VC | E326 | Richard Muirhead, Fabric Ventures on decentralizing the future - Web3 x AI and why does it matter

35:30only one and actually unfairly I forget which one I think might have been meta, only one had submitted any and the questions were well okay wait but this is going to slow down our release cycle but wait what is your mechanism for checking it but wait if you find something wrong what's the outcome of that, what do we do next and plus by the way this is for the UK but I'm kind of a US company is the UK working with the US? Is the UK working with the EU? Do I have to do it multiple times? You know, am I going to give away privacy? Am I going to give away my intellectual property? You know, it becomes quite hard to manage with web to centralized capitalist, you know, generally ad or e-commerce driven, you know, business model comes quite hard to take that and then have this question of managing AI sort of centrally and make it work.

36:28And rather, if we look at the systems that have scaled most effectively, that are actually most secure, they're all highly open and inspectable systems, open source systems. That wasn't always the conventional wisdom, of course. In the first company that I built, we were seen as somewhat of, you know, more than somewhat of an outlier to use open source software in anger in a carrier scale, you know, context. And that was still the case in the 90s and in the noughties. But obviously, we saw this huge shift towards Linux defeating proprietary Unix and defeating Windows along the way. And I think that we can build something much better that way.

37:12And so I think the AI safety thing is a tricky. I think we need to come up with other approaches, more emergent approaches to monitoring and inspection to make sure that we understand the data sets this has come from. and we also understand who's using it. I think also from a geopolitical perspective, it's like, are we not better off winning by out-innovating by our open source approach to things in the West versus hoping that Google and OpenAI won't do a blunder and have Chinese operatives inside their companies? It kind of seems like that's not so far been the case that they've been able to lock things down.

37:54So how safe is a closed system, really? Since what you're saying is it gives a very specific and helpful attack vector for somebody, because you just need to get on the inside of that closed approach. And now it's no longer safe. Yeah, exactly. Look, it's not without its challenges, the open approach, in terms of making things in multiple levels work in a decentralized way and to protect privacy. and to get consensus and deal with governance. There are many challenges. That's what the Web3 world has been wrestling with. But does it feel like it can both be better and safer? It does to me. If you were to steel man the case of open AI, meaning not the company, but open AI development, is unsafe or more unsafe than the closed approach, what would be your arguments?

38:55I think if it's a steel manning, so if what we're postulating is that the open decentralized AI approach is safer and more aligned, the Achilles heel to that is actually kind of the business model, the incentive side. That's what I think is, if I'm steel manning and critiquing it, it's what is the business model. And because I think, and I'll come back to that in a second, because I think there's also an answer and it's part of the, it's an exciting part of why it can be better. But if you go back to all of the other elements, it's like if you, if things are transparent and immutable and inspectable and decentralized and incorruptible and all of these good aspects that come with the Web3 architecture, you know, if you go to, I think it's Marc Andreessen who said, premier several pundits, fantastic pundits if we're worried about what the algorithms are capable of doing but they have biases then we need to worry about the data that we're being trained on and if we're worried about the data being trained on we need to worry about the humans that are making the decisions about which data to train it on and the way in which they make those decisions so we need to go all the way back to the individuals either the ones who have a certain identity and are creating data individually are the ones who are kind of coordinating that.

40:18But go back to those humans. Who chooses those humans becomes a very critical question. And so I think in the closed world, basically, you're not going to know who those humans are, let alone who chooses those humans or how they're chosen. You know, that's not going to happen. You can have that within an open approach. And I think that's kind of the case for open. But if we go back to this question like, well, fine, but we understand at least the business model for how AI can be used to deliver some utility to people and then can be locked up behind an API and people can be charged with that API or they can be milked for advertising revenue or whatever it is, it may come to a similar thing.

41:06And you inject results in the inferences that are paid for. But so we do understand that model. We didn't always understand that model. There was a point at which it wasn't clear whether advertising would work on the internet. Advertising wasn't baked into the system. I'm old enough to kind of remember that. There was a point at which Google hadn't worked out what its advertising model would be. Prior to its IPO, it had to bring different elements together and then, well, it borrowed a bit from Overture and a bit there, and it all came together and took off. That requires work. And I think we see the same thing in token economics.

