollar Backed Wayve

0:00 close 1.3 billion dollars. A lot of money. Plus, we want to be the first company to get to robot taxis in 100 cities. That's Amar Shah, co-founder of Wayve, Charm Therapeutics, and a secret company in stealth. He starts with Wayve. And we just have this thesis that if you look at evolution, it seems that the heavy lifting is done not by the sensors, but by the brain. So, we tried to create a demo which which would be admired by the likes of Elon or Demis Hassabis or founders of OpenAI. Elon Musk reached out to us. Emergency forced him to change lanes and apply his AI knowledge to the healthcare space.

0:43 Unfortunately, I got very sick in 2019. Had a very late diagnosed illness which was partly because I was waiting a long time to get a scan on the NHS. And when it came, it was very clear what kind of problems I had. Had to go into the hospital and was there for 2 months. So, the drug discovery uses something called transformers, particularly spatially invariant ones which are quite different from convolutional neural networks. Going to have a drug in clinical trial for leukemia next year.

1:14 He's also a prolific angel investor and invests in some of the most moonshot deep tech companies. My first company was in space technology, for example. I've always been fascinated by space. Briefly worked at NASA. In total, I've invested in about 80 companies. Half through a fund and half personal. And now, he's turned his immense knowledge to one of the world's biggest problems. I would like to eradicate terminal cancer. That's what we're going to do.

1:49 Been trying to get you on for a long time cuz you've got a great name, just like me. But, this is actually the first time we've ever recorded where behind the camera over here, we've got people from our work experience program on as well. So, it's a great new first experience. You're like a guinea pig for having an audience. But, people who don't know are going to find out just how much you've been able to achieve in your career. But, where did it all start off? So, you originally going more down the science route and studying for a PhD. What took you down that road? Thanks for having me. Um pleasure to have a live audience. That's exciting. Uh somewhat true, not quite true because I studied math at university and I did have a job for all my sins at Goldman Sachs as a banker.

2:25 So, I always had an eye on the business world, finance world. I did miss studying, I have to admit. And I did go back to do a PhD in machine learning. I wouldn't say I was intending to stay in academia at any point. It was always an option, but I just knew this was a skill set that was likely to be incredibly useful going forward and it would open up doors either in academia or either back in finance, potentially even in entrepreneurship. I really didn't know which way I would go.

2:55 But, I knew I wasn't ready to decide. So, I would say it was more about uh keeping options open. I also think in this country we decide what we want to do for the rest of our lives far too young. And being in the bank at 22 years old, I just felt like I didn't know enough about the world to commit to this for the next decade. So, going to do do the PhD, yes, it was science, but it was actually a sort of holding position to understand myself and where the world is heading so that I could make potentially a better informed decision and take a little bit more time, especially cuz we're probably going to live 80, 90, maybe even 100. So, um I just felt there was no need to rush that decision. So, that that was more of the philosophy of it. How was it actually doing that? Because I guess a lot of people you were at the bank with would have been thinking, "Okay, next promotion, I need to keep earning more and more money."

3:43 Whereas you got off that train and looked at longer future term options. Yeah, it's a good question. I think since young age I've had delayed gratification uh ability. I also saw the trap associated with the glamour. I saw like when I started saying I was planning to leave or thinking about it, my salary increased like 80%. And I was thinking, well, that's quite a juicy carrot, but if I bite that carrot, I'm always going to have more carrots being dangled. And so and I noticed this amongst my my seniors, people who would have been like me 15 years ago that thought they'd be there a year or two, but ended up there for 15 years. And I just looked at those bosses of mine and thought, they're great people, I like them as humans, but I just did not aspire to their lives and how they allocated their time and just really viscerally realized if I don't go now, I'm probably never going.

4:38 So, I think that was the calculus which made it a really rational decision to leave even though I took a 90% pay cut to do a PhD and I'm very glad I did. One of the best decisions in my life. I guess like with what you've now gone to build is good for everybody else who did it as well, right? I hope so. And so once you did that PhD, I would say you said your thing about doing that is like learning and getting better at things.

5:03 But then how did you decide which direction to go into after that? Mhm. So, it's important to note that purely out of luck, I started this PhD at probably the perfect time. So, the year I started was 2012 and in that year was the first time an Nvidia GPU card was used to train a neural network. And I suppose the rest is history. So, this whole wave of AI Deep learning started there. Generative AI here. Exactly. That's That's what I mean by that.

5:31 Um Yeah, I got to see it. I was in the room when Zuckerberg came to announce Facebook AI research. I saw the forming of Open AI. Um DeepMind were hiring like crazy. There was just so much opportunity we had at the conferences we went to, like incredible parties just for hiring people. It felt like we were rockstars already and we're getting a lot of attention. So, it was really clear to me throughout the PhD program that there was a lot of demand for this talent. So, um again, it would have been really tempting to take a high five five salary about five figure salary job at that time, but we realized actually if we started our own company that was close to little downside.

6:12 I think a lot of people fear that, you know, if you start a company, you could lose everything. In our case, we didn't have that fear. First of all, we live in a quite a comfortable welfare state. It's highly unlikely we'll ever starve. And if our company failed, which was more than likely, we would more than likely also get a high-paying job somewhere. So, the whole risk profile of starting a company looked different to me than what the, I don't know, typical dogma is around it. And on top of that, we thought, well, if we've got this unique position, maybe a once-in-a-lifetime, maybe generation um position of having such low downside, we're not going to do something incremental. We're going to do something as big as we can imagine.

