 Hello, welcome to the special CUBE Conversation here in Palo Alto, California. I'm John Furrier, Co-host of the CUBE Co-Founders SiliconANGLE Media theCUBE. We're here at Naveen Chhattah, Managing Director of Mayfield, Mayfield Ventures. Naveen, thanks for joining us today. Absolutely, it's a pleasure to be here. So first of all, I love talking to the guys who are writing checks, funding the next winners. Great to get some insight from you, so thanks for coming on. It's a pleasure. So you guys, Mayfield, have had changes over the past 20 years. We had original founders and then evolve 50 years now. Mayfield, like the blue chip venture capital firm in Silicon Valley. What's that current situation in terms of the numbers, fund under management, how many partners, give some quick numbers. Absolutely. So currently Mayfield, we're just entering the 50th year, next year if you will. Currently we are like six investment partners. We have a team of three partners who are focused on operational services and marketing, business development and talent, and we have an operating partner. So 10 people sought out for running essentially the firm. Currently investing our 15th venture fund, which is over $400 million, and also investing from a later stage select fund, which is $125 million, investing in select companies in our portfolio. Got it, so you guys are looking at early stage, also growth, mostly operational support, which is key these days, rapid growth. It's not like the old days where you do an A round, which is like a C round, and then the B round, maybe C round is growth. Now you get in growth in the early days, early rounds. So that seems to be the standard you guys are doing, right? It's very, very important, right? Like Mayfield has a philosophy of being a people first firm. We get involved early in stages of the company, primarily at Series A, and you need to help the companies. All of our partners have been entrepreneurs, have been execs, they have worked at startups, but then you can only be on so many boards and do so many things yourself. So that's where we compliment ourselves with other operating execs who have done this at successful companies and seen excellence, and can bring those best practices to our portfolio companies. That's also great, now you're being a little bit humble, I'll also add that you guys have good tech chops, you see some, you know the trends and you're on top of things, so I think that's the key as well, you got to know what the tech trends are. So I got to ask you, what's your investment thesis right now for you as a partner? I know you do, certain kids, your other partners would do some health care, some other things, but you're investing in what we love, which is cloud native, blockchain, machine learning, AI, all the cool stuff, IoT. So you're right where we're programming too, so our audience is interested. What is your investment thesis when a startup comes in or later stage growth company comes in? What do you look for? What are some of the investment theses that you're going after? So I'm very focused on people, because my strong belief is people make products, products don't make people, and companies are built by great people who want to change the world. So I start with people and with people, and the people we bet on are climbing new hills, and our job is to figure out, are these going to become real hills, or they're going to be just vaporware, if you will. So the current areas where I'm spending a lot of time are applications of AI and ML, autonomous being a key one, if you will. Cyber security is another one as an application of AI and ML. The second area where I'm spending a lot of time is decentralized web or web 3.0 driven by blockchain computing. Another area is the post-Mordslaw era, which is the second innings of Silicon. People thought Silicon Valley is dead, no more Silicon investments, it's becoming a media valley, the world is changing. So in the post-Mordslaw era, there's a lot of innovation that is happening, and Silicon again is in vogue, if you will. Also spending time on cloud native and microservices because believe the whole tech stack is essentially being redefined. So it's a very exciting time to be an investor in the enterprise space. Well it's a nice thesis because it kind of threads the data, you got cloud data, which is going to basically bring scale and compute. The Silicon kind of fits into that because it's obviously at the edge and also in the data center. And then that's going to enable ML and AI. Yeah, so the stack needs to support that. And then blockchain is potentially the next potential hill, which is not distributed computing, it's decentralized. Yeah, and it's the whole thing people have talked about. It's edge computing, peer-to-peer computing, and it's going to redefine the economics of centralization and it's going to disintermediate some of the bigger players. So one thing that you didn't bring up, and I want to throw this at you because I think this is interesting, is that we were just talking with Google this morning, I was at a briefing because Google Next is coming out. Data, and big data has been a sector, and to kind of know how that turned out. It's now called AI. So a little bit of marketing, but I think AI is legit, no problem. But data as a category, data warehouse, and I think that's kind of older, but it's still a big enough market. But there's a whole new category emerging that's the intersection of data center, cloud, blockchain, and AI. And that has horizontally scalable data. Does that stand on its own in your thesis, or is that just a native piece of the thesis that you're looking at? Yeah, so my belief is, if I look at the last decade, it was driven by cloud, mobile, social, and big data. And in my mind, big data is like crude oil, but when you can do something with it, it becomes refined oil and that's AI, if you will. So to me, cloud, social, mobile, and big data are just enablers of the things you and I just discussed. That sounds like the original SiliconANGLE editorial framework. Totally agree, that's at the table. It's table stakes, right? Like basically we have a lot of data, compute, storage, and