41:43I was once told by someone who's a Bitcoin maxi, John Pfeffer, that this sort of postulating that token economics was a very broad design space, but that within it, the incentives are aligned in a way that I believe many exciting things are possible and will lead to great economic outcomes. I was told by him, and I hope he doesn't mind me saying that, but he said it to me, and we then debated it, was the worst investment thesis he'd ever heard. But I believe it's true. We've seen it operate at proof of work. We've seen it operate at proof of stake. We're seeing it operate in various, you know, deceptive physical infrastructure contexts, whether it's kind of helium or hard map there, or we've got one called FrodoBots we've invested in, which is gamifying the collection of data in the last mile, or then these Airbnb for GPUs for AR training.

42:43And I think we'll see it in governance. We'll see it in data collection and incentivizing people to share your data in different ways. So I think token economics can be the killer blow that helps us really unlock all of the value of the open approach. But to your question, if I'm still mounting it, That remains the bit that is still unfolding. I'll say less proven, still unfolding. And so therefore, it's fair for us to challenge ourselves in that area. And I guess that there's also that problem of, so when you look at, the very first thing you look at when you look at a startup is what's the moat?

43:25And so how big can this get? And if it gets really big, how well can we protect it, right? In the blockchain world or in the Web3 world, if you as an example say, what makes Meta so interesting, like the Facebook types of models, was very much that you could see that you could create a login on all your users and then you could increase the take rate. But that is not really, you know, the case with that. The whole point with a blockchain system is that you probably you build it with a with a locked in take rate from the beginning. But then that kind of makes it for the early investors that makes the bet the whole idea of investing in that blockchain company a bit less interesting.

44:18Because what you want to see is that you can actually log everyone in and then you can bump up your your prices afterwards. Do you see that? How do you kind of see ensuring that you build a really, really strong incentive for the investors that make sure that you get a technology into what we have always called hyperscaling mode, right? or blitzscaling, which is kind of what have allowed us to get these behemoths in the Web2 areas. It was so profitable to put those early investments in, plow tons of cash in because you knew that once you had it built, you'd have a winner's take-all market. But I guess in the Web3 world, decentralized world, you will have less of that type of economics, or will you?

45:19So, look, I think it's clearly a very good question. I think if you look at the, there was a FAT protocol thesis from the folks over at USP Play sold at Jim and I, and Chris Manisky and so forth. And, you know, look, I mean, it's a good of a way of thinking about stuff, but it was saying there are protocols appearing in Web3 versus a situation with HTTP and SMTP and the kind of first wave of the web. there's actually an opportunity to monetize them because they can be sort of tokenized effectively. And so the argument was that, you know, the kind of layer one blockchains and layer two blockchains can be tokenized and can be, they can build their little sustainable digital economies, you know, through fees and staking and so forth and burning and watering the whole interplay of those different economic elements.

46:14And they can be sustainable and they can build value. And I think we've seen that play out, you know, kind of Bitcoin worth trillions, Ethereum worth hundreds of billions. And there are numerous layer twos, at least for the moment, worth tens of billions. And that's locked in through sort of security and speed and for this, you lock people into using that kind of that particular environment. But it's more of the infrastructure layer. It remains an open question, arguably, exactly how the interplay economically is going to look like between the applications that run on those environments and those platforms, those substrates themselves.

46:54I mean, I think if you look at something like Uniswap and Maker, those are both generating fees. And they, you know, through their utility and then the Uniswap's agile liquidity, I guess, in both cases. But, you know, in the liquidity that is operating with them in the most DeFi primitives that you might want to operate, you might want to use. that creates a network effect that people are not locked into it, but there's a network effect that means that they want to remain within it. But it is certainly, I think, one way I think describing what you're saying is that you shift a little bit from kind of a golden cage to golden handcuffs or whatever, from golden handcuffs to golden cage.