6:54 And so, this is why uh it helped being a bit younger and naive and bright-eyed bushy-tailed. And we thought, given the training we had had and the research we had done, um and the timing, we graduated around mid-2010, so 2015-16, that was just when computer vision, which is computers understanding what are going on in images with what what's called convolutional neural networks, started getting superhuman. So, they could identify things in pictures better than humans could.

7:25 And we we we wondered, where could this be applied? And we thought that actually generally in robotics, this will this will unlock robotics. If you think about most moving animals, birds, fish in the world, they have pretty simple set of sensors, usually a pair of eyes. Whereas at the time, self-driving cars had spinning lasers and radars and and lots of sensors. And we just had this thesis that if you look at evolution, it seems that the heavy lifting is done not by the sensors but by the brain.

7:58 So, simple sensing and very, very complicated, powerful computing is how you move around the world. And this was a contrarian belief at the time. Most of the investors, they took meetings with us. They took us meeting seriously, but I think once we left, they kind of laughed behind our backs because these big companies, Google and Uber, were doing something very different. But I read Peter Thiel's Zero to One, which you probably have here somewhere. And the first chapter says, "What is a belief you have with high conviction that most people disagree with?"

8:34 And focus on that in building your company. And this really ticked the box. And we we realized that we strongly believed that end-to-end deep learning with basic sensors is going to be the way eventually of how robots move around the world. The simplest robot was a self-driving car. Because if you think about to to move a self-driving a car, you need two numbers, which is acceleration and steering. Which is actually way less complicated than even your hand.

9:05 Like the actuation required for your hand has maybe hundreds of variables. We have so many different joints and and muscles. So, as a robot, a robotic hand is actually really complicated. The hard part in driving is understanding what's going on around you. So, it's much more of a computer vision problem than a robotics problem. And computer vision was doing really well. AI really took off. We had a chat GPT moment, and we saw this idea that as you increase data and compute, performance improves. So, over time, we really rode this wave of selling that mesh same message and we at the time when we started we didn't know how long it would take for computer to improve or data to improve but we knew it was going up and to the right.

9:48 Luckily it's happened probably quicker than anyone anticipated and we've just been able to piggyback off that but I think from the point we started we realized directionally that was going to happen. Perhaps we didn't know the rate but nobody really did. So that was the contrarian bet and felt like there was little to lose going for it. So you said how investors maybe were laughing at the beginning as well. And as you said you mentioned Uber's going in this area, Tesla. These huge names, right? But how did you get people to believe in it in the early days right in the early days because you're going against some of the giants of big tech.

10:22 You've got a contrarian view to what everybody else is saying. How did you get people to believe in you and to back those early prototypes? Yes, so what you have to realize is 8 years ago AI was not commonly known about. Most of the general public had no idea what it was and even investors would have had very peripheral knowledge of what this thing is. And and this was the feedback we got once we were pitching in 2016-17 um that there was like a huge education process necessary if anything.

10:51 Uh but there were a select few leaders who you know were the kind of gatekeepers of what makes sense in AI and we realized enough after our seed round what we needed we knew we were not going to have revenue anytime soon. And so in the absence of revenue how do we show progress in a deep tech company? This was the the art we had to sort of figure out. And what we realized is we would keep going for funding every 2 years and we're going to face the same barriers.

11:19 We don't understand stack. Other people are doing something different. Why should we believe this? So we realized our target audience or our target market not as a customer but um uh for our progress was actually the thought leaders of AI. So, we were what we thought about early on is how do we get intellectual buying that what we're doing makes sense. So, we tried to create a demo which which would be admired by the likes of Elon or Demis Hassabis or founders of OpenAI.

11:48 And with their endorsement, maybe people will say, "Okay, these guys are doing something useful." So, in 2015, um Google DeepMind made a a demo where a computer could play Atari games, these these old video games, and they could play them just by looking at the screen and trying over and over, trial and error. And then they were learning policies which were even better than any human could play. And so, this was the first interesting example of uh an a reinforcement learning algorithm learning to do a task better than a human. And we thought, "How can we map this idea onto a car?"

12:27 And so, our first demo was called learning to drive in a day. And we took a car that had no knowledge of anything. It was completely blank-minded. And we turned the car on, it moves. As soon as it does something the driver deems unsafe, it's turned off. That's it. That's the only feedback it gets. So, it's very Pavlovian training. And what we found was after 20 minutes of doing this, it learned to follow the lane. Not only did it follow that lane, it followed other lanes in different weather conditions and lighting conditions. But this was the first This is not a product. You can't sell this.

12:58 But it was the first demonstration that you don't need to tell a car how to drive. You can make it learn how to drive itself. And so, this kind of sparked a thought in many people's minds about what these guys are doing might make sense. Particularly, Elon Musk reached out to us. We met him in 2018. He was fascinated by it as were some of his autopilot team. And subsequently, a co-founder of OpenAI, Ilya Sutskever, invested.

13:26 So did Yann LeCun, uh uh, head of AI at Facebook, and the chief scientist of Uber. So, uh, with their, not just, uh, verbal endorsement, their financial endorsement, we were able to, uh, convince investors to give us a, meaningful series A of, um, $22 million. And, um, so that was the real big first step. And, uh, that was the point we moved to London. So, all until this point we're in Cambridge. And so, yeah, that that's what we did. And we had to do that for quite some time, even all the way up to series C, to be honest.

13:57 And obviously now, is it over a billion has been raised? Yeah. >> an exact number? 1.25, I believe, is the number. Maybe close to 1.3 billion dollars. Uh, a lot of money. So, um, yeah, the last round was, uh, a bill- 1.05 billion last year. From SoftBank. And we weren't actually, uh, trying to raise that much. We were going for maybe 200, 250. And we're starting to get term sheets at that level. But I think what happened with SoftBank is that they realized we're going to keep going every 2 years and having sort of fluffy metrics that don't really show progress. But And they said, "You know what? We've got money.

