The Changelog: Software Development, Open Source
You'll rent chips and be happy (Friends)
Fri, 18 Oct 2024
Zac Smith left his role leading Equinix Metal in June of 2023. Since then, he's been thinking deeply about the present and potential future of data centers, OEMs, chip makers & more.
Welcome to ChangeLog and Friends, a weekly talk show about hydrogen non-bombs. Thanks to our partners at Fly.io. Launch your app on Fly in five minutes or less and join three million others, including us. Learn how at Fly.io.
Let's talk. Hey friends, you know we're big fans of fly.io and I'm here with Kurt Mackey, co-founder and CEO of Fly. Kurt, we've had some conversations and I've heard you say that public clouds suck. What is your personal lens into public clouds sucking and how does Fly not suck?
All right, so public clouds suck. I actually think most ways of hosting stuff on the internet sucks. And I have a lot of theories about why this is, but it almost doesn't matter. The reality is like I've built a new app for like generating sandwich recipes because my family's just into specific types of sandwiches. They use Braunschweiger as a component, for example.
And then I want to like put that somewhere. You go to AWS and it's harder than just going and getting like a dedicated server from Hetzner. It's like it's actually like more complicated to figure out how to deploy my dumb sandwich app on top of AWS because it's not built for me as a developer to be productive with. It's built for other people.
It's built for platform teams to kind of build the infrastructure of their dreams and hopefully create a new UX that's useful for the developers that they work with. And again, I feel like every time I talk about this, it's like I'm just too impatient. I don't particularly want to go figure so many things out purely to put my sandwich app in front of people.
And I don't particularly want to have to go talk to a platform team once my sandwich app becomes a huge startup and IPOs and I have to do a deploy. I kind of feel like all that stuff should just work for me without me having to go ask permission or talk to anyone else. And so this informed a lot of how we built Fly. We're still a public cloud.
We still have a lot of very similar low-level primitives as the bigger guys. But in general, they're designed to be used directly by developers. They're not built for a platform team to kind of cobble together. They're designed to be useful quickly for developers.
One of the ways we've thought about this is if you can turn a very difficult problem into a two-hour problem, people will build much more interesting types of apps. And so this is why we've done things like made it easy to run an app multi-region. Most companies don't run multi-region apps on public clouds because it's it's functionally impossible to do without a huge amount of upfront effort.
It's why we've made things like the virtual machine primitives behind just a simple API. Most people don't do like code sandboxing or their own virtualization because it's just not really easy. There's no path to that on top of the clouds. So in general, like I feel like and it's not really fair of me to say public cloud suck because they were built for a different time.
If you build one of these things starting in 2007, the world's very different than it is right now.
And so a lot of what I'm saying, I think, is that public clouds are kind of old and there's a new version of public clouds that we should all be building on top of that are definitely gonna make me as a developer much happier than I was like five or six years ago when I was kind of stuck in this quagmire.
So AWS was built for a different era, a different cloud era, and Fly, a public cloud, yes, but a public cloud built for developers who ship. That's the difference. And we here at ChangeLog are developers who ship, so you should trust us. Try out Fly, fly.io. Over 3 million apps, that includes us, have launched on Fly.
They leverage the global anti-cast load balancing, the zero-config private networking, hardware isolation, instant WireGuard VPN connections with push-button deployments, scaling to thousands of instances. This is the cloud you want. Check it out, fly.io. Again, fly.io.
I accidentally got myself into three community orchestras. So I'm like super busy.
Three community orchestras. What do you play?
I play bass. Okay. Stand-up bass.
Cool.
And one I've always been in. Well, not always. For the past three years, I've been in what's called the New Conductors Orchestra. So we're a community orchestra that gives a platform for emerging conductors to practice. And then we do three concert series a year. And then... I joined this one in the summer because they just had like a really good program.
It's basically high school and college music teachers that come during this when they're on break and play together. But they always need. I mean, this is the story of my life. Everybody needs a bass player. Right. So they needed some more bass players because we were playing this smaller symphony, which had like as many basses as you can possibly get on stage.
And then a buddy of mine harangued me into a different one, which is a chamber orchestra. So, you know, that's what I'm doing tonight. And it's kind of odd for me because we're rehearsing in Lincoln Center in what used to be the Juilliard dorms, which is where I live when I first came to New York. And I haven't been in that building in like 25, 30 years. So all kinds of memories.
I'm going to come back when I was just a pipsqueak.
So tonight is the first time back to these dorms?
Yeah, because I guess they borrow a rehearsal room there. And I haven't been back since I was 18. You had to live in the dorms freshman year at college, which was good. I was 17 years old and had no right living in New York City on my own. But then I moved out of the dorms and didn't really ever visit again.
Any jitters to go back? What's your feelings? Nostalgia? Fear? Concern?
Oh, no, I'm all good. Me and my college years are fine. We're at peace with each other. Good. No jitters.
That'll be fun.
