Why One Internet Pioneer Thinks the Original Model Broke

Key Takeaway

Run a 20-minute privacy audit today: list the apps and sites you use most, review which permissions and trackers they rely on, and remove anything that is not essential. Treat your data trail as cumulative—not as isolated purchases or clicks. Then move one sensitive conversation or large file transfer to a service designed for encrypted direct communication rather than assuming every “secure” messenger or VPN eliminates intermediary risk.

Episode Overview

Joel de la Garza speaks with internet-infrastructure pioneer Barrett Lyon about DocsNet, a network designed to restore direct, encrypted peer-to-peer communication without centralized servers storing users’ activity. Lyon argues that conventional VPNs and messaging apps often leave users dependent on opaque infrastructure, while pervasive tracking becomes more consequential when combined with cheap, scalable AI.

Full Transcript

Transcript of Why One Internet Pioneer Thinks the Original Model Broke from A16Z. Auto-generated from episode audio; may contain minor errors.

We're really at an inflection point because you have the birth of essentially coin-operated intelligence at scale. The internet has really kind of stagnated a bit from protocol design. There's like not a lot of new stuff, all of it's built on top. Erasing that is kind of what I'm trying to go for. What you're saying is you've actually built this transient way for folks to connect and chat without there being any record. The really big problem is on WhatsApp, Signal, iMessage, Telegram, all of those applications disguise themselves as you sending something to somebody else, but you're really sending something to a server.

The question is like, okay, well, whose servers are those? It's coming at such an important inflection point for like the future of our country, tech, and to just have this conversation that seems to be manipulated to some end. It feels very manipulated, and to me, the future of our tech is needing to build this infrastructure now if we're crippling it. The internet was designed to connect people directly. Today, much of what we do online passes through someone else's server, along with an enormous amount of data about us.

In this episode, A16Z's Joel de la Garza sits down with internet infrastructure pioneer Barrett Lyon to talk about DocsNet, his attempt to rethink that architecture from the ground up. Barrett explains why VPNs aren't enough, how DocsNet enables encrypted peer-to-peer communication without a central server sitting in the middle, and why the company owns its infrastructure all the way down to the hardware. They also discuss what happens when decades of online tracking meet cheap AI, why Barrett thinks internet protocol development has stagnated, and what it would take to bring back some of the openness and experimentation of the early internet.

Well, today we're joined by Barrett Lyon, perhaps one of the most interesting people in tech. I have a lot of stories to share, and I am always kind of blown away every time we catch up because you tell me a new story that just puts everything I've ever done in ale and comparison. And I think the most interesting one I heard recently, and I had never heard it before, and this was during a period, I believe, where we were investors in your previous company, was you are a pilot, you have a pilot's license, and you decided to optimize your life by flying places, and you had an unfortunate experience, I believe, flying back from one of our meetings.

Yeah. Well, I mean, one day I was driving to work in the Bay Area, and I realized it was almost 45 minutes, which used to be 15. And my family lives up in the Tahoe region, so I'm like, well, I think the plane can fly there in about the same time, maybe I could just commute that way. And that was before work from home was codified by COVID. So I had a need to physically be present. So yeah, I started flying. So your daily driver was a daily flyer.

Right. Yeah. So I put 1,300 hours on, I'm serious. And yeah, one day after meeting here, I left and got in the plane, took off from San Carlos, and somewhere over Oakland Airport, engines started acting up. And I thought maybe it was like some sort of water in the fuel or something, but I climbed up, cleared Bravo, and felt like I was okay. And then right after that thought, lost pretty much all power. And what went through your head immediately after that? Yeah. Actually, I messaged my friends and said, hey guys, I have an emergency.

Love you. I don't know, I just declared emergency and just tried to think, okay, where do I crash and not hurt people? I found a field, and then it turns out I was pretty high, and there was a giant airport right next to us. Not Oakland, but it was Buchanan Field, I forget the name of the town. And it's a giant old military airport. Oh, in Alameda. No, it was the other one. The other one on the other side of the mountain range there. Buchanan Field is the name of it, I believe.

But anyway, I landed there, but before that, I kind of went to this field, and I was like, all right, well, I've got a parachute, ballistics parachute. Do I pull it? And I was thinking, all right, well, when I read all these crash reports, the reason the pilot dies is because they didn't use the parachute, so I'm like, well, the airport's right there, and I'm still in the air flying. And isn't that a funny quirk of that plane, that it's technically the safest plane ever made, but yet it's involved in the most crashes because people never use a safety feature?

