Behind The Tech with Kevin Scott - Mike Volpi, Partner at Index Ventures

Episode Date: May 14, 2024

Mike Volpi is a longtime venture capitalist who joined Index Ventures in 2009 to establish the firm's San Francisco office and North American operations. Prior, he was Chief Strategy Officer at Cisco, overseeing a run of acquisitions still studied today as a model for technology merger strategy. Mike invests primarily in artificial intelligence, infrastructure, and open-source companies, and currently serves on the boards of multiple companies including Aurora, ClickHouse, Cockroach Labs, Cohere, Confluent, Covariant.ai, Kong, Scale, Sonos, and Wealthfront.     In this episode, Kevin and Mike discuss Mike’s early childhood, how he got interested in the study of engineering, and his career experiences—including what led to Mike’s long career at Cisco and his current Partner position at Index—including his board experiences with multiple companies.     Mike Volpi | Index Ventures  Kevin Scott    Behind the Tech with Kevin Scott    Discover and follow other Microsoft podcasts.    

First of all, I'm okay being wrong. And by the way, I'm wrong all the time. So having compassion for yourself for being mistaken at times, but also continuing to have the self-confidence to say, I believe in my own logic, I believe that this is going to happen and that allows you to take certain chances. Hi everyone. Welcome to Behind the Tech.

I'm Kevin Scott, Chief Technology Officer and EVP of AI at Microsoft. Today, tech is a part of nearly every aspect of our lives. We're in the early days of an AI revolution promising to transform our lived experiences as much as any technology ever has. On this podcast, we'll talk with the folks behind the technology and explore the motivations, passion, and curiosity driving them to create the tech shaping our world. Let's get started. Hello, and welcome to Behind the Tech. I'm co-host Christina Warren, Senior Developer Advocate at GitHub. And I'm Kevin Scott.

And today we are bringing you an interview with Mike Volpe, who is an investor and tech leader, and he's been a part of nine successful IPOs, including Confluence, Elastic, Aurora, Sonos, Slack, Pure, Zuora, Arista, and Hortonworks. Wow, that's a mouthful. Yeah, it's a lot of cool stuff. No, I'm super looking forward to this conversation because, you know, we rarely talk to people, or at least I rarely talk to people. You talk to people like this all the time. I rarely have an opportunity to, you know, hear from people who have this level of experience and investment in the tech world. Yeah, I mean, Mike is super interesting in that he had such a long career at Cisco when Cisco was really at the center of the tech world.

I mean, Cisco is still like an extremely important company, but it's just hard to overstate how important and central Cisco was as we were going through the internet boom where, you know, in a very, very, very short period of time, you had to figure out how to birth the modern internet to get everybody in the world connected to it and to get every business that already existed a presence on the web, and then to build all of the new businesses that the web created the opportunity for. And so Cisco, at one point in time, was just a fundamental part of the story that you were telling on all of those things. And Mike was just dead center of all of that at Cisco for years. And I think it's particularly interesting thinking about how that internet boom and this massive infrastructure build out and this massive scramble to go create a whole bunch of things that never could have

existed before back then parallels what's going on right now in AI. And like, I think him as one of the most successful investors in AI, you know, is no accident. So I'm just super looking forward to this conversation with Mike. I am too. Let's get into it. Mike Folpe is a partner at Index Ventures, a venture capital firm. He co-founded the firm's San Francisco office and North American operations, and he invests primarily in artificial intelligence, infrastructure, and open source companies. He spent over a decade at Cisco Systems where he led acquisitions and built a strategy for mergers and acquisitions that's used as a model throughout the tech world. Mike studied engineering at Stanford, where he earned a BS in mechanical engineering, an MS in manufacturing systems, and an MBA.

He serves on multiple boards, including Aurora, Confluent, and Sonos, and he's a non-executive director on the board of Ferrari. Mike, thank you so much for joining us on the podcast today. So great to be here, Kevin. Thanks for having me as a guest. Yeah, so you and I have known each other for a while, so I have been especially looking forward to chatting with you today. And, you know, let's just start with how you got interested in science and technology in general, because you had a really interesting childhood and upbringing. Yeah, I had a little bit of a different, let's say, just different upbringing than most people. So I was born, I like to say, a regular Italian kid in Milan, Italy.

My parents are both from there. And I lived the first six years of my life there in Milan or the area around Milan. And then my dad got transferred to Tokyo. He worked for a bank and they transferred him as an expat over to Tokyo. He worked for a bank and they transferred him as an expat over to Tokyo. And then I spent the next 12 years, almost 11 and a half years in Japan. So from basically first grade all the way through high school. And I went to American schools there, which is principally why I sound so American and not so Italian, disappointingly to my wife, of course. But the interest in engineering, I think, originally came from cars. I loved cars. I mean, I remember as a little kid, walls plastered with posters of

Ferraris and Lamborghinis and Lancias and all that. And, you know, Italy is sort of the Mecca for the supercar, sports car, hypercar thing. And, you know, it sort of started off with the idea of, you know, wanting to understand how those things worked and what the specs were and, you know, what a cylinder and an engine was and all of those things. So that's really the origin of it was there. I don't even know that I really appreciated it was engineering. It was just like a passion for these mechanical things that did amazing stuff and raced. So that, I mean, I went to, my dad, when I was young, took me to Formula One races,

which I thought was the coolest thing on earth. I went to Monza with him. I think he carried me around on his shoulder so I could see. Those are probably my first memories of that. It's really interesting, though, because a lot of kids, I think, particularly in Europe, watch Formula One with their family or with their friends and you know it's not unusual for formula one fans to look at the driver and say hey i want to be the driver like do you understand what it was that made you look at formula one and say hey i want to i want to actually be able to make this thing uh i i think it was um i, I didn't dislike the drivers, but the Formula One cars, I thought

looked cool, right? They looked amazing, but they, uh, I don't think that people meant for them to look amazing. They just meant for them to look fast and the fast would turn into amazing, right? Like it was, it was a form followed function in some sense. And I was sort of fascinated, like I would like design cars in the evening, like in my notebooks and stuff to make things look cool. And actually there is something about in that era, because today, when you think about how cars are designed, it's real engineering. It's wind tunnels and drag systems and simulations and all that. But back then, it was the idea of looking fast was in wanted to understand, like, you know, if you didn't have that engine, you weren't going to win. And so it was sort of like, well, you know, it's the technology that's actually making these people win. And that's what got me interested into it. Yeah, super interesting. You know, I'm also, you know, fascinated with, mom do? My mom's a journalist.