memory are cheap. GPUs are available, now you can crunch things which weren't just possible before at very low costs. So you can solve problems that humans couldn't solve using AI and ML. And these new abstraction layers are interesting because when you get software involved with open-source software, you have now more software innovation. How is that impacting, say, blockchain, for instance, one of the things that we are very bullish on blockchain and token economics, we think that's a whole other SaaS-like impact, we think token economics, although separate from blockchain is the technology, is kind of the next SaaS kind of impact. Still is unproven. What makes you so bullish on blockchain? Yeah, so I would say the first inning of the blockchain was all about Bitcoin in 2009. It was seen as another asset, Gold 2.0, if you will. Then came Ethereum in 2013 with smart contracts. But the five, six years of blockchain were all about speculation of Bitcoin, Ethereum. Exchanges were created. You had money transfers, remittance. But now what's going to happen is a new wave of distributed applications are going to get created, where they're not going to be controlled by a big power. So could there be a new Google? Could there be a new Facebook? Could there be a new WhatsApp? Where there's no central authority and there's disintermediation and decentralization and the end people essentially own it. But to make it happen, the world is not ready. What do you need? You need the infrastructure to enable decentralized and peer-to-peer computing, which is based on trust this time. Because what is blockchain? It's an immutable distributed ledger. Everybody understands what a distributed ledger is, but it's immutable, which means you need proof of work, you need consensus to change anything. So we are entering an era of peer-to-peer computing, which is going to be trust-based and consensus-based. And in order to solve that problems in the past, there was a lot of development that used to happen at the protocol layer, like TCP, IP, HTTPS, HTTP, UDP, but those things were never monetizable. You had to essentially go build applications and system software to monetize it. But today, with the way the world is going in blockchain, the way we are seeing it with Ethereum, people can monetize the protocols. So the opportunities are not only going to be the application layer, there'll be opportunities at the protocol layer, there'll be opportunities at the data layer and the developer layer. So if I look at the stack, it starts with apps, but the apps can't be built till you have infrastructure. So it's going to be protocol innovation, infrastructure to help scale it, make it secure. Then what do I do with all the data that is going to be stored in a decentralized way? And then on top of that, what does the whole dev tooling look like? And eventually, there'll be decentralized apps. I think the proof points are there. You're starting to see the developer community, certainly Ethereum interactive developers. That's smart contracts, decentralized application, check signals, see that. But the question is, is, well, let me just ask a question. Define for a second, because you mentioned some, first of all, I love what you said, that's great. Define what you mean by protocol, because the internet protocols you mentioned were based on standards and couldn't be monetized, you couldn't need those to run. Now the internet's an operating system, so I agree 100% with what you said. But explain for the folks out there, what does that mean internet protocol for blockchain or protocol for crypto? So let's see how work gets done on the internet, right? When the browser came out, the reason it had to be built, because nobody could use the HTTP protocol, right? And the only way Netscape and then eventually Microsoft could monetize it is by selling the browser. And then that became free and all the opportunity went to the server side. And that's what led to the demise of Netscape, right? In today's world, if I am a protocol developer, say I create Ether, I can have coins that people can have, not only as security coins, but also as access coins that they can come, use my protocol, and the value of these coins will go up based on the assumption how successful my protocol is. So think- Give an example of what a protocol is. Is it Ether or is it, can I create a protocol? No, Ether is a good example, right? Like of a thing which is happening at the protocol platform layer and there are many, many others, right? So let me give an analog which will be easy. Databases were very hard to use, generally by people. So what happened? A whole industry started around application software which OEM'd the technology from Oracle, from many other database companies and built real applications. That's what is going to happen here. These protocols will be the operating system inside or the database inside on which developers are going to build applications. And this time, for them to support these protocols, they're getting incented because somebody else will go buy the coin and if these protocols succeed, it's a greater fool's theory. It'll become more and more valuable as usage happens. Well, that's token economics, but I think that there's certainly some underbelly fraud going on in the coin market. But monetizing the protocol I think is legit and it's going to happen, I agree with you. I think that's the real value. You have the immutable blockchain infrastructure piece, the token economics with the protocol. That's going to create a lot of advantage. I totally believe that. We are convinced by the data that we see that you said is absolutely going to happen. The question I have for you is as an investor, how do you invest in that? Do you just say buy some tokens? So you're an venture capital investor. You invest in equity. They have security tokens now and they have utility tokens. So you have a dual token market. You got T zero might launch a security exchange and potentially new liquidity. As the managing director may feel, how do you deal with that? Do you look at say, do I put a couple bets out there? Do I take equity? What's your strategy? So at least