Yeah. Well, and I mean, it does work, but it destroys the plane, too. So it shoots a rocket out of the back of the tail, and it rips all this stuff out and ruins the plane. So, and then it's not a guarantee that you're going to land nicely on the ground. And in fact, the seats, I think, are made with a crumple and packed thing, so when you do hit the ground, it absorbs the compression. And this is all playing out over how long a period of time?

Oh, I don't know. Probably like minutes or less. Maybe less. And then, anyway, I just made the call to land and landed, rolled right onto the transient parking spot. I was calm. I was kind of frustrated and a little confused when I got out of the plane. And then, you know, those big fire truck things, the huge ones that look like some kid made them, but they're big? Yeah. The guy gets out of that, shakes my hand, and then that's when the fear hit me, because I'm like, that's the guy scraping my brains off.

I'm like, oh, no. This is the guy that puts the fire out. That was close. Then I'm like, now what do I do? I have to get home somehow. Rent a car, drive home. That was kind of the end of it for me. I was like, the risk is too high. And I kept the plane for a while and ended up ferrying it out to Las Vegas and selling it. Did you ever find out what the issue or the cause was? Yeah, it was like three really weird things at the same time.

One was a piece of metal on the air intake control valve, or kind of a lever broke off and sucked into the air intake manifold. And then it went down to the oil pan and it was crushed into cylinder three and six, I think. Oh, wow. And then that caused an electrical problem with something with the spark, which was controlled by a brand new electronic ignition system. That stopped working. And so then I lost half the ignition. And then at the same time, my magneto on the other side, which is the backup, also stopped working.

Oh, wow. It was like a cascading, strange failure. And then it was also weird, too, because on the ground, the magneto was kind of working a little bit. So it would start, but then when you went to higher RPMs, it just went bye-bye. Yeah. Anyway, yeah. So that's what it was. It took a while to figure it out. But yeah, that was an interesting triple. It happens. Yeah. Now you understand why the Pope kisses the ground when he gets off the plane, right? Yeah. I hope there's never another plane in my life that will have less moving parts.

Hopefully backup pilots and redundancy. Yeah. Yeah. And you fly a lot. You're increasing your, I guess, your risk level a little bit. Yeah. And that sort of probably triggered a lot of things. I know we're here to talk about your new venture, which sort of is something very, I'd say very different for you. It seems more like more close to home, more important. And so maybe we can talk a bit about docs. Sure. So maybe you decided to build, well, first of all, what is it?

What caused you to do it? And kind of maybe you're thinking around why this is important. Why now? Well, okay. So DocsNet, first of all, maybe I should explain the name. As a kid, my hacker name was Docs. And then I started a business in high school with my high school friend called DocsNet. It was before cloud and we were renting server, basically giving user accounts for money. Yeah. Totally legit business. And in high school, we were making pretty good money. Were these shell accounts? Yeah.

Shell accounts. Shell accounts were $5 a month, but we had a thousand users and we had all these features and people liked it. And there was actually a little community around it for a while. Got out of high school and shut it all down and started my career, but it kept a domain. And also I'm very fascinated about protocol design and the fact that the internet has really kind of stagnated a bit from protocol design. There's like not a lot of new stuff. All of it's built on top.

So there's that. And then just noticing the shift of censorship and people's privacy being eroded and all of this kind of thing kind of, and then the uprise of VPN made it really interesting to make kind of an alternative network that kind of is designed from the ground up to protect people from advertisers, trackers, malware, and foreign adversaries trying to learn about what you're doing. Well, I mean, the original promise of the internet was this incredible marketplace of ideas, unrestricted, no censorship, right? I think a lot of original or a lot of the original people around the founding of what became the internet were free speech maximalists.

It was sort of a very specific crowd. And I think over the last couple of years that, to your point, has shifted pretty incredibly, right? Yeah. And it's in the hands of a few. So, you know, for example, China is a good example is China.com, China Telecom. And those, or is it China, but yeah, anyway, either way, it's a pretty good chokehold of the entire population. Using that network as an opportunity to force people into different decision-making processes. And the internet originally, like you said, it was kind of designed with this like more open kind of concept.

So bringing that back, one of the things that I was kind of interested in doing, trying to figure out new ways of doing things that might kind of leverage before the internet Way back in the day, wireless networking was what AlohaNet or something like that. It was like wireless through packet modems and things. But now we have really high powerful Wi-Fi on the edge, incredible opportunities to create better mesh protocols and better P2P and kind of really one of the biggest stains of the entire internet is carrier grade NAT.