Yeah. Instead of begging to take my money for school, now they just try to beg my money to give it to them. So it's different. A lot less pressure.
bass is an interesting instrument because we take it for granted. Don't we? Like it kind of, when it's, it needs to be there.
And it's like, it's all about that basement.
It is about that, but you kind of forget that it's there when it's there. And then when it's gone, you're like, something's missing. That's very important here.
Do you see the little line between my professional career doing internet infrastructure that is absolutely invisible and nobody realizes it's there unless it's not there? And then everybody's like, wait, it's not working anymore.
It's not working. The website is down. Yes, definitely parallels those.
I like a supporting role in my life.
Fantastic. Do you ever get solos, though? I mean, at least in rock bands, the bassists will get a solo every once in a while. But in the orchestra, I assume, probably not.
We have a few, but not a lot. There's not a lot of double bass solos. But it's OK. I don't mind. I like playing in the section. Frankly, what's most exciting for me is playing in an orchestra with like 80 or 90 other people, because it's this auditory, very physical representation of what a good organization feels like.
because you have 50, 60, 70 people of varying backgrounds and things and whatever, all getting together, doing something. And you can tell when it works really well and they know how to work together, which is so much harder when you're building companies or working in organizations. The feedback loop is so much slower than it is. So I like it, being in a group. I don't need to be out in front.
You kind of feel it in real time because if you're off pace, coming too early, coming too late, not there at all, you know, it's pretty clear.
Yeah, it's pretty obvious.
Right. And you're letting people down.
Were you a fan of the, I know it's not the same instrument, but were you a fan of the movie Whiplash? Did you see the movie?
I didn't see it.
Bummer.
Yeah. Was that the one that was about the Juilliard school thingamajigger? I know there was one. Somebody said I should see it.
Yes. Let's see. Who's the Whiplash movie cast? Miles Teller, J.K. Simmons, and Paul Reiser. And a few other actors that were in there, too. But those were the main ones that were in there. Miles was the student, the new student. He was playing drums. And I believe the famed song they played in the movie was called Whiplash. Ah, okay. And it was jazz. So it really had this very fast-paced...
double time kind of high tempo symbol writing. And yeah, like just a horrendous beat to, to maintain as a drummer. If you're good, not horrendous. If you're up and coming, like a brand new student, like miles was, is his character, Andrew, then you're on, you're on pace. But the relationship between the conductor, JK Simmons, his, his character was named Terrence.
The relationship was very violent. For example, he was very kind to him initially. And then he ended up throwing, I think, a chair at him as his first attempt to maintain the tempo with everybody else. Like you're seeing like this importance of being together, this feedback to being tight. His feedback loop was probably pretty on for our ears.
But to the conductor's ears, he's like, no matter what you would do, you were wrong. And so it's a very interesting movie, I would say, as somebody who is outside that bubble, whereas you're inside that bubble being part of a larger band that plays fantastic music together in concert. Well, for people, you know, so I thought maybe you've seen that.
Well, the intensity that you have to have to be in any high performing thing, whether it's like, I'm sure we started off talking about the Jets or something like, man, you know, like any high performing sports or music or, you know, I'm sure other intense engineering organizations or business or sales teams, like it can get, you know, people when they all want the best and they have a lot of A players, you know, all shooting for the top, right?
Yeah. can be pretty intense sometimes. So there was several uncomfortable moments I had when I was at school, but I, you know, you develop thick skin as an artist because you basically spend most of your time paying people or seeking out ways to get criticism.
Wow.
Right. Which is like for people to tear you down and tell you, you don't go to a, a lesson with your teacher for them to tell you about all the things that you're doing. Well, You want them to help you be better, which requires criticism. Right.
And that's for me as a as a manager was so difficult to transition from being like this music major, you know, criticism oriented, you know, self-criticism, you know, dominated music.
you know, human to like, that was my relationship always with people who were senior than me to then doing that in business world where it turns out that that was not always motivating to people where you would constantly just forget to tell them a good job, but you know, where you could improve is here. You just go immediately into all the places that it could improve.
And so I had to constantly remind myself when I, you know, over the years as a manager to like, remember to tell people the things they're doing well.
comma here's where you could do other you know the sandwich method yeah you gotta sandwich it up it was is there a myth isn't there a method called the sandwich method jared yeah it's compliment criticism compliment that's right yes i'm not sure if it's actually called the sandwich method but i think it is okay cool good job yeah that's how you could have done better but you did great i yeah but you did great
Yeah, that's something I've learned as well as a coach, coaching young children coming up playing basketball. I coach basketball and baseball. And as a driven, ambitious athlete that I once was, used to just a guy yelling in your face when you mess up. I just figured that's what everybody's okay with and it works all the time.
But there's a lot of situations where, maybe not sandwich method, but people don't realize that when you constantly criticize someone, It's not because you don't think they're good. It's because that's what you think they're there for, is for you to tell them the things that they could improve. And every once in a while, you got to focus on something they're doing right.
Because especially with young people, they struggle with that. But I've never thrown a chair at anybody. I don't know, Zach, did you ever have a chair thrown at you from a conductor? I mean, that seems like it's over the top.