Yeah. So, and like, I don't mean this in any kind of way, but like, you know, they're non, not technical people, like not engineers, right? You know, so like, you know, like super, super highly successful people, but like your family or your immediate family wasn't a bunch of engineers. Were there any other influences? Like, was it just a pervasive culture thing? Like I know both,

you know, Italy and Japan are both amazing engineering countries. Like particularly when we were growing up, like Japan in particular was like, you know, this ascendant industrial power. Did any of that influence how you thought about the world? Neither of mine, my parents have done anything technical, nor are they particularly good with tech. Apparently, my grandfather, who I actually have never met, he passed before I grew up, but he was a chemical engineer and was quite the engineering mindset. I think what happened to me was I liked these things. And then as I was going to school, I liked doing math. I liked the science classes. And that sort of self-reinforces,

you know, and after a while, you're like, oh,'m just, I'm a person that likes this sort of thing. So it wasn't so much that I was influenced per se, although in fairness, both countries, Japan, Italy, countries that build a lot of mechanical systems generally. In Japan, I was probably more influenced by the electronics part of the culture because the Italians are not, you know, they don't have a great history in electronics, but the Japanese do. And I remember growing up a little later in life where, you know, the don't have a great history in electronics, but the Japanese do. And I remember growing up a little later in life where, you know, the Walkman was the thing to own. It was like the coolest thing. And once again, like I was the kid that would get the Walkman and like undo the screws and look inside and go like, oh, how do they do this thing? How do they make it? You know,

how is it that it's so small and it does these cool things? And so, you know, I was always just enjoyed like the, how is something made was something that always just interested me. And it got reinforced probably because the environments there's a, there's actually, you might've been there, Kevin, but in Tokyo, there's an area called Akihabara where they have an electronics district. Now it's a little, cause the world become much more software than, than hardware. But at the time, you know, my mom would take me there and you could just buy anything you wanted that, you know, everything from resistors and capacitors, pre-made pre-circuit boards, um, you know,

oscilloscopes, like anything you wanted to go, go get there. And I was like, you know, I just loved it. I, you know, every time my mom would take me there for the weekend, I thought it was the super, she thought it was incredibly boring, but I thought it was super cool. So it just sort of bubbled out of me, I would say, but the environment helped for sure. Yeah. And so how did you get from, I guess, high school in Japan to Stanford? Uh, so my high school was, I went elementary, middle, and high school in English. So it was kind of obvious to go to an English speaking place to go to university. All the way through high school, science and math was my thing. So I wanted to go to a university that had more of a engineering orientation. US seemed obvious. It was cool. Like the US was where cool things were made. Like, you know, you know, the companies that we admired at the time, like I had my first computer that I had

at school. It wasn't my computer. It was a school's computer. It was a Commodore PET. It was like 64K or whatever I was. And it recorded on a cassette tape when you wanted to, you know, storage was like literally a music cassette tape. And then I got a PC that was a NEC-made IBM clone. And all these things were American, like Commodore was America, IBM was America, like all this cool stuff was happening. So going to America was sort of the obvious choice. You know, I applied to a couple of schools on the East Coast and I applied to a couple of schools on the West Coast and the weather was a lot better on the West Coast. I went to visit schools in April and it was like 22 degrees in Cambridge. And, you know, it was like beautiful and sunny like today is in California.

And I was like, OK, I know where I'm going. So that's what got me out here. Yeah. Well, and in a world of no bad choices, like, yeah. I feel very fortunate. I mean, at the time, I will say, you know, in the modern era where it's so hard for people to get visas in the U.S. and stuff, because, you know, I wasn't a U.S. citizen, obviously, at the time. And it turned for me back then, it was actually pretty easy to go through the whole thing, get a visa, get citizenship and all that. I feel for the folks today that don't have a chance to do that as easily.

Yeah. Yeah. It's amazing how difficult we've made it now um but anyway that's a conversation for a later time um so you you get to stanford this is in the 80s um and like it's a really interesting time in Silicon Valley in general. So how did you, I mean, I sort of understand you're mostly interested in mechanical engineering and making mechanical things. And MechE's a good choice because you also get to do a little bit of electrical engineering and software and whatnot. But what was it like at the time like was there this pull at the time towards computer science and symbolic systems and you know all of the software stuff that was happening in silicon valley or a bit i i would say early So, for example, my mechanical engineering class had 65 graduating in my year.

I think symbolic systems had six and computer science had 25. I think now Stanford graduates 400 computer science majors a year. Yep. Turns out your good friend Reed was a year behind me. So I was class of 88 and he was class of 89. So I actually knew Reed there and he was doing this when I met him. He's like, I'm symbolic systems major. I'm like, what is that? I've never heard of that thing. I didn't realize you were a senpai though. Exactly. But I did mechanical mostly because I thought I'd end up designing cars.

And mechanical at the time, I don't know what it's like now, but you did two sequences. You did the thermodynamics sequence. So that basically engine design. Yep. And then you did the mechanical design control systems sequence. So I did both. And I thought and I interviewed with when I graduated, when I was graduating, I interviewed with a bunch of car companies, actually. And so I'm just sort of curious, like we'll we'll get back to the interesting stuff.

What was the hardest class in Mechie at the time? Because like I. Partial differential equations. Yeah. The math was definitely the hardest. I mean, I was okay at math, but when we get to like partial PDs, I was like, okay, this is beyond my skill set. So the math was the hardest, I would say. The others were, they were fun.