It's just destroying applications ability to do really cool stuff. And erasing that is kind of what I'm trying to go for. Gotcha. And also, I think one of the interesting things is we always talk about sort of foreign governments restricting internet access, right? We harp on China and Russia, but I was in a state recently that has a know your customer rule and they were requiring or blocking access to certain websites, whatever they define as being inappropriate. It could be a health website, it could be whatever.

And so we're actually seeing the creep of kind of this censorship even within the United States. It's creeping into the US and it's being done in a way that's kind of odd because a lot of times they're asking the network to solve an identity problem and the network doesn't do identity. Yeah. It has a hammer and it's going to hammer the nail. An IP address is not your driver's license and asking people to do network restrictions based off of age and things like that, it makes a lot of sense, but in a different place in the stack.

If they want to restrict things for certain reasons, then that should be done in a much different way and not asking the network to do it. So yeah, I think that we're starting to see that in the United States. We're seeing it definitely elsewhere. And it's also not just with like age restrictions, but an interesting problem popped up the other day where Google or a Cisco umbrella broke for like a major carrier because they classified that carrier as France. They're not even in France, really, as far as their corporate structure goes, but France had a rule requiring them to block like 100 plus piracy sites.

And they said, no, we're going to go scorched earth. We're just going to block France. Then there's the blast radius of that, which is okay, well, then if your geo provider accidentally identifies somebody's wrong location, then you're also now part of that policy. Oh, yeah. And it's terrible. Those classifications are always the worst. I mean, we were looking at an investment opportunity in a company that makes chatbots that are focused on more like partnership type stuff or companionship type stuff. And half of the classifications engines classified it as adult content.

And you're just sort of like, okay, well, this is like a legitimate like, you know, e-partner, e-spouse site. Like it's not doing the untoward stuff. Yeah, we've run into that a little bit too, docs, because it's got two X's in it. And then also, there's doxing. But it actually worked out in our favor because we got flagged, I think, with Claude. And then we applied with, you know, the proper means to go through their registration process. And they opened up everything for us. Oh, nice. That's great.

I mean, I think so. So, you know, this creeping sort of the creeping information totalitarianism has given birth to a whole bunch of products of which we know we're all super familiar with. Right. And VPNs is probably the most common. And so maybe it'd be great to hear a little bit about sort of why the existing tools aren't enough. Why are you building docs? How is it different? Maybe we could double click a bit on sort of your thoughts around it. Yeah, sure. So a lot of the VPNs, I would say, are just entry and exit points.

So they lease a server from a landlord and set up their software on it. And then you connect to it encrypted. And then your requests and things that you do on the Internet come from that node. And that's to solve your IP addresses, your identity problem. Yeah, it kind of is a way around that, I guess. But it's also kind of people feel like there's more security to that. And I could argue against that. But it's very basic. It's not. If you think about it, those networks are real time adjustable.

You could apply APIs to everything and all of a sudden you could say, all right, well, here's an API that your agent can use that can build an entire network. And you're just using the VPN protocols as a way to get into that area. And then inside we can create either you can call it a walled garden or a parallel Internet or wherever you want to call it. But we built an entire mesh network, 196 different domains that don't exist on the Internet. You can register domains, you can lease IP addresses, you can lease mesh addresses or public addresses, create policy, all within like milliseconds.

So it's a whole different ecosystem in there. And then on top of that, we've been able to use that system since we can kind of control it to create P2P applications that actually work. So like phone calling, video call, chat, all of those things that sit best of class on top of the app run across the mesh network and it's all encrypted. No phone numbers, no email addresses, no names, no nothing. It's funny. It's funny. And that's phenomenal, because I think we're both pretty well steeped in the security space.

And I think I come from sort of I've had it drilled into my head that if you want to have a private conversation, don't do it anywhere with anything with a battery or an LCD. Electrical signal right right and so and so what you're saying is you've actually built this sort of like transient way For folks to connect and chat without there being any record. Yeah, so you create your account You just have to prove you're human by doing a little puzzle Once you do that you get a token.