14:38 Why don't we give you sort of three rounds worth of money at once? And it's kind of an all or nothing investment." Um, so it's exciting in many ways because we think we have like eight to 10 years of runway. So, you can really invest. And this is the massive advantage that people like Google have. Because they have such a massive balance sheet. Waymo's been going for 20 years now. And it's not profitable. So, it's been, you know, they've spent billions, maybe 20 to 40 billion dollars on it. No other entity in the world could really do that. And so, that's what makes it hard to compete with them. We don't have 20, 40 billion, but we've got a little nice buffer now so that we could really, uh, go for it.

15:18 You've got some, um, a real shot now. The downside is, and I think something we don't talk about enough in venture capital, venture capital is is the debt. You have to pay all that back before you make anything. And And if you think about if someone's giving, especially venture capitalists, if they're giving you a billion dollars, they want 10 billion back. So, they're happy for it to die in the pursuit of getting 10 times more. Whereas, if you own that business and it's most of your life, which is my case, um I'm not happy with such a skewed bet. So, you have to be really careful of this misalignment that you can have. Um and so, that's been personally a little bit challenging because it's great on one hand, but on the other hand, um other people on board may push the company into directions that are less comfortable for founders and early employees. There is a small moderate solution to that, which I think also should be more discussed in Europe, which is called uh secondaries.

16:13 So, the way to recreate that alignment is to allow early employees and founders to take a little bit off off the table. So, they sell a small portions of their shares just to be not to be wealthy necessarily, but to be comfortable and not be worrying about bills. So, that now they can really focus on the company and shoot for the moon. And uh this is something the US have been doing for years. It's It's much less common in Europe, but I I I can see it starting. I just was with the 11 Labs guys last weekend, and they're doing this really well, and it makes a huge difference on on outcomes in the long run. And also shows the ecosystem in Europe that actually this equity means something.

16:50 It's real. It's not just Monopoly money. And I think more people in Europe should get that experience to then encourage people to come back into entrepreneurship. So, it's something I'm trying to be an ambassador for across across the board. I think it was second result that bit that people are scared because if not if other people aren't doing it, and they want to take second is like, are they not as committed as other people are? But at the same time is like, if you you take the money off the table, which now means that you never have to worry about money for the rest of your life. You can just keep going purely because you want to have that massive outcome and achieve the mission of what you set out to do.

17:25 And I think, like you said, views are changing, but there's not enough yet evidence or enough people doing it to really change the market. So, when people like 7labs are doing it, it's going to help shift that mentality. >> Yeah, I agree. I think it doesn't even need to be an amount that means you never need to work again. Let me give you an example I'll give you, which I think is quite sad. Some of our early employees, very smart, they were, you know, we have slightly more modest salaries in a startup. They struggled to get a mortgage in London because a bank doesn't care that you have millions in equity. They just want to see a salary.

17:58 And it's frustrating and and I and sadly I think this is why really smart people sometimes favor job at the big tech companies cuz they get the big salary and that means they can get their mortgage. And so, it you just need to get over these these types of hurdles, which doesn't require selling all of your stock, but just makes you comfortable in putting a large deposit where you'll have where you'll have a flat or house so that you can keep working on the startup. And I think you have to get the fine balance where you're not overdosing in secondaries, but you're not underdosing either. So, um but that's I'm sure there are frameworks that have been figured out in the US for this and um but yeah, we just need to be a bit more um smart about this to create alignment across the board between investors, employees, um to to make sure we're all want to create in the same way rather taking early exits because we don't have any cash, which I think it has happened in Europe in the last decade on multiple occasions.

18:55 Hey, hello. Quick interruption to let you know a bit more about Bay HQ. We're the community for high-growth Asian heritage entrepreneurs, operators, and investors in the UK. You can join us totally free at the bayhq.com/join. There you'll get our CEO structure in your inbox every week with this content, events, and opportunities. You can also get access to our free startup fundamentals course by joining. Let's get back to the show.

19:27 So, obviously you ran Runway for the first 2 years before you had to shift away for different reasons. Can you talk about that? Like what made you need to shift away in that period as well? Sure. Unfortunately, I got very sick in 2019 and had a very late diagnosed illness which was partly because I was waiting a long time to get a scan on the NHS. And when it came, it was very clear what kind of problems I had. Had to go into the hospital and was there for 2 months. Luckily, we'd just raised our Series A and so a bit of pressure was was off.

20:04 But unfortunately, I did go back to Runway for about 6 months and then realized I wasn't totally recovering and that was a really difficult period because the company was in a great state. There was no real pressure. We could have just coasted and seen how how I went. And we did discuss potentially changing a role or having a bit more of a break. I didn't think that was for me. So, we decided including my health team and my board that why don't we wind down the role over 6 months?

20:34 Um basically hand over slowly and then I'd have to you know, just depart. I think I wanted to clean break, but it's very tap challenging because it was incredibly rational and you know, one where head and heart are saying completely different things. But I'm glad I did it. I think it was good for me personally and my health. Uh and for the company because I was nervous of taking both things down. And so I think it was the right decision in the end. Uh but there were silver linings. So, in my illness and recovery going through many variations of medication where I had side effects and um intolerances, I started getting interested in medicines and how they worked. And um then we had the COVID lockdown and I was a little bit bored, and so I started studying biology again. This was my least favorite science, honestly. I was more into physics and chemistry, but you know, I wanted to know why these medicines weren't working and why my doctors didn't understand them either.