47:39You know, the door has got to be open. You better make sure it's a good looking cage that your customer will rent because the door ain't ever going to be closed. You know, and you don't get the chance to handcuff them to your business. However, shiny and gleaming those handcuffs might be or not, or how sullied and rusted they may come up over time. So I think, you know, you have to think a different way about building a business. And it goes back to the question that folks are trying to work out. you know and we see this with you know all of this sort of social media experimentation how do the the rewards for decentralized social media flow back to to creators and if if the rake is so small or the take is so small for creators currently you know or rather so much has been taken by the decentralized platforms that's got to create an opportunity it also i think needs to happen, you know, at a much lower friction way.

48:39And maybe some of the answer is that you, whether, you know, we're all going to participate in multiple different sort of creative networks sort of simultaneously and kind of in the background. And the same way we will participate effectively and contribute to different financial markets, there's kind of content market, financial market, health market, market for your health data. And you're going to have a, get a little a bit of value from multiple networks in each of these sectors all kind of simultaneously and that will add up to something that works for you the user open question what does that look like from a venture investment perspective you know i think it's definitely there's this disruption going on i think it's going to become easier and easier to bootstrap communities into existence through through means there's mean coins and speculation and and properly sustainable digital economies i guess one thing also folks have looked at is on the social side do you do start seeing kind of social networks that are more ephemeral you don't believe in sort of you know that it's more splintered it's it's more fragmented but it's also more ephemeral they come and go uh at different times yeah therefore as an investor if you're backing them you need to assume that that's the case with the the you know this horse is only going to be good for one race or whatever and and and so therefore you need to think about it in a different way so maybe you construct your portfolio in a different way or of course find something that is a point of infrastructural commonality across them uh that is going to benefit whichever one succeeds you know build the race course rather than pack the horse or whatever you know um but i think it's a i think it's a I think it's a you know an emerging it's it goes hand in hand with that that exciting area of Trican economics that is still nonetheless emerging and you could argue is is a challenge for the open space but you know then it's still a good open question all right Richard uh so now we've really walked down uh down the world of web3 and done a deep dive that I think uh it's been a while since we've done a deep dive as deep as this one.

50:54So thank you for taking us on that, Richard. Now, I want to take us to some of our more normal topics of the European VC podcast. And the first one is a firm building question that we got from Itzen when we said that we're going to do this conversation with you. And he said that he'd love to hear a bit about the operational side of setting up and growing the VC platform that you have inside Fabric. So I'd love to just ask you that question and also add in, as he also says, as Itzen also says, where have you struggled the most in building this up? Look, I mean, I'm only going to succeed at all in building the operational side by finding excellent people to work with me.

51:38Because I discovered actually, I recently did an Enneagram actually to look at my character or personality type and some kind of mix of achiever and enthusiast or enthusiast and achiever and and um uh you know for better or worse and i you know get extremely excited about opportunities but um and uh can can absolutely be diligent but on the operational side particularly where we need to be reporting diligently every quarter and on top of our finances every month and and obviously on a deal-by-deal basis there's a lot of you know it's specific you know this process we go through in the in the back end uh we've got to have a team to do that so i thank everybody at fabric who's doing critical and doing that from you know gian to max to to um on kind of more investment side to audrey and michael and so forth but So that's one part of it.

52:38I think that my experience has been that I'm still hunting for truly fantastic service providers to help us in this space in Europe. And I would love to hear sort of the contrary, but I don't hear raving reviews of any of the service providers on the fund admin side and so forth. I have some people I should connect you to. That would be great. And like, I think it's ironic still that the, and obviously we have specific requirements. we have to deal with with with you know a variety of financial instruments quite a high pace of investment a large number of lps we have like 300 plus lps you know and we're also making sure that we are compliant you know and i i we i would argue my observation is that we are you know more compliant than any other web3 venture firm have come across in terms of our rcml and so forth and both sides, both on the investor side and the target side.

53:36So all of that is definitely an overhead and we've taken it in-house. And of course, ironically, not just for a sort of software investor or technology investor, but for one kind of cutting edge of software where there's innovations taking place in how you handle data and how you handle kind of finances, you know, we've embraced some of that. You use the stable coins, for example. there's no great software out there for administering all of this either you're still left with with with software systems that you have to bring together so i think all of that has been you know a challenge i would i would love fabric to be of the scale that we collaborate with other firms to help you know build yet better software in the space and yet better service providers but we're not quite at that point that we can let's say focused on our knitting but so that But if that was the question about operations, I would say that has definitely been a, you know, remains an area of hard work for us.