That's yours. You can change it You can delete your account make a new one But once you're in there, you can create a certificate and share it with a friend You don't really know you're doing it just says you invite friend and it looks kind of like normal But it's actually doing a cryptographic exchange with each other and then you have extra privileges I think the key lesson of PKI was actually if you tell people it's PKI. It's not gonna work Yeah, it's it's well, it's not even PKI it's private key and it's not public So anyway, you you and your friend just decided okay, we trust each other now That'll unlocks the ability for the phones over the mesh network to talk to each other directly And then all of a sudden you have all these really cool fire file transfer stuff Which is no file size limit.

The really big problem is on whatsapp signal. I Message telegram and by the way telegrams creepy all of those applications Disguise themselves as you sending something to somebody else But you're really sending something to a server and then the servers then having the other person pull that So if you upload a file you're uploading it and also also the upload infrastructure isn't always as distributed Or as it doesn't might even be a little bit slow You might be in one part of the u.s. And you're uploading it to another so the uploads a little slow so When we use our applications it extracts all of that.

It's gone so it's really no middleman no server in the middle and your phone becomes a server to Your friend and it's fast. It works. Great. You can send 20 gigabyte file to each other 20 gigabytes private network point-to-point. Yeah latency Wow with like three layers four layers of encryption Gotcha. So yeah, it's post quantum. And I mean, these are really nice tools to have Yeah, so I know when you start stacking up all of these things. It's Doc's net is kind of like this giant Toolkit for freedom of speech, I guess Freedom from surveillance.

Yeah. Yeah, we should talk about that too, which is what's interesting I mean you mentioned that, you know telegram is creepy. I I I Would say I think I find all It just seems like a lot of the VPN I don't want to say all because it's an absolutist term But it seems like a lot of the VPNs are very creepy, right and and without bashing on anyone specific I think for me the most illustrative story was in them in the news There was there was actually an issue where a bunch of organized criminals were arrested across the world Australia I believe it arrested some the US had arrested some and it turned out that the FBI had built a secure messaging app And then marketed it exclusively to organized criminal groups.

And so these guys are like don't trust signal because that's the NSA Let's use let's use this app because this is safe and only five people use it And it turned out that they all got indicted because and there are I think most of them are in prison now And so like in the u.s. Right, we've seen this activity And then if you look on any App Store and you look for VPN, you probably find 50 VPNs Yeah that are all the free to broad. Well if it's free, there's nothing free If you're not paying for it, you are the product you're being eaten alive, you know, the United States has pretty cool laws we can use them to protect people and So docks nets us-based, but we also have a Swiss company that handles all of European operations, so there's separation between the operations between the two regions we're doing other regions like that, too, so it's You know corporate governance actually matters in these types of situations, but like, you know, I just don't feel comfortable With my data going to a company that doesn't talk about who owns it Why they exist other than privacy You know, we've made a very very conscious decision that we're very open about what we're doing who's involved Why why the decisions have been made those ways and it's it's also kind of to drive community, too.

But Yeah There are creepy VPNs out there. Oh, yeah, I mean and and you know, and it's not just I mean It's funny because we like to talk about nation-states. We like to talk about intelligence agencies Right like CIA NSA. These are all super cool But in reality like the risks are a lot more low-level than that, right like and just you know for some of these there were instances at the beginning of sort of the crypto the initial crypto currency waves where People's accounts were getting drained and they were using some of these services that were in other countries and it turns out that like, you know in some of these countries you have the meshing together of the intelligence state apparatus and organized crime, right and so and so you you see where these things start to bleed into each other Yeah, I mean and then people Look at the policies for the VPNs on the logs know this and The question is like, okay.

Well whose servers are those? Yeah, are they owned by? Big brand VPN or are they leased and if they're leased who are they leased from and what are they leased from? And what are those people doing with the data and what are their policies? But there's there's that whole underbelly and and Learning about that has been difficult because I personally was not up to be, you know Building infrastructure all over the world as rapidly as I could but I ended up realizing that that was you know If it's not your equipment You can't set policy It's just you can't and then the then the next layer of policy is the the carriers and interconnect beyond beyond your Cage or whatever it is.

So it's been an incredible amount of work to work around You know the bare metals of service kind of creepy underbelly and this is like one big difference with I think a lot of folks Right is that you own the full stack from like software to wire. Yeah software to wire and uh, even all the ai components that run routing and things like that are all on our on our metal Well, I think that was that was one of the funny things. I think when we originally started talking about this was um You were you were incredibly distrusting Of large inference and ai providers.