21:34 And then I realized how freaking complicated biology is. It's like, I think it's going to take us decades if not centuries to understand biology fully. But um I started started learning about how small molecule drugs work and how they bind to proteins. And also that there was a lot of investment going into AI for drug discovery, especially because the cost of capital was really low in 2020-2021 when interest rates were about zero. And so I saw like when you see big funding rounds, they were coming mostly from autonomous driving, but also AI for drug discovery. So, piqued my interest as to why is there so much interest here. And so I started speaking to people in space and eventually met somebody who was at a startup in Oxford uh doing AI for drug discovery. Um I think it had just become a unicorn.

22:19 But he wasn't really happy there thinking that they were not using AI anywhere nearly as much as they could have. And so um I kind of encouraged him to start coding with me. We started doing some projects together. And a few months later Google DeepMind published AlphaFold 2, which seemed to have solved protein folding with AI. And we were pretty blown away. I mean, this innovation won a Nobel Prize this year, for example. It was really a huge milestone for uh medicinal chemistry and biology. Um and we realized this framework, along with what we had been discussing, would make an incredible framework for drug discovery with small molecules. So, we started developing it, coding it. We found people contributing to an open source project which we eventually hired. Before we knew it, we had a lot of term sheets being thrown at us to to do this formally and we kind of just fell into doing another startup in 2021.

23:15 So, um that was the silver lining of of the illness and I learned a lot about medicines, chemistry, protein folding. So, that was a pretty exciting experience. So, it's obviously two very different paths there, right? Autonomous vehicles and and drug discovery. But, underlying, I guess it's the analysis and the machine learning. It's similar kind of methods you're using there. Or was it quite difficult Was it easy to change just transition your skills or was it quite difficult to transition into a new area?

23:44 I think the markets are clearly completely different, but I would say the abstraction has a lot of similarities. How do you turn or the problem, how do you formulate that into something that can be addressed with neural networks and deep learning? And I think that's there there are a lot of parallels in that part of the the world. But, honestly, I I really know very little about the clinical process and the trials of drug discovery and it's beyond me, but that initial part of turning a real problem into something that a computer could solve, I think there were quite a lot of similarities.

24:19 The types of models were slightly different, but relatable enough. So, the drug discovery uses something called transformers particularly spatially invariant ones which are quite different from convolutional neural networks, but we'd been tracking this research even at Wave because once we'd started, we saw transformers being published in 2017 and then BERT which was an initial like kind of a precursor to ChatGPT and these large language models. So, all of all of these types of models were really in the limelight, especially when well, the the culmination was the release of ChatGPT which came just after we started charm.

24:57 So, yeah, we'd been tracking that research. So, I'm not tracking anywhere near as much now. It's just moving too fast. Back then, we were really on top of it. Um and so yeah, there were still so many unknowns and I think that's something I should add that the only reason these were possible were because somehow we found great other people to collaborate with. And even even in Wave today, I probably understand 15% of the technology. And that's what blew my mind because when you do a PhD, which is what I was doing for a while, you really own the whole stack.

25:32 You decide, what is the question I want to ask? What is the experiment I need to do to answer that question? How do I have to write the code? I have to run the experiments, analyze the data. I have to go and present the results in a conference. Um and you do that whole thing as one person. When you're doing a company like like Wave, it's completely the opposite. You have to find people, delegate, and align everyone on the same direction.

25:59 This is the North Star where we're headed. I would like to achieve this. I don't know how to. Well, how do you think we should do it? And really relinquish decision-making, decentralize it, give people uh autonomy and responsibility and accountability. And I think that's what I I learned It was a big challenge for me, to be honest. It's so different from anything I've done. Not something you you really get taught. But I I It just shows you the power of if you get great people together on one mission, it's just really powerful force. And so, since then I've been less um dissuaded by multi-disciplinary problems cuz I think that's where magic happens.

26:39 And even even in nature, you see the most proliferation of life at the boundaries of ecosystems because there's interesting things going on when you mix different types of people or environments or weathers. And so, that's something that's been fascinating and and subsequently I I like learning, so when I meet people at Wave who've worked on cars their whole lives, that's super interesting to me in a very different way to what I'd experienced at university. And so, um yeah, this is probably why I like multi-disciplinary type of ventures.

27:12 And so, with Chart right, the idea is obviously applying these AI models, machine learning to that new area. What happened there? What happened when you built the business and where did it get to? So, the premise initially was to build software where we would build a tool which a pharmaceutical company could use to assess lots of drugs in a computer before taking them to lab. When we were raising capital, our initial offers were from software type investors, typical um VCs, and we realized that we were we had a big gap in our combined group which was a lack of knowledge or experience in creating drugs and that we thought that might hold us back. So, we went back to talk to life science investors.

27:53 Um and for them, they were typically much less well-versed in AI. And so, they saw it as a Wizard of Oz thing. It was like magic to them and they were quite enamored by it, honestly, and we were offered a lot more money, in fact. Um but their condition was that we had to use our tools to make drugs because that is all they knew. And so, um I think it was the right decision for the company, but less so for myself because I didn't want to sit around waiting for clinical trials. At the end of the day, even if you can accelerate the discovery, uh it'll always take quite a long time to show it's safe in a human.

28:30 And that can take years, maybe even a decade. And so, I kind of saw that the world of AI is moving so fast, if I was bogged down in two of the business that was doing clinical trials, I was going to miss out on this AI boom. So, I just I personally decided to reduce my role as an advisor. I I helped get it going and then yeah, the company's doing well. Going to have a drug in clinical trial for leukemia next year.