for us, we have decided we are in an early innings of a decade to two decade long opportunity. So we are sticking to our nets of essentially investing in the equity rounds. That's where the value gets created, help these companies get formed. And as what happened in biotech, IPO's used to happen as financing events. Companies weren't created on that. The same thing is our thesis will happen in this area. You start with equity rounds. You essentially help put teams together. You create the technology. You get product market fit. But instead of an IPO, you go do an ICO, but it's a financing event. It's not a liquidity event, if you will. It's project financing. Well, this is the thing that we talked about. Same thing in biotech. Well, there's interest in it. I want to get your thoughts and reactions. So in New York at the blockchain week consensus event, we had theCUBE down there. One of the founders said, the token is a project financing. You're bringing up another thing, which is another potential funding is company finance. You're a company financer. You invest in people and founders to build the company. We want to be on the cap table. We want to be aligned with them. No, but that's a vehicle you understand. So that's cool. I get that. But now you have another potential opportunity with project financing. So it smells like open source. So a lot of the guys are doing project financing and trying to do company financing at the same time. So I think there's some inefficiencies with that. Yeah, and I think this will be in both places. So it's not just going to be equity financing. It won't just be ICO financing. There is going to be some kind of regulation that will happen because it's very hard in ICO when there's no banker, there's no VC. We're just raising money from general public. Have they even done their due diligence? How do you even value something like this? Well, the game is over. What if there's a fraud? Well, we report on SiliconANG and we've been reporting and then this is the current is that the utility tokens game is over. It is financing and has been ruled by the SEC. The security token is going classic credit investor. That is a securitized investment with the utility upside once the utility capability exists. Potentially two liquidies or conversions, all kinds of things. So that's kind of out there. The interesting thing is that how will the liquidity come out? That's something we're looking at. And how does the team fund itself? Because again, product market fit is day one or day zero still. You still got to go out on scale. You still got to get customers. You have a global framework. You got domiciles issues. So this is complicated. It's very complicated, right? The good news is, as I mentioned earlier, it's still very, very early. We're in the first innings. There was some hype on the prices of Bitcoin, Ether and all the other cryptocurrencies going up. But the real companies and the real applications still have to be built. And what's going to happen is in the future, my prediction is just on a white paper, you're not going to be able to raise an ICO. You'll have to build a real product, get some real product market fit, get some initial customers. And then yes, ICO could be a means of funding, but it'll come with its own bells and whistles. Yeah, you got to show some exit potential on either revenue and or appreciation and some sort of token. I still think there's a great market there. I bullish on both. I think it's a, you have a great thesis over there. Obviously, other VCs are jumping in to see Andreessen Horowitz. I think Excel just got in on the Amazon kind of mining deal. I just saw that go out. So Silicon Valley looking like they've put the toe in the water for blockchain. Yeah, I think like. And you have any coming you can share with us or? Yeah, we have a few in stealth, but you'll be hearing soon about them. So you've put some money into some blockchains. Into equity rounds. Equity rounds, okay, great. All right, let's get to AI. First of all, I could talk about blockchain all day long, it's my favorite topic. AI and machine learning. Machine learning really is the meat on the bone. AI is, you can debate on whether AI truly exists yet. How does the machine learning and AI taking that crude oil, the data and making it scale with applications, how is that going in your mind? Does that tie with the new Silicon model you see? Talk about those two dynamics. Absolutely, so I would say one of the things that has happened already is the AI models from companies like Google, Amazon, Microsoft, there's a lot of research that has been put out there. So the technology exists. And if you look at the basic science of it the neural networks and other things existed 30 years back, existed 40 years back. But compute was very expensive. There were no GPUs, right? There were no ML processors. But today the world has changed fast forward 40 years. Now the critical thing in my mind is which entrepreneurs are able to figure out how to take this raw technology to solve problems which were not solvable before. So I'm very bullish on the applications of AI and ML which makes humans better, right? There's all this talk about, oh, they can do the work my admin does, right? That's not where the real opportunity is. The real opportunity is, can we apply AI and ML because compute, storage and memory are cheap to help with drug discovery, to help with precision- And the data tsunami is another factor too. There's so much data, you need streaming data. You need everything, you need help there. And the machine can do a much better job if you will, right? So early in cybersecurity, right? We can't have humans looking at and staring at like 10th a billion events a day. A machine using AI and ML models can process it and only flag to you the 10 things which matter. So that's where I think AI and ML applications will go. Autonomous is another good one. Healthcare is another one. We just talked about cybersecurity. They could be applications of it in horizontal applications like sales, marketing, HR to make these individual professionals more efficient, more productive, right? And as far as the compute and other