I had a baton thrown at me. How about that?
Okay. Close enough, right?
No chairs. No chairs.
I had an eraser thrown at me in high school from my teacher. Oh, yeah. I think everybody has, right? Yeah. Several. Oh, yeah.
Well, it's an interesting topic around criticism, right? What you brought up there around sports, my sons play sports and trying to seek out the right coaches and whatnot who they respect, who help them be better and also can be part of that motivational circle. I'd had a few mentors growing up where I'd had one which was a cheerleader.
And then I had another who was very much on the criticism angle. And I remember coming back from my first lesson with the criticism guy. And I was just crying. And I was like, this guy sucks and whatever. And my mom said, well, you should go tell him how you feel. And I did. I was like, why did you just tell me all the things that I did bad?
And he says, well, if I didn't think you could be better, I wouldn't have bothered even telling you that. Right. And that really stuck with me as like, well, criticism is there to help you improve. Now, that has to come from a place of positivity. That can't come from a place of meanness of spirit. Just tearing you down because somehow that makes me feel better.
So obviously you have to, you know, surround yourself or be able to tell the difference of mentors or leaders who are coaches, you know, who have your interests at heart and actually just want to see you do better. versus those who want to put you down. But I guess that's a maxim for a lot of life, right?
Totally. For sure.
Man, we got really into the deep stuff.
We did get into the deep stuff. That's the good stuff, though, you know? That's the good stuff. Straight in the deep end. You know, why small talk when you can deep talk? That's what I think. That's right.
Well, let's go to maybe some deeper talk. The last time that we talked, Zach, I would say you're on Friends because I would consider you a friend. I never made it to the data center like we talked about, I think, at the end of the podcast. That's okay, though. I love a good data center tour.
To just layer on some of the conversation we had, we had a great conversation on Founders Talk 84 back in 2021. So that's the last time we talked. This is like three years. almost three years to the month to the day almost. And the last time I talked to you, you were really on this new trajectory with Equinix metal.
Your company packet had been acquired by Equinix, the very large behemoth global dominator in infrastructure, you know, servers everywhere, data centers everywhere. And this was a, a great thing for you and great thing for your company and, And here you are essentially being acquired by Equinix and you go from startup to global company.
And so preceding this conversation, you know, managing, criticism, stuff like that, I got to imagine that this journey into the big world, the bigger world of automating bare metal, took its toll in good ways and bad ways.
What, this gray hair? No. Okay.
What happened with that acquisition, with your role there? Did you enjoy it? What made you take a step away? Because you're no longer there. That was 2023, a year ago, I think, roughly, you stepped away from that role. Catch us up and let's go from there.
Yeah, man. Well, let's see, 21, man. That was just the end of the mask-wearing COVID times, I think, right? I think we're just getting out of our little cages at the time.
Peeking out like the groundhog. We were peeking out.
And we're peeking out. We're like, is our shadow gone?
Right.
Well, I would say I've always been a small company person. I've had the pleasure of building a couple of startups and being part of, I think, the biggest company I was ever at for any substantive period of time was no more than 100 people. Packet at its heyday was about 120.
So I was really used to small organizations that didn't really have advantages in terms of global scale or amount of money or you know, brand or anything else like that, but you had agility, right? So you kind of like organized around speed, organized around ability to change. And that's what I love about startups because I love to experiment.
I love to lean in with new ideas and I love to go fast. I'm a person of action. And when we became part of Equinix, I would say after about one year where we had a little bit of a grace period, we were off to the side and kind of got our bearings. This is a 15,000 person company that had a very different kind of culture and They were a market leader.
They had decades of success doing their business. They had a lot of kind of experience and leadership there. But that comes with bureaucracy, right? It comes with more people, more checks and balances. They had a matrix structure, which I'd never heard the word of. But basically... you know, it took me a while to figure out, well, who makes a decision here?
You know, and the answer was, well, you know, everybody has to get aligned and then, you know, decisions can be made or something. Right. And so for me, that was super different. And I will admit, like at first it was very, very frustrating. And I had to take a, step back and really learn some new skills, which once I took kind of a learner's mindset or growth mindset, I really enjoyed.
I'm not going to say I enjoyed all the experiences of cross-functional work because it's a little bit different than my personality, but I enjoyed the learning of how to figure that out. Because you basically had to think, hey, how do you get 15,000 people at a company to march in a direction together? Right. Or to do things that might feel uncomfortable or to try new stuff.
And the answer is like not that much different than marketing that you would have to do to go convince customers to try your thing. So a lot of those skills I started to use internally and figure out how to explain and communicate things. in different channels and in different ways. And that was super powerful. I think those are skills that I'll take away with me and other things that I do.
And then you also get a little bit of a different lens on the world. Running the digital side at Equinix, I was more of the technology-focused person at a big global real estate company. And so I had a first person view working with our partners like Nvidia, like with Intel, like with Dell, with VMware.