Yeah, uh, and and you've built your own ai stack. Yeah That's and so so I mean gpus are short in shortage It's data power is hard to get everybody's anti-data center. Now. Like how did you pull that off in? Probably six months very efficiently Well, I mean we built uh so far it's like 26 sites around the world With redundant routing and uh, and how many employees in the company? I had a total I think the people involved is 12 Yeah. Wow. Yeah. Okay And we have no network administrator no sysadmin that's all done by ai so um, and it's actually working quite well I've I kind of i'm kind of on top of the nice thing though.

If you think about our footprint of like Who's involved, you know, we don't have a thousand employees with access to this network. Yeah me, you know the the uh, the blast radius of uh Just the corporate structure is very minimal. Yeah, no, that's that's impressive and I think you know the I think the the You know the thing we very rarely talk about because it's kind of uncomfortable Is just how great an extent insider threat is. Yeah, and how it's like the most crippling catastrophic cyber attacks Are always sort of the insider threat angle and now you're saying with agents like you're kind of elite you're Eliminating at least a good portion of that race.

Yeah, and it's running in its own, you know secure Environment that's not exporting data to like one of the big ai companies so all the contacts and inference and stuff stays inside and we have it secured and we monitor all our equipment with cameras and Uh, no one goes in and touches anything. Do you have agents that watch the camera footage? You actually well, I mean technically we don't but the camera companies that made the agents Yeah, yeah, we get alarms if something moves and something like, you know something goes around but um you know, we even thought about putting lasers in a couple of the spots, but Trying to make a mission impossible style i'm like why not it'd be kind of fun.

Yeah Um, you gotta have the one data center that you take people to yeah to impress them, right? That's the like the switch data centers in las vegas, you know, they got the the jeeps with the 50 cal machine guns I mean, I do I love lasers and I did this big anyway long story short uh, we did look into it and even got the the um light, um Actuated switches. So if the beams did break it would actually set off the alarms and things but uh, Just the problem was the data center doesn't turn off the lights So you don't get to see this really cool display.

So and there's no dust in there and there's no particles. Yeah So yeah, you can't bring in some talcum powder because that'll no it's like oh, let's let's smoke up the data center for the laser show That's the other thing that's driving me crazy is people like don't understand data centers at all. Oh, yeah, it's crazy I mean, I think you're seeing that in the popular press, right? Oh, yeah Oh, they drink water. Yeah, it's like come on. No, it's The data the data center has been around for a very very long time And yeah, they need water when they break and you have to replenish some of the water and there's some evaporation But it's not like they're running off of water.

Well, that's the crazy part Yeah, and it's it's just sort of like I think I think in in order of like most impressive machines humans have built Like one is probably commercial airliners, right? And two has got to be data centers Like they're so incredibly sophisticated. They're closed systems. Yeah Yeah, I mean the data centers themselves are kind of just boring buildings Yeah, and the the other thing is like the sounds i've never heard sound outside of a data center before. Yeah, maybe A chiller or something but I see videos where people are watching videos and it's like look at the hell And it's audio from inside being dubbed and people aren't smart enough to know that.

Yeah that it's uh, it's fake Yeah, I mean, yeah, that's all contained with inside these things people are being manipulated with this data center Isn't it amazing right and I think like it's going to be I think over as the years progress it's going to come out that this was all just some sort of like very interested group of Organizations or interests I completely manipulating public opinion to some end Yeah, but it's very interesting to be in the know and then watch this and then go.

Oh my goodness These people are missing the mark Have you seen I mean because you you've been to 20 something data centers in the last couple months Have you seen actual protesters outside? No, you haven't. Yeah. I know I haven't like no you see nothing It's just most of them especially like the big data center park in uh, virginia You know, you just see bunny rabbits in the field It's just a big building and a bunch of parking lot and then totally nothing Yeah Meanwhile, that's like one of the richest counties in the in the country because of all the data centers, right?

Right because it brings jobs the people that work in them aren't getting paid pennies and uh And it also brings high tech Talent. Yeah, and then also infrastructure and infrastructure powers everything you do Yeah, so if you have a data center nearby your internet's faster Things like that work better Well, and then depending on which one it is but and then the argument about data centers drinking water goes back to some quote that was like fabricated entirely like It's it's yeah Well, I think there was the switch data center built near reno and they built this big like lake out there and people Think that it lives off the lake.