Like, you know, I mean, the nice thing, I think for computer science is even more fun today because when you write software, you get feedback immediately for what you've done. Mechanical engineering is not so immediate, but it's pretty immediate, right? Because we had the PRL or the product realization lab at Stanford. So, you know, you would go in, there's Lays and Mills and CNC machines, and you would, you cat out your system pop it in and boom out comes the thing you built and the thing you built i mean you only made one of them of course but um you know you'd get it in a day which is pretty cool feedback loop right and so there's i like that a lot i like that ability to to get quick feedback on what

you're doing and uh also it's practical because it's very, you know, when you make physical things, it's very tangible. You can see them and feel them. Yep. So there was a lot of reinforcement, positive reinforcement in that major. Yeah. I, you know, I kind of feel like an old man dispensing, dispensing advice but like i think other than aside from actually making things being fun i think it's necessary if the reason that you're studying engineering is because you want to make things exist in the world that never existed before you want to like create some effects like you actually have to do stuff at some point and the quicker you can start doing stuff where you are actually making the things that you're trying to understand theoretically like the quicker you learn and the better your

uh better your career is going to be and it always struck me as weird because like i when i was um when i was learning computer science right around the same time, uh, it was really a fashionable thing to teach computer science in a way where your fingers didn't touch the keyboard. Uh, so it was like a bunch of analysis. Yeah. And so it's, it's interesting, but like, it's just not the same as actually doing a thing. Yeah, no, I totally agree. I think that, you know, that feedback loop of making something that does, you know, responds then to some behavior that you exhibit is, for a lot of people, it's super powerful.

You can do a lot of theories of things and understand why they work, but that feedback loop to me is super powerful. And I think the other thing that I find that, you know, I've graduated with a mechanical engineer. I worked as a mechanical engineer for years, but I've not really been a mechanical engineer for a long time. However, the methodology of saying like, I want a thing to end up doing this for me. In order for it to do this for me, I have to break it into a composite set of parts. And then I have to take each part apart and build and, you know, successively build it so that it plugs in together and then becomes that. That mindset of parsing problems, which is really the engineering mindset, I think it's extraordinarily

helpful everywhere. I mean, I find it as helpful in business and thinking about how to achieve business goals and so forth. So foundationally, I mean, obviously it'd be great if we graduated more engineers in the world because I think we can always use more. But even if you end up doing something else in life, I still think it's an unbelievable foundation for somebody to start with. Yeah. And were you, so we've, we've already talked about you being sort of a curious kid. So you were like taking things apart to try to understand how they work. Then, you know, you're, you're just like drawn to complexity, not daunted by it. But were you also a skeptical kid? Let's see. I wouldn't say that I was extraordinarily skeptical. I mean, I think I didn't like it when people BSed about things,

when people made up stuff that I couldn't then sum up in my own mind right if you say like it's x and then you go like well how is x composed and that doesn't really kind of make sense I didn't like that very much yeah but I was aspirational as in like I always felt drawn to the idea of creating something that doesn't exist today. So in that sense, I did have some skepticism, but I was more motivated by what's around the corner, what's next, what's cool, and how can it be practical, right? How can it be useful day one? And I'm sure we'll get to some conversation about AI, but there's also a distinction there about like, oh, well, it's this amazing thing over there. But I always try to think about, you know, what is how is this practical for me today, as opposed to this pure vision? Like, I like aspirational, but I like it to the extent that I can practically apply it to what I do today. Yep. Yeah, which I think is a good impulse. So you graduate Stanford, and then you go get a job as a mechanical engineer.

What was that job? What were you doing? I was a product designer and product design engineer for HP in the optoelectronics division. So we made, HP at the time made various things that used gallium arsenide semiconductor material to create light. And, you know, you make a lot of products that create light. They can be anything from displays to automotive components to fiber optic connectors to all sorts of things.

But I was an engineer in the in the group there and what did what did that look like um because like i i think when people hear things like that uh like depending on what your background is you summon up a bunch of different images you know were you sitting at a desk all day, like, you know, doing schematics and drawings? Were you in a lab tinkering around with stuff? Were you with customers, like helping them build things? No, I was pretty much so a mechanical engineer's principal tool or CAD CAM. So HB had its in-house CAD CAM system. I grew up on AutoCAD and CATIA, which were the kind of industry standards, but HB didn't use those. So, I was on ME10, ME30, which at the time ran on workstations, big computers.

With your desk, you design this thing. It was a three-dimensional renderer that you did stuff and designed it in various shapes and forms. And then typically, you would design one out. You would send it out to the prototyping shop, they would make it, then you would test it. So you would take it into a lab. And, you know, in my case, it had to have certain thermal characteristics. So it was an interesting job because it involved both thermal, well, when semiconductors are pounded hard with electricity, they create heat. When you create heat, the performance degrades and also it creates,

if you involve it with plastic, it melts plastic around it. So I had to do a lot of thermal stuff to make sure that we weren't overheating these systems. And then mechanical design to make them fit into certain mechanical spaces

was what I was doing. I would design it. The prototype shop would build it up up we would test it and then when it was tested enough to our satisfaction we would send it over to what was called new product introduction engineering NPIE those were the guys that basically took our designs and turned them into a production line that they would make them at the time our products were made in Malaysia so we had a bunch of NPIEs over inia that would take our product design and then put them in and try to build them and you know usually you made something and it didn't work uh so so then you had to go you know spin it

another time and do it right but it was probably uh yeah i mean it's a long time ago but i say 30 of the time cad cam 30 of the time working with other people to, you know, facilitate a bunch of time testing and test eval was a lot of time. And we had to run a lot of tests that ran long periods of time and then we had to do temperature chambers to cycle them and those sorts of things. So that was basically the job. Interesting. And so at some point you made the shift from being a mechanical engineer at a Silicon Valley company to like was programming in basic in high school. I programmed a little bit in Fortran also in high school. I took a bunch of Pascal classes in college. So it was, you know, it was sort of there. It was funny. I was talking to my dad's always been a good advisor mentor to me and he

mechanical engineering, he can understand. But when I was like, well, cause when I was thinking understand but when i was like well because when i was thinking of majors i was like me maybe cs i was asked my dad is like computer science that's like games and stuff you don't want to do that that's a really bad career path stick with mechanical engineering way better so i think thanks for the advice dad really appreciate it we worked out for you you know if you hadn't given me the advice, I could be CTO at Microsoft now. Yeah, you made the right choices, my friend. But yeah, so I always played around with software. But what actually happened is my boss at HB had an MBA.