Those were all very different conversations than I would have had as a startup because we had something that they really wanted and needed. It was very different than, you know, conversations that a startup has with those kinds of big organizations. So that was super cool because we did some awesome stuff and we were able to unlock things that would have never been possible
I think, for a smaller organization. But then after three years, which was, I would say, a very measured process that we had with Equinix, they wanted to acquire the business to help make their platform, which is hundreds of data centers around the world, more digital.
And so the first two or three years were about kind of working through at the pace to integrate some of those capabilities from the packet acquisition into that broader company. But we always knew that there would be kind of an integration process.
And then, you know, for me, after three years, it was the right time to go, say, use my skills elsewhere, primarily because that kind of integrate and scale function as a leader is not actually my best quality. Yeah. I'm more of a driver and creator rather than a kind of a manager and a process person. So for me, that was a good transition. I think it was right for the company as well.
I was a little bit more mentally prepared than my first startup in 2012 I sold with Raj. That was really emotional because he was like your baby. And you're like, don't touch my baby. But I was much more mentally prepared going into the transaction with Equinix around, hey, this is your company now. And how can we make it work for you? And how can it last well beyond me? which is the point.
And so it was easier to step away when the time was right. Still had a lot of, you know, relationships and touch with the customers that I, you know, obviously miss, but it was the right time and it was a good, healthy transition, I think, for all parties.
I think it's so cool to build something as a startup that is attractive enough, not just to be acquired, but to be a hopefully integral function of scale for a already decade successful global leader. You know what I mean? Like how bad ass is that?
That's pretty cool. Yeah. Right.
That's cool.
Well, I, uh, I took away from it something, and I saw this with a lot of larger companies, and it really gave me this newfound respect both for market leaders and for startups. And I used to see them as very in conflict. And actually, I can see why great market-leading companies have muscles around acquisitions.
Because in market leading companies, there was a really good book, I think it's zone to win. I'd ever read the book zone to win. It basically talks about, you know, there's four zones, a company's lifecycle or where its products are. And the upper right zone in the quadrant, just think of a two by two, is the kind of market leading. Like it's, you know how to sell it. You're leading the market.
You kind of, everybody knows that you do it. Like that's where it's all working. Then you write kind of like below it, you have the stuff that's like starting to work. And I think they call them like horizon one and horizon two, something like. It's right here or it's like a future opportunity. And then they're over on the bottom left. And I'm totally bastardizing this.
So please, like somebody Google it and it will make a lot of sense. But on the bottom left, you have like innovation, which is this kind of like incubation zone.
uh which is not even a thing yet we don't even know right it's super weird it's super scary we might fail probably will fail we're not sure and then you kind of got this like adoption zone where it's like oh it's starting to work what i realize is that you know these the zone up in the right hand side where it's winning you know it's that way because it actually focuses you know it figures out what works for example with equinix interconnected co-location works
They dominate the market. They've been doing that for a while. Everybody knows that that's what they do. And they then build processes that keep that market leading around it. And so internally at the company, and I'm sure that this is at other market leading companies like, say, Google right now with their ad business. there's a very protectionist approach.
Like don't distract from winning at the thing that is working. Like don't get us off track because we're nailing it every quarter. The customers love it. We're dominating the market. Keep doing that. But what that can do is that can obviously take something that looks not working yet, e.g. incubation or some sort of innovation and kill it. Right.
Cause it's a distraction to the market leading component of the business. And so, um, that was definitely one of the challenges we had at Equinix was they were trying to figure out how this 20 years of market leadership they had around a single thing, you know, well, what happens to the next thing?
You know, what did they say about like, there's this, a funny phrase, you know, it's like, I went broke slowly and then all at once, you know, it's like, it's like it was working and I was the number one until, I mean, look at Intel right now until I wasn't. I think that there was a good self-awareness at Equinix. Like, Hey, how are we going to upgrade our,
products and our experiences for the next thing. And so acquisition is part of that. And I think startups play a great role because there is no past. all they have is future. So there is no, you know, zone that is swapping away innovation. Everybody's like all hand, all ideas are good because we need to figure out how to survive or, you know, make the next wave.
And so I think that's like super interesting. The companies who have figured out how to do that over and over and over again. And like, whether they figure out how to incubate, like what, you know, maybe certain companies do with labs divisions or open source or side projects or through MNA. Like I, I had a, very much, you know, a renewed appreciation for that, for that cycle.
So you're saying that being able to flex the muscles of M&A or incubation is, is something that I think startups should, since startups have no past, they only have the future. How can startups, they're all incubation. They're not really performance productivity because they're learning what should be
Yeah, they're not in the performance zone. That's what it was called. You got the performance zone.
Well, I, you know, I have chat GPT here. And so I just said, chat GPT, summarize zone to win.
Summarize in three sentences what zone to win is all about.
And so that was the performance zone, the productivity zone, and the incubation zone.
There we go.
There you go.
Well, yeah, I mean, that's what we say with startups. They're finding product market fit, right? Startups are finding fit. They don't even have anything working yet, which is an awesome muscle.