28:54 Um but then yeah, I started doing much more angel investing after that. Yeah, so obviously you've already invested how many companies now? Or is is it public information? I've in total I've invested in about 80 companies. Uh half through a fund and half personally. And so obviously once you got to that level and obviously with your proof of a wave in particular, you must have a lot of people coming to you who see your background, see what you can add to them as a strategic investor as well.

29:24 How did you go about learning who should invest in who shouldn't? Because obviously you're going to get so much people so many people coming towards you. Yeah, gosh, I still don't know the answer to this day honestly. But when I think about my personal investing, I have many things I'm trying to achieve. Um a small part of it is trying to get a financial return, but I don't think that's the biggest part of it. I think a bigger part of it is to learn and to just be able to follow someone's journey. Maybe in an area that I don't know much about but would like to learn more about. It's also entertaining because I often like these people and just want to be a part of their lives.

29:59 And so there's multiple objectives I'm trying to achieve which make it a bit of a difficult decision. So um but usually it's somebody I look more for because I invest very early at pre-seed and I'm mostly looking for personalities and do I believe in the person and do I like the vision to where they're headed towards and it's slightly less so about what they've achieved so far. And of course we would like to bet on people where we can believe in that vision and their ability to execute and achieve it which includes recruiting others onto their boat and journey, uh investors, customers, etc. And just seeing that all play out. It's it's kind of like a slightly expensive movie you're being to which lasts a decade or more. Um and so and also so I tried to have some diversity of um, types of companies.

30:47 Um, my first company was in space technology for example. I've always been fascinated by space and briefly worked at NASA. And so that was my first uh, first investment. And then things like underwater robotics, you know, quite wild and few more space ones. Um, yeah, all all sorts of things. Can you remember every company you've invested in now? I don't think I could. Um, if if I had the list I'll I'll I could tell you exactly how I met them and how what I thought about them, but I probably couldn't list all of them now, no.

31:18 Is there any that's really stood out that's gone and through really well after you've invested in them? Um, or highlights? There's some that are doing really cool things that uh, I think they're yet to show they're great businesses, but they're doing epic things. Uh, one is called SpaceForge, which is actually based in Cardiff. They actually do what's called the manufacturing in space. So they they send a machine into um, low earth orbit, manufacture things and then bring it back. It sounds wild, but there are some some real benefits but for doing that because you have uh, zero gravity and zero degrees Kelvin. And you can get much more precise manufacturing for silicon or um, antibiotics or things like that.

32:05 Uh, I think it's incredible and yeah, definitely the future. So that's one I I find really exciting to just read about. Obviously investing in 80 companies that's a lot more than most angel investors, right? I think I've got six. So it's not quite your level yet. What's your plan for that? Are you planning to continue angel investing or you how you think about that portfolio diversification and that building that overall picture for yourself? I definitely won't go at the same rate. I was there's this idea in reinforcement learning called exploration and exploitation. Very, very loosely, whenever you're doing anything new, you should spend the first section just exploring a lot to gather information.

32:46 And then once you've got enough information, you can narrow down and focus and exploit your knowledge. You can use this in job search, in dating, you could use this in anything. And so I think I probably went a bit far, but that that first eight is very much exploration. Now I think I'll be a bit more discerning and um be a bit more selective. I was also doing quite small investments, to be honest. I don't think any of them would materially impact me financially. But now I think I'm going to concentrate a little bit more, spend more time with people. I think that eight has given me quite a lot of feedback on what I like, what I don't like. Where was I right about my PCs, or was I wrong? And by the way, even eight is probably not enough to know if you're good at this. There's a lot of stochasticity and luck and timing and so many different factors involved. You can't really tell if you're a good investor for decades, I would say. So now, yeah, I'm probably going to be a little bit more concentrated.

33:43 Have you heard of the secretary problem? Yes, I was thinking about whether I should talk about it or not, but yes. This is one of the only sort of theorems from my master's degree that I use in my life. Yeah, probably the only one. Shall I explain it and you can correct me when I do it? >> you should explain it. Yeah, do it. So, the idea is that whenever you're trying to make a decision about anything, and I've heard it applied to dating for example as well, right?

34:04 You should get a reasonable sample. Let's say if it's 100 people, you you as a the first on a dating app, right? You'd swipe through 33 people and no matter what, you say no. But then you keep track of who's the best person. And then after the best person, you after that is when you say yes to. So in an investing sense, you want to look at a sample of a reasonable reasonably large sample of companies before you say yes to anybody.

34:29 And then you now have your benchmark of what good looks like. And then you invest in the the companies that meet that benchmark. And now you can explain it much better than I can and correct me. >> Yeah, pretty pretty much right, yeah. I was like I felt pressure there cuz I was like, why am I explaining and you've done like 10 times many investments as me and you've got a PhD. I don't think I use this principle accurately in my my approach, but yeah.

34:52 I think that's a big thing why I actually give this advice to founders is that when you're looking for angel investors, it's actually a good idea to look at people who haven't done many. Because they are often a bit more trigger-happy. Is it somebody now tries to pitch you and you've seen you've invested in these companies, you've seen thousands of thousands of pitch days, it's way harder to impress you. Whereas when someone that first starts angel investing, they've often got that like new puppy energy, right? They're like, "Oh, this sounds amazing. I'm going to invest."