54:31And I think it's an, I don't want to whinge, but I believe it's a higher burden across the UK and Europe than it is across other parts of the world, including the United States. And I think, you know, there are already other things that are a challenge for us. And it's a shame that that's an additional friction. Yeah, absolutely. So now I want to ask you about your three biggest learnings in venture. One of the learnings definitely is that you've got a lot of tensions in being, I think, a great or building a great venture investing firm. One of them would be you have to think very long term to identify the growth successes and hold on and back those winners and so forth.

55:14But especially in Europe, there is still a need to deliver short term results. obviously venture capital formation, venture capital fundraising is particularly a challenge right now due to the macro and the overweight situation that LPs find themselves in kind of private positions. But especially in Europe, that's a challenge. And so you're wanting to think super long-term, go deep into the kind of gen curve on losses individually and across the kind of fund and then come out screening out the other side. I saw the other day, what is it like? Microsoft has gone up 4 ,000x since its IPO or something.

55:51And you want to do that because there is a tension there between long-term thinking and short-term thinking and actually building the ecosystem we want. So that's one of the things is learning how great that tension is and how to manage it. But guess as much as you can through relationships with your LPs. And maybe it's related to my second learning, which is results do count. and I always suspected this, but brand masses, you know, at the end of the day, you know, the thing that you're seeking is that, and I think this is really the case, is that you should not really be hunting for the best opportunities and the best founders.

56:30You need to get to a point where the best founders are seeking you out. And you shouldn't spend too much time then analyzing to death, especially at this pre-seed stage. stage, you need to have a very strong feeling very rapidly that you want to work together. If you're doing too much analysis versus synthesis, you're just not prepared enough for rooming. And from a portfolio perspective, you're not spending your time in the best way. You're not going to get greater clarity and conviction, and you should not be trying to achieve consensus across the entire group. You should be moving forward with that partnership, and you should be seeking the next one.

57:12And I guess if I am still keeping it to three, the other thing that I felt very much when I was a founder and that I understood is that it's famously, I think it was Fred Wilson who said this, but you know, venture, you've got this situation where building a startup that is venture funded, you've got this kind of connection between the venture business model and the portfolio business model. They're kind of actually intertwined. And the kind of counterintuitive thing is that actually a good venture investor, on each individual bet that you're placing, you want to take more risk than any right-minded founder really should be taking individually themselves and what they're building.

58:01And that's arguably because the founders they're looking for are not analyzing it from a risk perspective. it. They're cold to do it. They're only building it. I would say it gets you back, it gets us back to where we started, where you said we're looking for the right type of crazy, right? Because as a founder to take on the type of risk that a VC takes on 30 times across a portfolio, that is, that requires the right type of crazy. Now, I'd love to ask you advice to young people in the industry because you've really built something with fabric and you've been with Axel, you've been with Open Ocean.

58:39So you've taken a trail that I think many would love to learn from. I feel like I'm very much still on the trail somewhere up the hill checking how much water I've got and how much tucking into my kind of carbon loading snacks. But so I think, you know, I think I mentioned before in the context of more kind of the back office and everybody's contributions to the back office. But, you know, you've got to find your team. It's a long journey. There's going to be challenges, fractious moments. Everybody's tired and, you know, personal issues arise to them. You've got to find, you know, you've got to find that team.

59:23And if you are trying to find it, you know, don't go on sort of trying hunting sort of something shiny and gleamy. gleaming, you'll probably find that the team you need and one is already known to you and you've been working together. And if that's not the case, then probably you're not ready yet to go do it. And that can mean, I guess, go solo. And so if you can be solo, then do that. And there are other aspects of course, to speak about server GP, so there are other team members I've seen, like, I can say from the LP angle to anyone listening in, like, You'd be surprised how often we diligence first-time fund teams to realize that we in a diligence process where we reference check, uncover things that the other partner in that firm didn't know about the other party.