Yeah, and The reality is, you know, if say you have a chiller failure and you'd have to drain all that closed system out You need to refill it. Yeah Drain it to the lake. Yeah, so anyway yeah, so it's interesting to see people like I know I work in these buildings and And then you see the conspiracy theorists and it's like I don't know we're we're way off here. I'm, sorry this feels just like the like the people that think the moon landing was faked or like it just feels like classic sort of but but I do think like this is it's coming at such an important inflection point for like the future of our country tech Yeah society in general And to just have this conversation that seems to be manipulated to some end.

It feels very manipulated and to me The future of our tech is needing to build this infrastructure now Yeah, and if we if we're crippling it then we're making bad decisions. Yeah, and we're making bad decisions on instagram posts Oh, yeah, or tiktok, right? Like which who knows where that content is or like some crazy Anti-data center music video. Yeah That's public opinion is being forced Framed by that. Well, and this is the second half I mean, I guess the I know we we had talked about sort of intelligence agencies in the big bad scary government, right?

but like and now we're talking more about Clearly an operation that's probably being done by some sort of commercial interest to manipulate public opinion Yeah, right which goes to kind of the second half of why you built docs which is like You know advertising. Oh, yeah, which is which is sort of like the like people like to think the nsa is incredibly sophisticated but you know Advertisers are probably far ahead of what they're doing And I think you're seeing now where intelligence agencies are actually using data brokers and using the techniques of ad agencies yeah, so Like all your apps have these trackers Uh, whatsapp has dit.whatsapp.net Apple even has apple telemetry.

They call it telemetry or metadata whatever they want to call it but it's Some level of it is a tracker And you can block them. It doesn't break. The app doesn't break Most websites don't have any problem with that telemetry data breaking because the rest of the site will still load But you know somebody asked me Very controversial wealthy individual said well So what someone will won't know that I purchased uh underwear off of target.com And then i'm like well, it's not really that one It's all of them when you combine all of the tracking that's being done On every different app that you're using or every different website And it accumulates and then it's being stored by who knows who where?

And being sold and then when you aggregate all those data points together You could know if you're a coffee drinker what time you wake up in the morning Uh what what your income is where you shop? Um how often you purchase things and if you're preparing for the birth of a child or perhaps for a divorce All right, some other big life event, right? They probably know these events Sooner than your spouse would yeah. Yeah So turning that off doesn't bother me at all I love all that stuff being turned off.

I hate that kind of stuff. So that's a big piece of the app is, or the network. It's not just part of the app, it's part of the network, is to kind of deconstruct all of that and get rid of it. And just kind of like the, get back to the old internet that has the ability to connect to things and has the ability to create and then also not be spied on by advertisers and governments and everybody else. Well, and it seems like we're really at an inflection point because you have the birth of essentially coin-operated intelligence at scale.

You can deploy millions, you've got the proverbial millions of monkeys with millions, or infinite monkeys with infinite typewriters. And so we're sort of looking at this world in which there's more telemetry, there's more data, everything you do is being tracked, and you can combine that with intelligence, right, to actually arrive at some pretty scary conclusions. Yeah, you can take all that and just dump it into an LLM and say, tell me about this person. You get an answer. Probably do really effective targeting against things like data centers as well, right?

Yeah. Like you start to see these mass movements orchestrated across multiple platforms and you start to think like, huh, that's pretty clever. Yeah, but people should be targeting those, maybe not the data centers, but the tracking is pretty bad, ad tracking. That kind of stuff really can go away and your life would be better. Yeah. So, yeah, at the core, it's like trying to recreate that vibe that the Internet used to have and then see where we could take it. With bigger resources and things like that, you can start creating more protocols and things that don't exist yet.

Yeah, absolutely. I mean, the Internet succeeded because it was free. And so thank you for trying to make it free again. Trying, but you have to pay me. Yeah, exactly. Just kidding. Awesome. So I don't know if you know this, but our paths crossed a long time ago. I was looking back through some of my archives and I did a lot of work sort of in the early 2000s on DDoS bots. So stock will draw TFN, TFN2K, and then obviously a large denial of service distributed attacks.

And you were at the forefront of building a lot of the solutions to prevent that, like at Perlexic. Yeah. Like, how did that get going? Well, actually, it kind of goes back to DocsNet. In high school, people were attacking my servers. And so we had to figure out better ways to protect them and techniques and tricks and all the different things you need to do to survive a DDoS attack. And I started writing kind of my list of what you would do to survive, and then somehow I presented that at SANS.