35:18 And then after a bit they realize, "Actually, wait, there's 10 other companies doing the same thing." But as in a founder, it can be quite good sometimes to get these people on board, right? Yeah. I don't know. I think you Ideally, you don't want to end up feeling like you conned someone as well cuz this is a long-term partner. So, you want you want to make sure you're both doing something you both will long-term be happy with. So, I'd probably put a little bit of a question mark on that idea. But you're right, I think that is what would happen. But you don't want them to then say, "Oh, I wish I hadn't done that." Then they just, you know, check out. Yeah, I I guess it's still a case of them wanting to invest.

35:53 It's just that they're not as quite jaded yet, right? With the longer you've been in the game, the harder it is to impress you, I think. I think so, yeah. But you also mentioned that some of your investments are through funds as well. And how do you think about that in terms of investing directly into people versus now becoming an LP or investing in funds instead? Cuz I've moved more towards either investing in funds because I get too much deal flow and because of the community, if I keep saying no to every people, I'm going to like me anymore.

36:19 So, if I were to say invest in funds instead, then it helps me manage that side of things. That's true, yeah. Again, I think you're doing the same process when you invest in a fund. You're you're backing the the GPs and the team and their ability to get exciting deals. So, actually the process is somewhat similar. Um the outcomes of funds are have less variance, I suppose. And so, in a way that's nice. I think I view the funds more as efficiency.

36:44 Like I don't have I if I'm doing other things, I don't have the time to see so many companies. So, you're kind of paying for the service of somebody you know, looking at a lot of things. And I think there is a a place for those, too. I think I I I'm um The slight challenge I have is some of funds have s- s- interesting um incentive uh because of the way they're paid, the 2 and 20 type model. And um I think that's something to be mindful of. It's usually aligned with the LPs, but um I think for some founders that can be not the right thing to take a lot of venture money for that reason. So, typically a fund, because of the power power law distribution of outcomes, funds slightly exaggerating, but in like push their founders to all shoot for the moon. And um that may not be the best decision for some founders.

37:37 Um but it it it is probably the best way to operate a fund. And so, I'm very uh conscious of that potential misalignment having been a founder. This These are the only small caveats I have about fund investing. But usually, if you invest um pre-seed funds are usually quite their equity is quite similar to that of the founder in terms of seniority. So, usually these things are mitigated, but but yeah, slightly technical. So, obviously with your background as a founder as well, what's some of the advice you're often giving to the founders in your portfolio?

38:12 Is there any way you tend to direct them a certain way or to help nudge them to try and avoid some of the mistakes you've made in the past? Yeah, it depends on the the type of company. Most of the ones I work with are so-called deep tech, so typically have revenues a long way away and have to do a similar sort of game that we did to keep people excited and and buying into the mission all the way along the path.

38:36 So, a lot of the work is about communication, actually. How do you convey your idea in a palatable way that would resonate with the audience. So, I'll give you an example at Wave, right? So, at the time, not many people know knew what deep learning really meant. How is it different from um uh SLAM, which was the alternative at the time? And we thought about what is our or who is our audience? Typically, they were people in their mid early 40s, GPs of funds, uh still quite young, but, you know, they may have young children.

39:07 And we had this idea to anthropomorphize our idea and make it relatable to them. So, I had an angel investor who was in that age group, and he had a young son. And so, in our one of our pitch decks, the first page of a picture of this boy, and we said, "In 10 years, he's going to learn to drive a car in 40 hours of training. And here's a Waymo car that's had 10 million hours, and it's worse than he will be in 40 hours. And why is that?" And then we talk about this boy learns with a neural network. And when he learns to ride a bicycle, you don't give him a set of rules to follow.

39:43 You He kind of has to wobble a bit, fall over sometimes, and then eventually he'll know how to do it, and he won't be able to tell you how he's done it. You can't convey that in natural language or code. And so, that was like a bit of a light bulb for the audience to say, "Actually, this makes sense. This is why I can see why this approach works." And so, that was that process, it's not I don't know if it's an alpha science, but helping founders to convey their their technologies in a way that's relatable, uh digestible. And I found this hard because being an academic, you learn to be incredibly precise.

40:19 You don't like saying things that could be slightly off. So, this analogy is not perfect. It definitely is an approximation to the truth, but because it helps convey a message so much, you have to be comfortable with slightly loosening the boundaries of truth. And that's that's something I found hard, but I try to help my founders with when conveying their products and ideas. So, I just want to see tech companies because it's so long before you actually make money, right? And you bring the reason you're in.

40:46 How is that especially with AI now and the disruption? How do you feel about when you're missing in these companies where the future is much less certain. Well, nothing was ever certain, but it feels much less certain was in the past. So, when you're making a 10-year bet in 10 years time, I think you mentioned it with Cham about how the world is going to be totally different. So, how how do you think about that when such long timelines on deep tech? Yeah, it's definitely something that's concerning me. I look for particular modes that I think make sense. So, for example, in the AI world, it's very hard to compete with big tech on compute or or data.

41:24 They probably have a lot more than anyone else does. And their cost of capital is lower because they have huge balance sheets. So, where one area, for example, where I look is if is there a deep tech company which has the ability to create proprietary data, potentially of the type that doesn't exist yet. So, one I've invested in is a female health tech company where they're creating a menstrual device which is collecting a lot of menstrual data.