Sorry, I was trying to get the fourth zone. He said there's four quadrants, and he said three zones, and I'm over here without ChatGPT wondering.
Incubation, transformation, productivity, and performance.
There you go. You pass. You have an A. So you mentioned Intel and you kind of hinted at some sort of demise of sorts. There's definitely turbulent waters for them in our side chat on LinkedIn, which is how this conversation came back. I'm like, hey, it's been a long time since we talked. I'm sure lots have happened. Is there anything we could talk about? And you said, absolutely.
And one of the things you mentioned was chip wars. So I got to imagine this innovation mind you have, this vision, find it, achieve it, do it, not integrate it because that's not really where you thrive. And then being blessed with this world of Equinix where you have global access to AMD, NVIDIA, Intel, all the major players, a whole different conversation than you had before.
What was the secrets? What did you learn? What can you share with us that we would never know that only you would know? That's not part of your NDA.
Yeah, exactly. Well, I mean, I think what I learned, I'd always respected Intel as a company, but what I learned working with them pretty closely was how hard of a task they had ahead of them, right? I mean, first off, we think of market leading for 10 or 15 years in something. They've been market leading for 40 years.
It's really, really hard when you're used to own a 90% of the market to change anything. And I think that that has to be You know, appreciated because I didn't appreciate how complicated cultural changes when it starts to go off track or when you forgot or couldn't figure out how to innovate the next thing because you were so busy making all the money. Right.
And like, you know, you know, winning all the market. Right. Because why would you do anything else the next 86 years? you know, great data center CPUs and desktop CPUs because it's working. And so that's what I kind of appreciated out of that. I would say that I was surprised. I did find that all the chip companies really struggled.
You know, it's kind of like blindingly obvious and at the same time, not in the realm of possibility that the chip companies all needed some sort of recurring or subscription revenue. They have huge, huge increasing investment cycles around CapEx, whether that's designing their technology or that's producing it with Intel and their fab.
But they have these huge things where then they go and then they just sell the chip one time. And increasingly, you know, they're competing against, I believe, their biggest customers, hyperscale clouds, which are also making their own chips. But they have a nice, juicy recurring revenue model. So they get to make money on those chips every day or for years and years and years.
And I believe that that look to me – and I was very surprised. I was like, hey, y'all, we should work on that together. And – Every chip company, when we would talk to them, was kind of like, yeah, we should, but could you guys just buy the chips from us, please? And we're like, no, let's practice the other thing. Let's move to the Apple subscription model.
Versus the buy your phone through Verizon. Let's figure out how to do that. And let's also solve things like circularity and full recycling and reuse of components and all the other stuff that goes with that. It can be better, not just more stable revenue. It can also just be better. And it's kind of what customers wanted too. Because customers kept expressing, hey, I want access to this stuff.
I don't want to be sold more kit all the time. But it was so hard for the organizations, and I still think it is. I haven't seen any meaningful approach. And I'm kind of curious right now, especially given the dominance that NVIDIA has within its business you know, if it's going to try and change the model of transferring its IP, you know, the GPUs on as a service basis versus on a CapEx basis.
And I just, I haven't seen that yet. And I'm quite surprised by that.
You mean like cloud-based GPUs kind of things turn into some sort of service kind of thing where they're not giving you the actual hardware anymore. They're just going to give you access to the compute and power of it.
I mean, I wouldn't, you know, we can get into the crazy wild idea. That's part of this podcast at some point. Let's do it.
That's what we're here for.
I wouldn't be surprised. I think that the verticalized model of make your silicon and offer it as a service in data centers is amazing. working really, really, really well for a couple of big clouds right now.
I wouldn't be surprised that some of the chip companies do something like that, whether it's through a proxy like Intel, I believe, spun off part of its business in a joint venture called Articulate with an infrastructure fund. And I know that NVIDIA has just invested heavily in a lot of AI data center businesses, right? So they're making their moves to support that distribution.
But I wouldn't be surprised that they would go more direct if they could find a way to kind of get their heads around that. But just to go back to as a service, I think it's even more basic than that. And I'm like, just think of Apple, who I think just has a master class in this. They used to sell you phones, but now I just go on this upgrade cycle all the time.
And they take back the physical device from me and finance it on their Apple card or whatever. All I know is every –
18 months to two years it's super super easy for me to get a new phone and the other one goes back into the back of the store on roller skates you know or however like the apple cut comes out and brings me the new one and um that's for a physical device i hold my hand all day long the access of like most data center equipment is much easier because most people never touch it they never see it it just goes into a big data center in one of like 40 markets around the world
which means we know exactly where it lands. Nobody really touches it except for service providers. And so I think there's a real opportunity to kind of do an upgrade cycle or a kind of circular model where it's like, instead of buying computers with
an intel chip in it you buy a subscription to intel technology that allows you to have always access to the latest stuff and that could be in your data center or somebody else's data center i think that that's you know one part is financing and service model and the other part is like where does it sit but you know these things aren't going into random places they're going into pretty well-known places
Was that as a concept was difficult to sell or was it like the actual, how would we get there? Because we are this behemoth that's doing this other thing and we're in the performance zone.