And then it got in Google, and then people started calling me. And then we started dealing with real cybercrime, extortion, people trying to take, you know, one company sells watches, the other one is attacking the other watch company during Christmas sales season. Or, you know, Super Bowl is about to start and all the gambling sites go dark. Yeah. Things like that. And it got pretty serious. And so I built a whole company to protect businesses from these attacks. And the way I looked at it is, like, you need big scale.

You need to have the biggest, most expensive stuff, and not everyone can afford that. So we'll turn that into kind of a leased service as, you know, one of the first cloud services to exist. And it was pretty very successful. It still exists today. It's Akamai Perlexic. Yeah, absolutely. They even kept the name. It's just rare. And then we did DefenseNet later on, which became F5 Silverline. So in between all that, I really like big scale stuff and bandwidth. And so the other one we did was called Big Gravity, and that was one of the first, or the first, video-centric CDNs.

So big file, big object delivery, and we had some really great customers there early on. People got it, and then we had a lot of tech around it to save people money. You know, people didn't realize it hit play, and you're just downloading a progressive file, but before all of this kind of media management stuff existed, you're just wasting bandwidth just to scroll into the middle of a video. It costs people money. So anyway, we created all kinds of cool stuff around video. And so as far as like building a new network for consumers, well, actually, and it's a little bit for bots too, there's actually a lot of agentic networking components to it.

It's right up my alley because I love building big infrastructure. I love building scaled stuff. But now it's kind of like twisting it. Instead of doing this for corporations, we're doing it for people. I haven't really seen a lot of that being done, at least well. So it's going to be fun to see kind of how people adapt to it. But I think one of the more interesting things about that is we might even see some DDoS on our network. I'm sure you're ready for that.

Well, inbound from the Internet, what happens if somebody gets onto my network and starts using it inside? And I have to deal with those silly people. But, you know, the truth is, even if you have privacy, there's still laws. And you need to at least follow them as much as you can. I think one other interesting thing, so, you know, it's sort of the – and this is the kind of spending years talking with you. One other interesting thing is that you created an artwork that actually hung at the MoMA.

Yeah. And it's part of the permanent collection. There's two pieces. Yeah, a college project. I studied philosophy. Oddly enough, I studied – Everyone does, yeah. Yeah, I studied the art of aesthetics and some other stuff. A lot of it was like theoretical AI. Yeah. And I'm like, wow, actually a philosophy degree makes a lot of sense now. Totally. This is helpful. But, yeah, I started mapping networks because I was doing – I was a penetration tester and hacking, ethical hacking into other companies. And I found that casing the joint was the easiest way to start.

So you make a network diagram of where all the stuff is and how it all works and what you think is where. And then I started automating that. And then I got mouthy at lunch one day and bet some guys that I could probably use the same technique to map the entire Internet. And that started a $50 bet that I was pretty sure I was going to lose. And then I got hit by a car on a bicycle. Oh, okay. Yeah, and couldn't walk. And all of a sudden my life got really slow, and I sat down and started doing the code for that.

And that became the OptiProject. And that ended up in not just MoMA, but like Boston Museum of Science, all over the world in different traveling exhibits. And then I did animations of that too. And it's pretty neat because that project has gotten a lot of attention, not just because it looks cool, but there's kind of really interesting history in the video. As things are happening, you can see companies crash and burn and all kinds of stuff. And I was lucky enough to be able to talk to Vint Cerf about it and have an opportunity to ask my questions to him, and it's pretty cool.

That is awesome. And you've probably accomplished two of the harder things to do in this world, which is, one, starting a successful tech company, and two, being a successful artist. Being an artist, a successful artist, is probably much more difficult than starting a tech company. Yeah, it was actually kind of strange. I went to the opening in MoMA, and it was this giant party. And I was one of the artists. And they had an exhibit, and I had the piece in there. And it was just kind of like, oh, man, this is kind of cool.

There's a DJ, there's drinking. And then the curators, they took me to some really interesting parts of MoMA's other properties that they own, and it was a really cool experience. But the next day, I kind of had this imposter syndrome where I just felt super depressed. I'm like, wow, am I ever going to experience what an artist feels like ever again? Because this was kind of like accidental, you know? But, yeah, I got to do it again. But it was, yeah, a very strange, surreal moment in life.

Probably the only person in tech that can say that. Yeah, it's interesting. I have a lot of artist friends that are probably much better artists than me. And they're all like, what? How did you get into MoMA? Yeah, how did that happen? Yeah, and part of the permanent collection, too, which is neat. Yeah, you're one of a few hundred or thousand. I forget about that sometimes. I also forget about the book, too. That is true. Yeah, we have not talked about the book. I don't know, do you want to cover that?