41:52 And the device itself is relatively affordable. It'll be used by people. They won't make that much money off the device, but this data no one in the world owns. So, they're not competing with Google, with Meta. And once they collect this data, they can subsequently build therapies on it and do research on it. So, So is a deep tech company in a way. They're not going to make meaningful profit for a long time, but I can see the moat and that the value in that data in the future. So that's the That's an angle I I like to go go for with with deep tech, but um yeah, and the world has very much changed because now we're seeing companies get to like 100 million ARR in a year, and I don't think we've really ever had that uh before. So it's something I'm finding a bit challenging, honestly, and not sure how to play. The problem with the AI probably is for every one like that that takes off, there's probably 100 that completely flopped. So it it makes it very hard to invest, and I think we had a similar paradigm in the dot-com era, where yes, a small number of people did really well, but actually most people didn't. And I think we might be approaching a similar um situation now, especially with the price inflation of AI startups. The valuations are really high right now, so I'm kind of being very tentative about investing in AI right now. Uh maybe I'm just not good enough. I'm sure there are people out there that know a lot more about how to do it, but I'm tentative, I would say.

43:18 And I know right now you're working on something in stealth, which is related to some of these different areas we're talking about as well. Is there anything you can tease us with and let us know about? Sure. Uh it's in the intersection of hardware and software. Very much has uh the idea related idea to what I've just described of collecting novel data sets that hopefully will be monetizable one day in a very useful way to a society. The thing I really like about it is it's democratizing an incredible technology that um is Nobel Prize-winning, that's a big hint, um but doesn't exist in some many countries in the world yet.

43:55 Um and part of the mission will involve giving free machines to countries in Africa and Asia, um which is uh quite feels quite rewarding for me because I think some of the especially therapeutics one of the my hesitations with it is these therapies will most likely be used by the and the top 5% wealthy people in the world and that's a bit of a shame but I think the thing I'm working on now will be democratizing.

44:24 So you'll hear about it soon I think. And obviously we've having built such giant companies in the past how does that inform how you're building today? Is it you're building very differently to how you did in the past? You've have you been quite consistent in your methodology in the things you think are important? Yeah, I think the main thing I've learned is how to get myself out of a job sort of thing. Before I was doing everything I was doing the accounts, the payroll, I was taking the bins out. I I wired our office for ethernet cables and I've learned how to figure out what's important for me to focus on and delegate what isn't. Also my personal utility curve has changed and so everything I do now is just for the mission and less so financially driven. Of course if I have customers and investors I'm going to do what's best for them.

45:16 I'm technically not an employee of my current company which is strange. I'm just a shareholder and it's my capital actually and so that's been quite interesting. It means I don't have to have a 9:00 to 8:00 p.m. job but I probably do that anyway cuz cuz I want to and I think I've learned how to pace myself actually. I think I've become a bit wiser. It's possible why I got ill was at least in part due to stress and like not taking care of myself. So I'm very aware to be a bit more balanced in my life.

45:49 Also for the small things like I read so many more novels now since COVID and I've noticed how much that's helping my day-to-day work in just understanding narratives or conveying a message and and starting to see synergies in in different activities and prioritizing more balanced existence and would probably purport for people to think about you know, not letting go off off hobbies and other things because I think that you get incredible things in the intersection of your life. There's the famous story of Steve Jobs going to a calligraphy course when he was meant to be studying and that ended up becoming the type script in in in his first computers and it was very attractive for customers. So who would have thought that would have happened, you know, so that I think that that's what I'm learning more about how to, you know, not forget about wider life and make everything melt together to have a We only get one life so you don't want to just be building startups alone. Um I think that's where I've become wiser, let's say.

46:51 I think I've always found that the people who are really good sellers at stages, they're the ones who are the biggest champions of balance because I've always had people at the beginning like, "Oh no, I've got to do this." The people who've actually got a lot further they're like, "Well, actually I should have been way more kind to myself later on and earlier on in the journey." Mhm. And obviously we've got a bunch of business books here. But have you got any recommendations for the novels in the fiction side that you've really enjoyed? All right, I don't know about novels but the book that came to mind was one I read a couple of years ago related to this topic. I'm not sure if it will help founders who are big beginning but maybe it's worth reading earlier on actually. It's called The Second Mountain.

47:30 Mhm. The the high-level premise is exactly as you say like we spend the first part of our lives chasing, we're climbing ladders, we're building our education, careers, status, wealth. Um and then we get to a point where we're doing okay and we think and then we have a bit of a lull, think what's next? And we climb a second mountain which is more about giving back, contribution society to society, family and about other people. And I don't think life is that. It's not all that binary, but I think certainly for myself where I passed that transition a couple of years ago, I just think maybe it's better to have a smoother transition where, you know, we get off this very individual race and start thinking about contribution more broadly. Um so I I really learned got a lot out of that book when I read it. And then before we get to your final question as well, what are the dreams of the companies you have at the moment?

48:22 There's obviously Wave. What's the big mission there? What's the big mission for Charm and for the company you're currently building, whatever you can say? So I just look into that long term of like what impact you can make through the companies you've done. So for Wave, I think the first massive, well, the big milestone coming next year, touch wood, is launching a service in London. But the way we pitched even from the beginning was we want to be the first company to get to um robotaxis in 100 cities.

48:53 And I actually think we've got a good shot. We've driven in over 90, I believe. Um but deploying services is a whole other thing. But we have now offices in Tokyo, Stuttgart, California, Vancouver, and London. Um and I think we've got a really good shot of of getting to that. So then, you know, that's sort of touching world domination. Um so that would be incredible. I think it'll take another 5 to 10 years, maybe more. So yeah, that would be an incredible place to be.

49:22 With Charm, what we were hoping to build is an engine that just turns out drug candidates for all sorts of different diseases. As like a a machine and then each each drug will take years to develop, but we would like, you know, dozens of of drugs to be generated from this system over the next two one or two decades. My last company, I think this is probably going to sound the most bold. I would like to eradicate terminal cancer.