It's the concept. It's just like concept. Yeah. It's the concept would come back to, well, how would we, you know, sell you the chip first, you know? And it was like, no, no, no, no, no. That's not how it's going to happen. Like, well, who would be the OEM? You know, it's like, well, maybe there wouldn't be one.
You know, and like, you know, just those things that their whole departments of managing the OEMs would be like, well, what do I do in that? You know, and so I think that those are the conceptual issues that got in the way. Not like, could we could we physically figure out how to do it? I think that wasn't actually the issue. We almost never got there.
It was like, yeah, of course we could if we wanted to. But why would we? Because of all these reasons.
And some of those reasons, it sounds like we're justifying certain jobs or or even sub orgs inside of that, which would no longer be necessary. Or I mean, I wonder because to me, of course, maybe I'm also a small business kind of guy like you are. But to me, it's like, let's just let's just try that. Let's just give it a shot. Run with it.
Yeah, yeah, I totally agree. You know, the problem is that, you know, and I noticed this when I was at Equinix as well, an experiment is almost not that useful because it's so tiny. Oh, so what was wildly successful and we managed to do thousands of these things in this awesome new way, it would still be like 0.01% of our revenue. Wouldn't be meaningful.
So we should, you know, we need to do something at real scale, but we can't get behind something at real scale because it's super scary to do something that you've never done at scale. And so there's this little chicken and egg. It's a hard thing. It's like obvious, but it's hard to get through. And so I think that's why some companies have practiced. They recognize it. They're self-aware.
And they practice that through M&A or they practice it muscle. Like I really love the Google SRE handbook chapter. What is it? Two outage points or something like that, you know, where it's like people's bonus and the SRE team is not built around 100% uptime because that means you're probably going too slow and not taking enough risks.
So it's more like how do you balance outage points with customer pain and There's got to be something that's not 100 and not zero. It's something more in the middle. And I think that's kind of difficult in scaled businesses because you're really – you can tell it like even with Intel right now. They don't get a lot of room in the public markets to mess up.
And so experimenting is hard in a big business.
Mm-hmm.
Okay, we're here in the breaks. I'm here with Firas Bugadige, founder and CEO of Socket.dev. So Firas, you put out this fire post recently on X. And I'm going to paraphrase. You say the XZ package backdoor was just the tip of the iceberg. Give me just a peek behind the scenes of this incident and what you mean by it's just the tip of the iceberg.
Yeah, so I think the XZutils backdoor was really eye-opening to a lot of developers. It showed the vulnerability of the open-source ecosystem. You had this maintainer who had been tirelessly maintaining this package for 15 years, who was targeted by nation-state actors, who created, like literally, it's like a spy movie, right? They had multiple personas.
fake personas that were contacting this poor maintainer and, you know, working on him psychologically to convince him over the course of two years to add them to the repository and give them publish permissions. And they did this through a bunch of kind of negative messages, but also by being helpful and by sending good positive pull requests.
It's really like I really think it's out of like out of a spy movie, just kind of the level of effort that they put into this. And what they were able to do is get access to this package. This is built into pretty much every Linux server out there. And what this would have let them do is it would let them SSH into any server and run any command on the server.
without knowing the password, without being authenticated to the server. So this would have been like a world ending, potentially kind of an attack, right? It would have been probably the worst attack we've ever seen. I'm not exaggerating. It could have been that bad, but we were lucky. Through a total accident,
This backdoor dependency had made it into the beta builds of some popular Linux distros, but it hadn't made it all the way out to the stable version yet. And a developer who was testing out the beta versions of these Linux distros noticed some weird behavior. He noticed that his SSH connection was taking half a second too long.
And so he pulled the thread and traced it back to this backdoor dependency. And we were all saved because of this total accident. It's mind-blowing to me for a couple reasons. One, obviously, wow, there's literally states out there, countries that are trying to target open source now. Clearly, there's a team behind this. They probably didn't just work on this one dependency.
They were probably working on getting access to many other ones in parallel. If you just look at the time between the emails they sent to the maintainer, they were about a month between some of these emails. So they were probably working on other maintainers and trying to get access during that time. So that's really scary.
I also think it's pretty scary to see kind of the fact that it took an accident to find the attack. It makes me think, like, how many have we not caught as a community? How many have we missed if this one was caught by a total accident? It was eye opening to a lot of people and it made people realize that there really is a threat in the open source ecosystem.
And it's not because most people are bad. It's the opposite. Most people are good, but there are few bad actors out there taking advantage of the trust in the system. That's really where we come in. We're trying to give every company the tools to protect themselves from those types of attacks. And that's what we do at Socket.
Okay, friends, go to socket.dev. Security dependencies. Socket is on the front lines of securing the open source ecosystem. They're a developer-first security platform that protects your code from both vulnerable and malicious dependencies. Install the GitHub app or book a demo. Again, socket.dev. That's S-O-C-K-E-T dot dev.