Sure. We'll throw this on the list of great stories. Well, I keep forgetting about that it exists. And recently, so I have an AI-powered webcam that watches boats in Miami. Yeah, yeah, yeah, yeah. And it's 24 hours a day. It's been up for almost a year now. And it's got a little chat community of people on there. And one of them figured out who I was. And then the other day, like, I just finished reading the book about you. I'm like, what? What do you mean?

I'm like, oh, shoot. That's right. They figured it out. So what is the book about you and what is it about? It's been so long. So it's about, I think, the beginnings of Prolexic and tracking down these cyber criminals in Russia. And specifically, these guys, there's like a couple of them. They were extremely aggressive, extorting and damaging a lot of our clients. Yeah, yeah. And so we just kind of had enough of it and said, this has got to end. I mean, we had Christmases destroyed because we were just dealing with customers and network outages.

Well, and these are like real criminals. These are not just hackers. These are people that are doing human trafficking. They're trafficking narcotics and they have a side hustle. There's some stuff going on, yeah. It's like real criminals, right? Yeah. And then they went to prison for, I think, eight years, hard labor or something like that. And I don't know what happened to them afterwards. But, yeah, that was really – we went out to Russia and got turned around. We were there in Moscow for a while.

And then there was some really interesting signals of like, you need to leave now. Yeah. So we left. But, yeah, that whole book is about that whole saga of what happened through that period of time. It's really weird because, you know, when you live it, it seems kind of normal. But then you look back and go, oh, geez, that was kind of crazy. Yeah. I tell you, every time we talk, I hear new stories. Let's not talk about Burning Man. That could be a whole episode itself.

Yeah. Actually, when was your first – Burning Man is next week? Yeah, it's this coming weekend. Yeah, yeah, yeah. Are you going this year? Yeah. Awesome. I think it's year 14. Year 14. Somewhere around there. I can't remember. We lost a couple to COVID, so I don't remember which ones I skipped. And you have a famous car. Well, my partner has the famous car. Yeah, okay. So Pablo created the Mayan Warrior, and I created the Mayan Warrior Camp. And so we were partners in that whole crazy thing for a while.

Boy, that car went from a fun project to a whole global phenomenon. People paid money to go see it. It's crazy. But no, it was a lot of – what I was thinking about this morning is when you're passionate about something, it kind of creates that aura around you, and then other people pick up on it. And we were very passionate about that project. And I think it kind of inspired a lot of folks to do really interesting other projects that are similar to it. And this is the first year we're not taking it to the playa.

Oh, wow. It's not going. We're done. Will you be doing your laser show, though? Nope. Wow. All right. Well, there'll be something, but it's not mine. Pablo's got a beautiful art piece. There's two art pieces that he's bringing this year that'll be pretty cool. Awesome. No more traveling superpower soundstage. That's gone. How's the vibe out there now that you can get Starlink Internet access? It's actually – I mean, you can argue against it and you can argue for it. I find that if you're running – well, for me, like when you're organizing a camp or dealing with, you know, oh, we ran out of water on the art car, it helps because you can communicate.

Yeah, totally. Because we always bring high-powered Motorola radios. Yeah, yeah, yeah. People are using shortwave, too, right? Like it's sort of – Yeah, so you have the radios, and then that's kind of fun. And, you know, it reminds you of being a kid. Yeah, yeah, yeah. But it's also like, you know, it's still a communications device. So out there, I don't know, it takes the edge off when it comes to like finding things or organizing things. People don't get stuck at the gate very long anymore.

Yeah, yeah, yeah. But if you're sitting there watching movies, that's probably the wrong use for it. You don't go to Burning Man to Netflix and chill. No. If you're using it for anything other than like local communications, then maybe it's like the wrong use. But it has changed it a bit. Awesome. Well, as you said, right, like you pursue your passion, and that takes you down all sorts of amazing rabbit holes. And you are clearly someone that is always pursuing his passion. So thank you so much for stopping by.

I am an incredibly happy DocsNet user. And thank you for building it. It's good to get that old Internet vibe back. Yeah, well, it's just getting started. And thanks for trusting me to help build it. Thanks again for listening. And I'll see you in the next episode. A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see A16Z.com forward slash disclosures.

Go deeper

• All A16Z episode guides