He was the manager of the engineering group that I was that was working in. And, you know, he he liked me a lot, was very kind to me. And he you know, he was like, you know, for an engineer, you ask a lot of why questions like, why are we doing this? This product doesn't seem to make sense. Why? You know, what's the you know, why are we making this one instead of that one? And why is that division over there? So you said you should really go and explore the business world to kind of understand and you should get an MBA. So my third year, three years of engineering, I applied to business schools. And I decided to leave and go to my manager actually wrote my letter of recommendation. I decided to go to business school to sort of understand the business likes of it. Not, not knowing what I would do afterwards.

I assume maybe I'd go back to doing what I was doing before, but I'd learn it. And I think that's where sometimes in your career, you get lucky because doors open. But in my second year of business school, I was organizing a conference, a technology conference for the business school at Stanford. We were inviting these speakers to do things. And originally back then, this is circa early 1994, when you advertise the conference, you would put posters up around campus, say, you know, come by, we have these speakers, whatever. And one of my friends was like, well, we shouldn't do posters, we should do a website. And I was like, what's a website? And they're like, oh, there's this thing called

the web. And I was like, oh, like the thing that you do email on? Because we had email, I even had email at work back then, but the web wasn't a thing. And in early 94, what was then Mosaic and then eventually became Netscape had launched the web server, web browser. And so instead of making posters, we made a website and we would have we had a very simple little database where you could enter your email and then we would then follow up with an invitation on email. And, you know, it got me to exploring the Internet and that, you know, the light bulb went off. You know, seeing the web, I was like, wow, like this is, nobody's ever going to make posters anymore. It was one of those luminary moments. But no, I sort of got the bug. And then, then I actually said like, whatever I'm doing, I have to find a job that has to do with the web. Like, it doesn't matter what. And at the time, 94,

eBay was started in 95. Yahoo was started in 95. There literally weren't, there were no internet companies in 94. And so then typical engineering mindset, I was like, well, if there's no like web companies, like how's the internet made? I started doing research on how the internet works. And there's this thing called TCP IP and there's routers and switches. I was like, well, who makes those? There's a company called Cisco makes them. So I like, I cold called the CEO. I just like, and at the time you actually called people, literally cold call people. And his assistant was nice enough to put me through.

And this was Craig Chambers at the time? No, this was still the first CEO was a guy called John Morgridge. Okay. John was nice enough to like literally take my call. And he invited me down and I did a few interviews and got passed around. And finally, the guy, the CTO actually of the company hired me to be sort of technical assistant, basically. It's like you come and do whatever I tell you to do. And then one thing led to another.

But it was literally like that conference and making a website. Like I just, you know, we got on. My friend and I got on, made some HTML stuff. And I was just like hooked. Yeah, that's awesome. And so what did that job look like at Cisco? So like you were the TA to the CTO. What were some of the technical problems that you all were trying to sort out? What, 95? context, Cisco was a small company then. It was about 2,000 employees. It was very informal. We had grown up very, very fast as a company. It was a couple hundred million in revenue, I want to say.

And so roles were a lot less formal than roles are today in the organization. They basically hired me because there were two technological themes that could affect their core business. They had routing and they had two technological themes. One was broadband to the home, right? Because back then it was all dial-up. And so they wanted to evaluate broadband technologies. And then the other one was that there's a technology that sits right next to routing called switching, which is sort of a more hardware-oriented approach to moving packets on the internet, which they were afraid would disrupt routers. Turns out to be true.

It did. But they were worried that that technology could obsolete routers, and we had to evaluate from a business development perspective whether we should be making acquisitions in that sector. So those are the two projects I got a chance to work on with the CTO. The broadband work led to essentially us partnering and investing with a few of the cable companies to make cable modems. So they were invented at Cisco. And then on the switching side, we ended up acquiring a bunch of companies, like four of them, including Andy Bechtolsheim's company at the time, to build the entire switching product line at Cisco via acquisition. And that turned out to eventually become the largest part of Cisco.

Yeah. That's kind of incredible that you were there helping to sort all of this stuff out strategically, right. As the internet was beginning to take off. Yeah. Well, I think what was, um, what I, I, what I learned to really appreciate at the time was my boss at the time, the CTO, was very good at sort of seeing, okay, this is what we're doing today. It's great. Business is humming. We love it. But there are things that might happen two years from now that will dramatically disrupt what we're doing. And all this happy stuff that's happening now will finish in three years if we don't pay attention to that next thing. Right. And that, you know,

the natural tendency of a business that's going well is to just keep doing what it's doing. And he was very good at saying like, ah, like that's the exact moment that you have to go and lean into the thing that might kill you. I think I was mostly there helping because he was a big strategic thinker, but it was not very hands-on. And so I was like hands for him. So he would say like, okay, go explore that one. Go do this. Talk to this guy over here. And in the process of doing

the work for him, osmosis set in. And I was like, oh, that's how I got to look at the world of business. And that's how it works. And it's like being an engineer and you're like sort of sitting there saying, oh, if only we did it this way, this can create another opportunity over here. That was the biggest lesson for me at the time. And so you were at Cisco actually for quite a long while. So 10 years as far as Silicon Valley, 10 years ago, that's a long tenure. I mean, by Microsoft standards,

like people stay here for like, we just- No, it was even longer. So it was 14 years I was there. Oh, was it? Yeah. So, well, I guess two questions then. So like one, like what made you stay so long when I'm sure every month there was some new opportunity? Like you had this great career. You were probably getting called all the time to come do other things.