49:54 That's what you're going to do. For that right? So that's three fairly bold missions then. I think the last one is probably the boldest here. What we're doing as well, so we're 250 something episodes in. And in 200 episodes time we get people back again. So that's in 2 years time. 2 years time. Okay. Where do you think you could be in 2 years time and what we'll do we'll have a little TV up here. And we'll play back what you say right now.

50:16 Jesus. And then we can then see like how close did you get to it? You mean for these companies that I mentioned or >> For for I guess what's your next for the next 2 years what's your main Mhm. Um For for the company I'm working with the project I'm working on now, I'd like to have a working system that is deployed in one of the countries I mentioned in Asia or Africa starting to provide value to them. Just one. I think that's what I'd like to achieve in 2 years. Okay, we'll play that back in 2 years time you sat here again. Okay. So going to be tough.

50:50 Time for the quick fire questions now. Yeah. So first one is who are three Asians in Britain you think are doing incredible work and you'd love to shout out. One is a founder I've met a couple of years ago called Bobby Medakra. He is the CTO of Hippo Swipe. They make an airbag for the knee but most excitingly he and his co-founder built their prototypes in the lecture room at university cuz they were finding it a bit boring and I just saw a demo of theirs and eventually became their first investor.

51:23 Found them more investors to um to back them and they end up leaving their university so I feel a little bit guilty. So I don't think his mom likes me very much but um I think he he was a really cool guy to to work with. And then uh I met a really inspiring lady called Harshini who is a student in in high school. 15 year old lady and I saw some robotics project her and her friends were working with.

51:49 And they were looking they'd you know just done this off their own pocket money. Their parents gave them a little bit. The school saw and thought wow this is cool. They gave them a few hundred pounds. And then they were invited to a competition in America and they needed about 8,000 pounds to go. They couldn't raise it. So somehow I found out about it and sent them to to to the wave team and they ended up being sponsored to go. But it was just really inspiring to see a group of very young girls just doing robotics for fun and actually taking it seriously.

52:23 And so that was yeah an inspiring one for me. And I think slightly different track I'd say Rishi Sunak. I think he's uh he had a hard time in his role. I think on the whole he was mostly liked but um you know he was caught between rocks and hard places in his role. I think he's quite eloquent generally kind and thoughtful. Um he's also an East African Asian. So I think he's done well for for our community. He's brand in a way. He's also recently become an ambassador to Prostate Cancer UK which is related to what I'm working on now. So I'm hoping to meet him soon to discuss it. So I think those are three that come to mind. So the first one you mentioned there so they're actually like going viral this morning. I don't know if you saw the story. I'm not yet. So they're building in their apartment.

53:15 Mhm. And obviously what they're building it looks like there's wires coming out of it everything. Yeah. So they're basically reported by somebody who thought they were building bombs. And then like the police knocked down their door and like broke in. Oh wow. And like with like a gunpoint and then they had to like showcase like look no this is we're building something and we've got investors. But that was like a new story that broke this morning. I clearly missed it. He's definitely been stopped on the tube because they used to meet on a platform at I think Waterloo station and build there cuz one lives in one part of London the other lives in the other.

53:46 And they've had people to report like something looks dodgy and they've definitely been stopped there. So, they're probably on on like wanted posters in the stations now, yeah. It's quite wild, but yeah, very bold guys. So, next thing is if people want to find out more about you and what you're building now, where should they go to? Um you'll hear about it soon, I think, yeah. Um we'll be publicizing in um later this year, probably in conjunction with move over.

54:11 Too many hints today. Should they follow you on LinkedIn or how can they Exactly, yes. I'm generally not very present online, but yeah, I will I'll probably post about this there, for sure. And then is there any way that the audience could help you today? Just curious what interests you, which areas of technology do you think are underrated, underinvested in, and why? Um we all know about AI, of course. There's plenty of investment there, but if there are areas of especially science that you think deserve much more attention from society, investors, and entrepreneurs, very interested to hear your thoughts.

54:46 So, thanks so much for coming on. And I thought the weirdest as well is that with what you've built, like amongst all the podcast guests we have, like it's almost intimidating for me just how much success you've had and how much you've been able to build and how you met as well, I should probably say this at the beginning is that you just came to one of our coffee events and you were very casual, very relaxed.

55:06 Then when we added you to the LinkedIn, I was like, wait, like I had no idea about what you built in the past. So, the way you've come across is very humble and just when we have events and I can see how you're interacting with people, especially early stage founders, which you don't have to do at this stage you've been able to grow to, that's really cool for us to see and have that model of people who have been from the community and been very successful who still are able to keep their feet on the ground and really help out the next generation. So, thanks for that. Have you got any final words yourself?

55:37 Appreciate you saying this. I I think it comes from realizing um how how fortunate I have been in in all my endeavors. Like most of the things I've done, I have observed somebody before me do something kind of similar and think, "Okay, maybe that means I can I can do it." And also in all these companies as I mentioned, it was so much a joint effort. Like I think very little was about myself and my skill sets and my knowledge are pretty limited compared to the groups. And I think the only thing I did well is getting myself out of other people's way, but giving them a platform to operate. And um so I realized, you know, everything requires having incredible partners. And um also it's great to meet other people kind of to live through them vicariously because that period was fascinating for me and I'll never get it back. So, I'd love to help people do that part of the journey and make sure they achieve to their fullest potentials. I think there's a lot of negative self-talk, especially in the UK and Europe. We're not as boisterous and out there as our American friends. And so, just making sure people are not underselling themselves, I think is important to me because I think we have a lot of potential and sadly not all of it is realized. So, I think that's kind of what I would like to contribute going forward if I can.

56:57 Thank you for watching. Don't forget to subscribe. See you next time.