So are you saying that NVIDIA and these players that are providing these expensive AI GPUs, things like this, this new tech, I suppose is probably the easier way to say it, should learn from the Apple model of this? Every 18 months, there's a recycling of tech. I'm getting the latest, greatest. The older version of it is cycling out. And I can rely on the best being available to me.
And the company like NVIDIA can lean back on this predictability of a sales cycle, essentially. So as they have CapEx and innovation, this new thing they're working so hard on has a place to go.
Yeah, you said it very well. I think to even dumb it down a little bit more, it feels more consumer. Because that's what we have a lot in our consumer life. And what you said there is like, hey, I want access to the latest technology. And then you can see, well, okay, well, how do I remove the friction of an upgrade cycle? And how do I make that beneficial for all parties?
And usually that's what SaaS is about. You have to buy packaged software once every 10 years. Just give me the latest updates all the time. I'll pay you monthly. Yeah. Let's be aligned in terms of I'll keep paying you if you keep delivering me great updates. And I think that that alignment is where the real unlock comes.
And I just believe that you can do it with hardware models and I think operating models. And most people are just desperate to say, nobody's raising their hand saying, please ship me all the computers and all the cables. I love putting them all together. Yeah. Nobody says that. They just like, I need to have the outcome.
One person says that.
You don't talk to the same nerds we talk to.
We talked about racking and stacking last time we talked and there's joy in it.
I personally enjoy it.
But not, yeah, there's some joy in the process. Businesses don't enjoy it, but people inside of businesses sometimes do.
Yeah, but doing that all over the world and in 25 markets and dealing with import duties and I'm missing the optics for this thing and this one broke and nobody loves that part. And I think that the upgrade cycle is really important because I do believe we're moving into this new architecture, right?
We were kind of going from, in my simple brain, I've been living in the world of IT that is basically client-server plus Ethernet for 25 years. String together a bunch of computers. It didn't really matter if the computers were the same type or version. we're going to run Linux, we're going to connect them with Ethernet, we're going to scale out workload is going to be hunky dory.
Now we're moving more almost back to like mainframe style integrated systems where the new architecture of computing with machine learning and AI is much more dependent upon a rack scale or a pod scale or system level architecture with the right networking, working with the right hardware, with the right software actually makes it possible.
And I think that that architecture shift going from commodity COTS computers into highly specialized integrated systems represents an opportunity to change that experience of how you consume it. Which is like, oh, you're going to upgrade this on a regular basis and it all has to work together. You don't just get to piecemeal, bits and pieces whenever you see a good deal on DRAM, you know?
Like, I think that's going to make a real difference in how people go through these cycles. And we've been used to an Intel CPU TikTok cycle that's more like six years. Like, I mean... maybe they were doing it every two, but really you didn't need to upgrade, you know, except for every four or five years. There was no true need for the, there was no true need, right?
You played the TCO on power and speed and density, but like, it wasn't like we were, we were, the Moore's law curve had slowed. We're not, we're just the beginning of that curve right now in accelerated computing.
Yeah.
So I think that's going to represent an opportunity for a more delivered model versus here, buy the kit, call your Dell rep.
I want to talk about that, but I want to put, I hate this phrase. I'm going to say it though. I want to put a pin in it just temporarily.
Just a pin in it.
Because I want to throw something out there because we departed from one part of this conversation that I want to just throw out there for both of you. And if we like the Apple model of this, you know, every X months or every X years, I'm cycling out my phone and we're all on Macintosh computers. We're all in Apple, M1, M star computers, all three of us in this conversation.
Why haven't they done that with the Mac computer? They've done it well with the iPhone. I'm not swapping out my phone, my, my computer the same way I can with my phone. if it's such a good model, are they just not eating their own dog food or drink their own champagne in terms of this model with the Mac?
Is there, is there a reason why they haven't conquered that market where I'm swapping on my Mac every two years? Cause the latest greatest is there. They're taking the old thing away and they make it an easy process. Now you can, you can go and buy the new thing, but the process that they've put in place for the phone is so turnkey. Right.
I think it's a great question because they've got a trade-in program. They'll take your old MacBook. It's like no real difference, but you're right from a conceptual standpoint. I don't think about it, and I think most consumers don't think in the same way. I think this is the power of... You know, reframing, like, why do you go to the grocery store?
And it's like, same grade formula, just a new box, right? And it's like, wow, it's purple now. It must be different. While they're telling you it's exactly the same. There must be some science spy and consumer branding and marketing where the phone was just a different model. We were used to it. being an upgrade.
We were used to like, oh, you know, but our desktops, no, we put that there and stuck there and, you know, we did it and we bought it and we owned it. I don't know. Maybe there was this because phones came through carrier. That's what I think it is. And it was just like the consumer behavior was more used to it.
Somebody can subsidize. No, there's already a monthly payment. We're already paying our phone bill.