Like you could have stopped being at Cisco at any point. Like what was the thing that kept you there for as long as you did? do other things? Like you, you, you could have stopped being at Cisco at any point. Like, well, what was the thing that kept you there for as long as you did? And then I'd love to understand, like, you know, when you finally did decide to move on, like, why did you choose, uh, what, what came next? Net net two reasons. I think the first was that my, my role kept changing and I was learning something new. So I was doing more or less what today would characterize as corp dev and strategy work for the first six years there until like 2000. And then I was kind of ready to leave at that point. And then John Chambers, who is the CEO, said, well, if you want to be a real businessman, you need to own a business.

Like you're just sort of like a deal guy. So you should really learn that here. And he gave me the opportunity to take that switching business, which I mentioned a minute ago, that we had assembled through acquisition and become the general manager for that. That meant I was now, you know, I went from managing like 50 people to managing like 3000 people. And we were doing everything from chips. So we had teams that were designing the silicon, custom silicon, electrical engineering, mechanical stuff, thermal stuff, software, core OS software, application layer software, management software, all that stuff in the product, plus product management, product marketing. So it was essentially a GM of that business. And I just thought to myself, this is an amazing experience. Like who gives a kid who's never like, I literally went from being an IC engineer to managing 3000 people,

you know, less than 10 years later. And I thought this was really good learning. So that, you know, I had that business unit, then they gave me more, and then they gave me more. So every time I thought I was different competitors, different challenges in the market, different customer sets, I went from like selling stuff to Wall Street trading floors to selling the big routers to internet service providers, people like Verizon and Comcast, more customer interaction, which I found super fun. So I felt like I was constantly learning. That was one. And two, and Kevin, I'm guessing you feel some of this as well being at Microsoft. You know, Cisco is not as much of a center stage company today as it was back then, but then it was very center stage. Like every opportunity, every important event that happened in and around the Internet was coming through one way or another. You were touching it.

And so you get spoiled because you're at the big boys table. You're sort of like, yeah, I'm really moving the big chess pieces around this thing. And I had a very good relationship with our CEO and he really brought me into it. I mean, it was at all the board meetings and everything. And so that was fun. Feeling like it's center stage and I'm learning a lot was amazing., uh, was amazing. Yeah. The thing that got to it is at some point networking lost its glance. It was sort of like, you know, you don't, well, now you think about networking because the data centers, you go nuts. It's back again, man. Back again. But for a while there, it got a little sleepy, right? It was just like,

you know, you plug in the internet works, who cares? And, and also, I got to age 40. So, you know, I ended up spending 14 years there, I started when I was 26. And it was just one of those life moments were like, yeah, am I going to be a lifer here? Or is there something else to do in life? So that so those two, like the sleepiness of the business and a realization that I just had to do something different in life. So why, why become an investor? The investor thing happened later. So I thought I was going to be an operator for a long time. So I ended up joining through some friends that I had at Sequoia. I ended up joining a company in their portfolio called Juiced, which was doing, was basically today you would call that Hulu or Netflix. It was streaming video. I was CEO for two years.

I learned what I didn't know, which is I understand the tech. I had no clue on the content side. All this stuff about content licensing and rights and copyright and all this stuff, no clue. So it was good tech, terrible content. company went pretty much nowhere it ended up getting acquired and uh i had gotten because of my corp dev background i had gotten a lot of calls from a lot of vcs over the years and then my good friend danny reimer at index who was on my board said why don't we just give this vc thing a go and uh you, I had thought about the idea, but I never thought I would be a VC. It was, I liked products and I thought I was, I'm a product person. And, but then, you know, maybe it was also a little bit because the startup

journey was so shitty, like it really didn't work. And I felt maybe a little defeated. I was like, well, maybe, maybe I'm not a product person. I'm just like a, sort of a tech trends kind of a guy. And, uh, you know, I, I enjoyed my court dev time and at Cisco and maybe VC is like that. So, you know, I decided to join forces with Danny and become a venture capitalist. But I do think you are a product guy and you're technical. I remember one of the first times we met, you're citing research papers to me and you're constantly you're constantly reading like a bunch of highly technical literature and uh so do you think that that has helped you be a better investor that like you you've got this curiosity and like you've got all of the foundation there where you can just sort of dive deep into

really gnarly technical stuff uh yeah i a hundred I mean, you know, what I've learned over the years is that I'm sort of this weird product as a human being that is both very keen on technology and understanding why things work and so forth. But I also, I like to jump around a little bit and see different things with different context. You know, I call that like professional ADHD. But I have both this love of tech and professional ADHD. And the venture thing, I think, worked out because at least the way I prosecuted it,

it was always a game of like seeing a lot of different things, seeing how they fit together, understanding how they work. And I do think one of the most important things, there's a lot, you know, you can have very different styles of investors. Some investors are just very good at seeing heat and jumping on it. I fancy myself as an investor that is good at understanding when a technology can actually become a business. And I think in that context, being technical enough, because I can't, you know, I can't write code. I mean, I can write a little bit of code, but nothing that would ever become a product. But I can read a paper and know and go like, ah, you know, this is cool,

but this doesn't feel like it can be part of a business, or maybe this is a little too early, or it's not far enough along. So that, you know, that ability to assess a technology's maturity and its ability to become a product is how I've hung my hat as a venture capitalist. And that doesn't mean that's the only style, but for me, it's worked. Well, so I think that's a good segue into AI because I think you've made some of the most interesting investments in AI. And I would argue that AI is an especially difficult thing to invest in because it's changing very rapidly. It's kind of complicated. Like there's some strategic stuff

that, you know, I think a handful of years ago wasn't obvious at all that is, you know, that you have to be able to see early or like, otherwise you're going to be too late uh like you you invested in scale which is like one of the truly interesting data companies you invested in cohere which is like one of the most amazing teams in the era of generative ai that's like building a really interesting business so and and like i guess the other thing too that we we'll say like, again, you know, now a lot of this stuff seems obvious, but like handful of years ago, like all, almost all of the things that people are excited about right now, most folks, you know, investors and tech executives and lots and lots of super smart people were like, yeah, this is nonsense. You know, like none of this is going to work. Like, you know, you guys are just full of crap. So like, how did you decide that this was interesting, and you wanted to go deep on it?