1:00:15It's like, you know, it doesn't necessarily have to be a deal breaker, but it's like, that's difficult to come back from. And the big LPs that will have signal value and so on, they will do that work. So it's definitely something that you really want to think about. I think, unfortunately, we have EIF as a very important party in Europe that really does not like solo GPs too much. But it's like, in many cases, that's a better decision. And I would rather bag a great solo GP. Yeah, it's a better risk. We haven't got it at least this time, but I think there's definitely a lot to be said about, you know, what is the state of the venture capital ecosystem across Europe and how, you know, we're still sitting, got a bit of 40 year disadvantage to other geographies.

1:01:08And, you know, I still think that LPs and congrats to you, it's still, you know, the lack of LPs is still a major, major issue, you know, that are native to Europe. But I mean, you know, other stuff I think for kind of emerging managers is, you know, I mentioned before we ended up, yeah, looking for something that's the right side of crazy but a focus around web 3 something that was away with timing and where we could be specialized and distinctive i think that's you know and there's data to support that you know get to cambridge associates and look at emerging managers should be part of the mix specialized managers should be part of the mix if you're an lp if you're building a portfolio and so that makes sense i think and it's more tractable generally for us you know for you as an emerging manager and we see great ones i love what Nathan does with, you know, Air Street, Christian's down over 2150 or whatever, you know, there's a lot of specialized ways of going.

1:02:02And then, you know, anchor is really important. You know, you need your lead as a founder who's building a venture-backed company. You know, you need your anchor, you need that relationship. It's going to be pivotal. It's going to shape you in a lot of ways as a manager. And so, and the things you do before that, like the team you have and your your theme and your your edge it's going to be about less about what money you can make for them but all about how you can make a match with them um and so i think that's worth about worth thinking about and then that will unlock a lot if you've got that in place and you really focus on that then that's likely to you know give you the kind of the the impetus to be able to get to your first close and beyond i recently did an episode speaking of that finding your edge and yes finding your edge, but also finding the right, the wave to ride.

1:02:54I think that's actually something that's a little underappreciated by emerging managers. I think that if you are someone who can really build and who can invest and are good at that side of the business, you can actually, instead of going and saying, okay, I'm going to build a firm that's generalist, well, maybe do the work and find find the vertical and the technology and so on where you're seeing that we should actually have an emerging category here, that there's room for a small firm to build in, which I kind of think is what you also described around Fabric. Well, yeah, exactly. I mean, one thing I was worried about this morning actually is that at Fabric, we are happy that we feel like we saw this kind of intersection of distributed, accelerated computing with self-sovereignty and identity and own your own data and value and so forth with AI.

1:03:49And we saw it well in advance, arguably as much as like 10 years in advance. But you don't want to get too self-congratulatory and think like, oh, great, you saw it because, you know, if you're not careful, the time for outsized returns in those areas might have passed. So what is the next thing? What is the thing that LPs are not hearing from other GPs. What is that? I think that's a really good area to focus on. And you really need to push yourself constantly to evolve and make distinct your thesis. Do you do that continuously or do you tend to do it more when you revamp the thesis up for a new fundraise?

1:04:25Do you have a cycle? Well, there's a difference between what we would like to do and hope to do and obviously what we end up have ended up growing but um but i think you hit a key point actually in the last you know we have kind of get together like not uncommonly on a kind of quarterly basis you know in a proper oscillate type structure and you know in person and something i've been kind of trying to drive through the organizations we we talk about ourselves as being kind of having a core dna as ex-founders you know we're founder-led we're research uh driven and we're community first we We were community first before as we were kind of playing around with the concept of being part of Web3, where, of course, community is extra critical.

1:05:09We've always seen ourselves as a collective. And on that research-driven side, it's almost like as well, it's a way in which I'm hoping that we kind of manage ourselves rather than kind of I manage the firm, is that if we research in public and we engage with our founders and our advisors in a constant cycle of writing and testing and exploring, that process, you know, re-energizes and refreshes the thesis on our ongoing basis, like you described, and you just bake it into the way we operate constantly. But it also, it's kind of not, it's not extra work outside of the work of encountering the founders and having them be interested in working with you.