things is concerned, again, we are in an early innings. And that's where- We're trying to be an investor. Yeah, and that's where if you look at it, right? Like we are in a data-centric world. We invested in Fungible. Fungible, the founder is Pradeep Sindhu, the ex-founder and CTO of Juniper Network. Yeah, he's awesome. Also, his co-founder is Bertrand. He was the head of all software at Apple. What these guys are saying is like, hey, today the bottleneck is not CPU. Is Pradeep doing a new deal? Yeah, it's Fungible. He's the CEO of Fungible. We backed it two years back. You should talk to him. I love, he's a great person. Yeah, he's an amazing guy. He's so smart. And still going at the age he's going at, right? Yeah, he is phenomenal. So Fungible thesis is simple. Microprocessors are hitting the Moore's law. They can only do so much thing. And what's been the tendency? You throw everything at the CPU. It burst. GPUs are a good example. You offload, floating point processing. You offload graphics processing, AI processing. Performance goes up. You can do more job with smaller space. So what's Fungible trying to do in a data-centric world? They're saying like, hey, if you look at people are running networking, security, storage functions, streaming functions on the CPU, why don't we just offload them? So anything which is data-centric or IO-centric, run it on microprocessor. So we're going to see more and more of this stuff. I think you're right. You mentioned earlier silicon is in vogue again. I think that's interesting. I saw a post by Steve Lasalski. He was a former Microsoft guy. Just wrote a story around how Intel missed a big wave because they were so focusing on what their architecture was. They missed some of the new things. NVIDIA comes right in. So we've been seeing what NVIDIA is doing. I still think there's a whole another silicon generation coming. This is what you're saying. AI on the chip, things are going to be on the chip because the Tesla is still a computer. Correct. We're going to see more of that, right? And I think that AI chips won't just be the training chips where NVIDIA is going. They will also be chips which go into autonomous. They'll also be chips for inference which go into cameras. So this whole area is very, very fascinating where not only will you have CPUs, they don't go away. You'll have GPUs. You'll have vision chips. You'll have sensing chips. You'll have inference chips. You'll have network processors that are going to get built. So the whole area which was written off in the last decade is going to come back. NVIDIA, great to have you on. I want to get your final thoughts on cloud and cloud native. Obviously the IT world being decimated by what cloud's doing. Pretty clear if the folks who have been doing DevOps for years saw this coming. But I think now we're seeing mainstream this tornado hit IT. Yep. Impact to IT is pretty obvious. Automation, DevOps, infrastructure as code, microservice are certainly going to do that. But one thing I want to get your reaction to is that the cloud is horizontally scalable. So you mentioned databases. So that changes the notion of a stack. Yep, it does. Your thoughts and reaction up. Because that is a landscape altering event. Correspondently scalable resources. When the world was stovepipes of vertical stacks. Your thoughts. And I think my feeling on that is I tend to agree with you, right? The world is moving towards scale out, not scale up. And it's moving to hyperscale. Not web scale or not the scale we used to have in the traditional database world. And I think what this will do is what we have seen with Amazon, right? Making compute a commodity. Making storage a commodity. And now we are moving into the containerized world. But soon we are going to go into the serverless world. So our belief is this underlying technology is going to create an opportunity to redefine the complete application stack. And the way software is written. The way the applications are built. And this is not any different. It's already happened. Than what happened with mainframe. Yeah. Going to client server computing. Going to web. Going to mobile. And now to cloud native, right? So I think this pendulum in IT architecture changes every 10 years. And when that happens, the whole stack gets redefined. And right now we are only seeing custom applications. But think about all the SaaS apps you have. How will they run on a serverless architecture? Can you have a different kind of pricing model? Rather than charging per user per month, could it be transactional based? So a lot of new things essentially are going to happen, right? And then you have the whole notion of decentralized computing. Where now you can run computer nodes which are not used. You can build a CDN using blockchain. You can build a new Dropbox using essentially blockchain on unused decentralized nodes. So it's going to be a very interesting time on how this whole hyperscale, microservices, horizontally scalable stuff comes along with peer-to-peer and decentralized computing. Naveen, great to have you on. What's next for Mayfield? Dry powder in the fund? Lots of dry powder in the fund. Only like 35% invested from our new fund and just looking for entrepreneurs. What size checks are you guys writing? It varies, right? Like a seed deal is essentially 250K to a million dollars. Series A's are trending up. They're more from like four to 10 million dollars from us. And we do selective B's which are like eight to 14 million. But the A inflation has been there and it's like four to 10 million dollars from a single VC. And on A they had to product up and running honestly. No, a lot of the companies we are backing, they're paper and pencil ideas. And the main reason, absolutely, the main reason is it's all about people. And if you back great people, they'll build the product and they'll get the product market fit if you will. Well congratulations, Mayfield. I've always known them as a great firm. 50 years old, blue chip venture capitalist. Thanks for coming, I appreciate it. It's a pleasure being here. Thank you very much. Cube Conversation here in Palo Alto. I'm John Furrier, thanks for watching.