Yeah, well, I mean, you know, in some ways, you look back at my life, and the dots always connect in a similar way. And so what actually happened to me was that I got really interested in Waymo or back then Google's self-driving program. I thought that was like super technically cool. And even though I was a little skeptical about the timeframes they were talking about, I was very interested in how it worked. So I started doing some research on some of the core elements of the technology there. And the thing that popped up was deep learning for computer vision, right? So that was, I was like, okay, what is this deep learning thing? So, you know, you, you, you wind the tape back and you

go back to like ImageNet and AlexNet and you go back to like, basically what machine learning systems look like and how they were applying them in multiple methodologies because at the time self-driving was not one big model but it was like a series of separated models some for vision some for motion planning and so forth and that got me pretty interested and so basically i started chasing ideas around self-drivinged up meeting a common friend of ours, Chris Ermson, invested in his company. And one of the beautiful things about investing is that you meet these incredibly smart people and they'll actually teach you how this thing works. So I was sitting there with Drew Bagnell, who's the CTO, and he's just like literally giving me a PhD level class

on how this stuff works. And they were like, yeah, we need to use labeled data. And I was like, all right, we'll label. And so they explained to me why, you know, vision objects need a label. And I was like, well, how do you get labels? And I was like, well, people have to put labels on it. I was like, really, how do you, how do you do that? And I was like, well, we use this company called Scale. So I was like, wow, that's interesting. Let me let me introduce so i went and met alex and he was like look this labeling stuff you know it seems mundane but you know in order to train machines you you need that and this is way before llm like llms were this is 2016 so they hadn't even written

the attention paper yet google hadn't published their birth stuff stuff and so forth. So, you know, a lot of it was when Alex Wang, who was the CEO at Scale, explained to me why this was important. And it just resonated with me that we would need a lot of labels in the world. Like over and over again, we would need them. And people said, like, well, we'll do simulation. And I was like, yeah, I don't see how simulation is going to work exactly well here. Like, I think so. I bought the story. We invested in that in that company. And then, you know, one thing leads to another because because you do computer vision, you see robotics. So we invested in covariant. The covariant guys are are friendly with Jeff Hinton and the folks at University of Toronto. That's where Cohere is. So we get connected to them.

And, you know, one thing leads to another. I think at the time, to be honest, Kevin, a bunch of my partners thought I had lost my marbles. Because, like, nobody invested in AI in 2017. And much less, like, you know, OpenAI was a nonprofit at the time. And people were like, well, how do you invest in a nonprofit? You don't make money in a nonprofit. Big mistake. But it just struck me at the time that humans are, in some sense, the pinnacle of biological technology.

We are, as far as we know, at the top of the pyramid of biological technology. And I felt like a system would not copied it, but attempted to emulate how humans worked and did so via a computer system had to have a really, really bright future. Like it just was, it was sort of like saying the horse is the best instrument we have to move around. If I only could build a mechanical system that mimicked a horse, I could actually improve productivity a lot. And somebody invented a car, right? And in the same way, like a human is the best brain system that we know of.

And if you try to build, you know, you're not, like a car is not a horse. They're very different. But they serve a similar purpose. And that's at least how I think of AI, which is actually AI systems don't really work like the human brain. They sort of loosely, but they serve the same benefit at the end. They just because of the nature of scaling and computing, they can be much bigger. Yeah. And so that that's sort of like that light bulb went off. And I was like, I don't care what my partners think. This is it. Like, we're doing this. And so I did a series of investments and got more and more involved. Yeah, and it's it is.

I mean, maybe just talk a little bit more about how. I mean, maybe this is a thing that's run through your whole career. It's difficult sometimes to see a thing and have conviction in it and say, I gotta go do this. And everybody else around you is like, yeah, this is dumb

and we don't understand it. And you somehow or another summon something, whether it's contrariness or, you know, like some kind of primal determination or, you know, like cold rational calculator, whatever it is, but like, it's sort of tough to row against the current, so to speak. And like, you know, it sounds to me like just sort of listening, you know, to you talk over the past hour, like that's a lot of what you've done over the course of your career.

Yeah, I think, Kevin, maybe, you know, maybe you feel some of this stuff at the same way, which is when I'm wrong about stuff, I tend to be wrong because I'm too early on it. Like I see things and I can understand how you can get there, but maybe it takes five years instead of two and I'm wrong about it. And so what I try to do in my own mind when I look at things is say, well, that's my natural tendency. I tend to underestimate the amount of time it takes to get there. Applying that filter of bias when I look at something, do I think that it has extraordinary potential? And within a reasonable timeframe, can I nail that? If I'm convinced of that,, if I convinced that it can work in my own sort of architecture in my head, don't really care much about how other people feel about it. Um, because I think a lot of other people are either too aspirational at times, you know,

they're selling you like past the hash pipe type things, or they're in the mindset of like, yeah, that'll never work. Yeah. That'll never work. Yeah. Right. It's, it's most people I find are in those, in those corners. And, um, I think that actually a, a really good way to be a business person is to find that happy medium in our industry. And that important because the majority of people are not going to be where you are. So you really have to build your own belief system based on essentially logic of why it is that I believe this. And if I really believe it, first of all, I'm okay being wrong. And by the way, I'm wrong all the time. So having compassion for yourself for being mistaken at times, but also continuing to have the self-confidence to say, I believe in my own logic. I believe that this is going to

happen. And that allows you to take certain chances. Honestly, I think in AI, I was maybe a little bit early, but more or less it's worked out. Yeah. And sometimes, you know, you even have to ask yourself as as an investor, you know, like whether you're a venture capitalist or you're a chief technology officer, like what the cost of being early versus late is and like where you want to be biased. So yeah, with AI, I think it's probably worse to be late than it was to be early. I think you're completely right. I mean, I think today it's very, very tough to be an AI investor. It is very hard. For one, the industry has matured where it's become a large player's game because the capital required is massive. Two, the minute you have an idea, since everybody wants the capital, throw capital behind it, 37 companies appear that do the same thing. So I think it's really hard now. I mean, in some ways, it's like you show up today and you go like, oh, I'm going to do AI.