1:05:52It is that work because you're talking to them about what's going on and you get introduced to the next person. You should talk to the next person. So, but, you know, look, it's not surprising. Not only is obviously venture a solid, you can earn a solid wage, or you can earn extraordinary amounts if you do very well. It's incredibly exciting and vital to the economy of Europe and for society itself. So my point is, as a consequence, it's very competitive. There's a lot of people who are kind of piling into it to be involved in it, hopefully in balance with a lot of people founding companies. And so it doesn't surprise me that the work that you have to do to apply yourself to think about how to be research-stricken, what does that mean in terms of your processes, how do you incentivize everybody, how do you coach everybody, how do you work with everybody internally, that takes work.

1:06:46Because there's no surprise, because you've got to try and do it better than anybody else out there. And if you look at Harvard Business School case studies, probably dating myself and talking about it in that sense, from the last century. You look at Harvard Business School case studies. You did that when you referenced remembering something from 2009 because I definitely do not remember anything about venture from 2009. So that's me dating myself. Well, I was reading the Harvard Business School case studies and stuff around both of the web and the internet written by John Hegel, McKinsey, and the various other.

1:07:23But anyway, if you look at them, you see firms that are successful because they've really refined the approach in a particular area. And they've put a lot of effort into it. And I think given the nature of venture, thinking about how do you stick to a DNA that your founder led? How do you be truly research-driven and make that really work for you? And how do you scale being community first? That's something that we're pushing on hard with, you know, and our head of the product and platform, you know, that all takes real work. It's not going to happen by just like cutting a few checks and sitting back.

1:08:00Yeah, I think you're completely, completely right on that. I always have the, like the thesis that if anything is, the more sexy an industry is, the more competition there's going to be in it. And for that reason, you should also just automatically assume that venture, you're not going to get rich here. unless you're good enough to make crazy returns on carry. You know, or you can get rich if you spend a ton of time, but it's not an easy job. You can't do 40 hours and then get rich on it because there's a ton of others doing 60 hours and 70 hours and they will be better than you. So I think for that reason, it's kind of weird when you meet people that want to do venture, but they're not willing to put in the hours and put in the years where you're not going to make much as an emerging manager, as an example, or as a venture partner, side hustling too many hours next to whatever gig you're doing already.

1:08:55So just a few notes on that for me. Richard, thank you so much for joining us for this episode. It was an incredible deep dive. I've been looking forward to talk to you about something that I think is going to be like in five, 10 years, everyone's going to be like, why did we not all see that coming? Why did we talk about it as something that could go away? Because I think it's going to be absolutely fundamental. Thanks so much, Richard. Nelly, thank you, Adres. A real pleasure. I hope to speak to you again soon. And to everyone tuning in today, if you did enjoy this episode, please make sure to go on eu.bc and subscribe and also go on X and follow Richard because Richard and the team at Fabric does quite some egg storms.

1:09:42And we will link to the most recent one when we put out this episode as well. So make sure to follow us as well as Richard. I hope that you enjoyed the content. Here's a few words from our beloved sponsor. This episode is brought to you in partnership with Zero 100 Conferences, which organizes networking events connecting LPs and GPs in private equity and venture capital firms across Europe. This will be the fourth and final chance in 2024 to connect with investors from the largest firms worldwide following successful events in Vienna, Amsterdam and Prague. Their upcoming event, 0100 Conference Mediterranean, will take place in Milan from October 28th to the 30th at Palazzo Mezzanotte.

1:10:24Attendees will include major LPs and GPs like the European Investment Fund, Taikao Capital, Vencap, Arcano Partners, Amandi Alpha Associates, P101, United Ventures, Merseyside Pension Fund and many more. Save the date. October 28th to the 30th at Palazzo Mezzanotte in Milan. This was their final show. Tear down this wall. It's more than just an alliance. This is a union of values. United and determined, we can serve as a model for other regions of the world. The nature of a problem requires a European response. Europe is a story of new beginnings. New beginnings. Let's start acting