It's like, OK, where's the insight there? Like, you know, being early, you know, you have to convince a lot of people. And you had to do this at Microsoft, I'm sure, when you were thinking through the AI stuff. Like, you had to convince a bunch of people that this was worth it. Because they were saying, ah, it's been done forever. It never works. You know, it's been 30 years. This stuff never works, you know. And you have to be like, no, no, no. But this is actually the moment.

Trust me, this is the moment. And maybe you were probably a little bit on the early side, too. But at least you weren't overpopulated with like a thousand people doing the same thing. Correct. Yeah, we were. We were we were we're early, I think, but like right on the verge uh like if you just sort of look at how things have unfolded like if if and there's some things by the way where i'm like i just really desperately wish that i've been six months earlier um like really like i mean it's

small handful of things we can talk about about those like off, off the air, but like there's a handful of things where I'm like just six months made such a tremendous difference where, you know, it's, it's just a big pain in my ass that like I didn't see it just a little bit sooner. Yeah. Yeah. I would say I probably have a few things that I went a little too soon. But, you know, at the end of the day, you know, sometimes you're going to catch them just right.

But the truth of the matter is you can never end this business till. And in particular, you know, the world we live in an AI, the curve is so steep that those it's not sick. It's six months in human time, but in technology development time, it's like four years. Right. So it is a very tricky thing to pace in time because the rapidity of the velocity of change, you know, the the delta is so high right now in this sector. So maybe a last couple of questions. What do you think is interesting going forward, either in AI or, you know, like anything else that's sort of happening in technology that you think is interesting that the AI side that I pay attention to. One is the physical embodiment of AI, which is interesting. I think the AI we experience through Bing Chat or ChatGPT or Cohere or whatever is the purely digital experience right now. And, uh, I think that, uh, you know, we are analog beings and at some level there needs to be a, a, uh,

and a physical experience associated with that of some variety. So whether it's, you know, robotics or it's devices or, you know, other things that allow a more physical embodiment of what we perceive to be AI, I think is super interesting. I do pay attention to technologies other than transformers. I think this is an investor bias. I don't see how investors can, you know, startup investors can win in transformers now anymore because of the capital requirements of it. And I would say for now, there is no obvious scaling boundaries to transformers,

but maybe there might be. And if so, might there be a different approach yeah and maybe maybe not but it's my job to explore yeah and for listeners uh who may not be ai people uh like when we when mike's saying transformers he's not talking about optimus prime uh like this is the uh this is the prevailing architecture for deep neural networks that is basically driving all of this crazy scale-based progress right now. Yeah, exactly. I mean, everything that everybody experiences today in AI is largely based on this technology called transformers. And it has very good scaling characteristics, meaning that if you know, if you throw more computing power at it,

it just gets better and better and better. That's not generally true with technology. Technology usually has a plateauing effect. Like you can throw more resources at it, but the pace at which it improves flattens over time. Yep. And, uh,

this particular technology has not shown characteristics of flattening so far. Uh, but, uh, it also means that the resources required get massive. So it means a lot of power, a lot of computers, a lot of data centers, all that stuff. Very hard for a small company to survive. This is very hard for a large company.

Yeah, I'm not kidding. Good for coal in West Virginia. So, yeah. So those sorts of things I'm enjoying and trying to pay attention to, to see, you know, is there something around the corner? Outside of the AI space, I mean, I think it's still sort of AI related, but the intersection between AI and medicine is an area that I've been paying a lot of attention to. I think that we're even earlier in the stages of that. hopefully a chat GPT like moment at some point where we define, you know, we can design a vaccine in three weeks, uh, or, uh, or find a drug that, you know, uh, uh, cures or, or addresses, uh, diabetes or something. And, uh,

that's the other field that I am paying attention to. Super cool. Well, um, last question. Um, and I asked this of everyone on the show. So you have this super interesting day job and you, you know, in addition to being an investor, you get to do cool things like be on the board of Ferrari. What do you do outside of work, whatever it is that means for fun? I do a little bit of sports just to keep it real. Play tennis, ski, that sort of thing. I read a lot. I actually hate business books. I think I read a lot of them and I'm totally done with them. But I particularly enjoy books that recount either people's lives who had some fascinating involvement with the evolution of technology or stories about how technology came to fruition, like how a technology became something that's part of our lives. I find those things absolutely fascinating. Do you have a favorite recent one? Yeah, so there's one called,

that I recently called The Empires of Light, which is the story of the war between Tesla doing AC and Edison doing DC and how basically Tesla lost. I mean, he died crazy and poor and Edison won. But actually, the technology Tesla developed turned out to be the right technology. So we all use AC at home. And there's a whole interesting sort of business subtext to the creation of the technology. And there's a lot of, you know, what made Edison a great entrepreneur?

Maybe not such a great technologist, but a very good entrepreneur. What made Tesla a great technology, but actually a pretty pathetic entrepreneur? His relationship with George Westinghouse and so forth. So fascinating book that really teaches you about how, you know, because oftentimes we sort of simplify and we say like, oh, well, the best technology wins or the best technology never wins. And neither one is true. There's a lot of sort of permutation. So that's a fascinating book. And then I go to the occasional Formula One race to enjoy the same thing I enjoyed when I was six. That's awesome. Super, super cool. Well, this has been an amazing conversation.

Thank you so much for taking time out to chat with us today. Kevin, it's always a pleasure. We have to do this more often and maybe we can do the version of this that doesn't shielded by Microsoft secrecy in person. Absolutely. All right, man. All right, Kevin. Thank you.

Bye. Okay. Well, that was a fantastic conversation with Mike Volpe. And as you were kind of discussing before your interview with him, his background is so fascinating, because as you mentioned, he had this experience at Cisco, where he was for 14 years. And as you mentioned, you know, it was kind of the center of kind of, you know, the, you know, dot com kind of boom and kind of the first real internet web boom and continue to be very successful. And what really

struck me, though, was, you know, you have this person who's obviously brilliant, goes to Stanford studies, mechanical engineering, and, you know, is working in engineering at Cisco, and then says, I want to learn about the business, and goes and gets an MBA. I've met a lot of entrepreneurs. I've met a lot of technologists. I know a number of MBAs. It's fairly uncommon. Has that been your experience? I don't know a ton of people who haven't had previous, you know, interest in the business side who are,

you know, engineering majors who've decided, no, I want to go get my MBA. Yeah. Like there's some, I mean, like quite famously, my boss is one of them. Right. But, you know, it is an interesting thing, like particularly at that point in time when you are sitting, you're sort of sitting in Silicon Valley at one of the boom moments. And, you know, maybe that's even a weird thing to say because Silicon Valley has more or less been booming for five decades plus now. But, you know, the 80s and 90s, when he was like an undergrad and grad student, were like an especially active time in Silicon Valley. And, you know, I know a lot of people who just don't even want to get an undergraduate degree because they're afraid that they're going to miss out on something. And so, you know, I think it's one of the things that I really do

respect about Mike is that his curiosity is broad versus narrow. And so like, he's just constantly trying to fill in the gaps in his experience so that he can give himself even more permission to be more curious. And I think that's what going to B school was, was about, you know, it's him, it's him, you know, like being an engineer and realizing, okay, well, like engineering without context is not super interesting. So like, let me, let me go figure out how to get context and then you know as we heard um you know the more curious you are like i think the more opportunities you create for yourself uh because you know him being at b school is the thing that created the connection for him to go to cisco right right no which which again i mean i think is so fascinating right that

he um had that curiosity and that kind of instinct to go to the next place, which is clearly set him up for all the stages in his career, both at Cisco doing the, you know you know, investment and stuff. And then, you know, as a CEO and now as an investor, I really enjoyed the conversation that the two of you had about being too early versus too late, because I think that's one of those conversations that people could have kind of endlessly for hours, right? But I did kind of want to ask you, what was it about AI in this moment? Let me be more specific. This moment in AI that we're having, this moment with Transformers and with generative AI and whatnot that made you

make the case to make some of the investments that Microsoft has made and be one of the, you know, earlier, you know,, that these systems were really doing interesting things as they scaled up and that they behaved more like a platform with greater scale. And so one of the things that I had been skeptical about with AI and I had been episodically skeptical about AI for most of my career. So when I was an undergraduate and in grad school, we were in an AI winter. So like we had in the early mid 80s, like people got super enthusiastic about AI and, you know, like lots of energy and lots of people studying it and lots of companies forming. And we like we had a really catastrophic bust. And this has happened multiple times, like where you've had moments of enthusiasm and then the air goes out of the bubble.

It's so frequent. In fact, you can go search for AI winter and there's a Wikipedia article explaining the bust part of the cycle. And so the thing that I'd had skepticism about for a while is like if the problem you're trying to solve with AI is the intellectual exercise of understanding in very precise terms the mechanism of intelligence, which is what a lot of AI had been about. It's like, okay, like we've got this brain, it must be like this subtle symbol processing system. And just like Newton's laws of motion

and Einstein's theory of relativity, there's some mathematical beauty sitting in the neurons and axons in your brain. And if we just sort of think about it hard enough, we self-reflect. We're going to find these pure, simple rules that explain how cognition works. I was super skeptical that that was going to be a thing that goes fast. Because the nature of scientific breakthroughs like that is it's just sort of slow. You go slow and then you have a breakthrough and things change. But predicting progress is hard.

And what happened in 2017, particularly with Transformers, is we had a mechanism whereby things could go fast. And you didn't actually have to understand the fundamental nature of human cognition. You had this thing that, as you applied scale in data and compute, could emulate parts of human cognition and the parts that it could emulate were useful enough where you could build a platform on top of it and so like that that's you know it was a pretty straightforward argument like it just seemed blindingly obvious uh and you know if you are a company like microsoft that is a technology platform company and your job is to equip people to go solve problems with technology, if this is going to become a platform, you must make sure that it is a great, excellent part of the overall platform that you're building. You can't really have a technology platform, I think, in 2024 without having highly capable AI. No, definitely not. But certainly, that wasn't clear to many people in 2017, except for people like yourself and people like Mike,

who were already investing in companies even before some of the seminal papers were being written. So as always, I'm appreciative of both of your insights because I think we've obviously needed this, especially as Mike was pointing out, this has become only more capital intensive and that early era of probably a generational thing where you know there are very few uh when i look back in history like moments to kind of get in at that same stage and now it's it's just about identifying you know the the next big waves um probably associated with these things yeah and there's still i mean the thing i'll say to everyone who's listening there there's still like an unspeakable amount of stuff to be done it's just uh yeah like what what folks ought to be thinking about right now is like how do i go

harness the power of this platform to go build amazing new things to solve the education problems and the health care problems and the just myriad business problems that uh you're going to be able to solve now that you've got a new tool in your toolbox to go after them. So like that, that's to me, yeah, the fun stuff. Like I always enjoy, like I derive a whole bunch of satisfaction in building products for other people building products. And but, but like at the end of the day, like building tools is kind of useless unless people use the tools to do something that matters. It's so true. It's so true.

Like, yeah, for platforms to work, people have to actually use them when to build things out of them. And I think we're lucky we're in a space right now where people are definitely doing amazing things and are going to continue to do amazing things. And people like Mike are going to continue to invest in those things, which is great.

For sure. All right. Well, that is all the time that we have for today. Huge thanks to Mike Wolfe for joining us. If you have anything that you would like to share, you can email us at behindthetech at microsoft.com. And you can follow Behind the Tech on your favorite podcast platform,

or you can check out full video episodes on YouTube. Thanks for listening. See you next time. you you you