And often you finance your phone through the carrier. So I think it was just consumer behavior was very comfortable with this. And I think the fascinating part for me in data center is we're very comfortable with this with the vast majority of our computing called cloud. We get it like that. We don't have to think about the cable that broke or this, that, that. We just get upgraded EC2 instances.
And so I think that it can translate, but you have to rebrand it in some way, or maybe you have to say it. You can't just say, hey, we'll rent you the server versus or the chip. It has to be a totally different value proposition or at least presented as like, I don't know. There must be somebody out there who's got a creative left brain that can think of a way to say this.
Because I think the emancipation for most customers is, hey, man, I'm in the banking business, not in the buying computer business all the time. Could you just really make sure I have access? And I think that's been the attractive of public cloud challenges is it's still a bell curve offering, right? It's still like a mass market multi-tenant bell curve offering.
And when you have this really differentiated technology coming out of diverse players like Nvidia, this is where enterprises make their alpha, right? And they're like, I want to adopt that now or at my pace, in my way. And I think therein lies the opportunity to bring that as a service consumptive style desire. Maybe it means operated. Maybe that's just the thing that I'm noodling on.
But I think there's an opportunity to bring that to enterprises beyond just like a lease. You know, it has to be a better experience than just financing.
Would you subscribe, Adam, to a Mac subscription? Would you subscribe?
I would certainly consider it if it was available. And if it was productized like the way the iPhone is available to everyone, I do understand that there is a subsidy there that carriers are playing. There's an underwriter that is essentially making it easier for me to consume.
I'm not sure if I would personally, but I know there's a, I would say a minority, but still very large market that pays attention to the latest capabilities of Apple's machines as a desktop.
And so I may not be personally, but I can appreciate it being available because there's people who really, every time there's the newest version of it, whatever the major advancement of it is, and it might be every two years, they're the ones buying it
But similar to the way you buy a car or a large liability that seems like an investment, which is what a machine – this computer is an investment because it makes you make money – But it's still a liability in the fact that it goes down in terms of its cost or its value to someone else.
Yeah, it's depreciating.
So you may buy it for like three or four grand, which is a pretty good average for most decent MacBook Pros. Let's just say four grand, right? I buy it for four grand today at the latest gen price, and I want to recycle it two years from now. Apple doesn't have a program where I can just go and buy the new thing.
They do have the trading program, but it's not predictable in terms of what I get back for today's investment. And I think that's what they've done well with the iPhone is that I know pretty much, well, if I'm on latest tech, I probably get 800 bucks-ish for my old tech to cycle out for the new tech. There's some rebate program out there.
Or just these programs you can pay for and just you never own it and you're always happy. Like the New World Order said, like the World Economic Forum said, you will own nothing and be happy about it. And I kind of feel like you're saying that with this stuff too. Like this Apple model of this is good. And I kind of hate that.
I kind of like it as a consumer and the fact that I get access and this promised, predictable, priced-in access to latest tech. But then the other part of me is like, man, I'm renting everything I own. Everything I have has a subscription attached to it in some way, shape, or form. So every new dependency in my life...
whether it's food, a bed, and I won't name a name, or something else is reaching into my pocket with some subscription. And it's kind of essential to my life, personally or professionally.
Well, I think therein lies the opportunity. I think the most interesting part about a subscription, in theory, I mean, I know why businesses love it. They're like, dude, I get $10 a month from you or $20 or whatever, right? I sold you a bed, now I get $20 a month. It's amazing, right? I'm not sure there's that much value being given out of it for the app or the thing or whatever.
But what it does in theory and what I've always loved the concept with chip companies is that it aligns your interests. to some degree, right? If you're in the sales business and you're just like, here, buy this stuff, buy the product.
Then when I have new product, I'm actually like the business model of Intel was specifically built around ticking and talking to you so that you would have to buy new technology on a regular basis. But what if I'm totally fine with the old technology? I want to keep using it. What if you could just do updates to the chip you made three years ago that made my
Linux kernel optimized and I get to keep running, I'd be happy. Or maybe just lower my price a little bit. That'd be great. We would be aligned in terms of that. Right now, that's not the case. They got to sell you new stuff.
And I believe beyond that doesn't align the overall economic interest between the customer and the vendor, that it actually doesn't align a whole bunch of the sustainability issues that come out of that. The amount of waste that we produce in
creating, making, shipping, delivering effectively a bunch of sheet metal and copper so that we can sell you an upgraded CPU is asinine, which there's not repeatability. And it's kind of funny because people will argue over, I'm not paying, and this is a CPU line, maybe it's a little different than GPU, but CPU would be arguing over
$10 on an $1,100 CPU while we're throwing out $1,200 of power supplies, which is like forest through the trees here. Why don't we figure out a way to have reusable power supplies and reusable all the things around it and slot in the new upgradable CPU thing, blade or whatever, and everybody goes home happier? But those incentives are not aligned.
That's one thing you notice being in a data center company is just the amount of old crap that has to get thrown out. There's a lot of work in this industry to get stuff into data centers, and there's very little about what to do with it after.
You used a word earlier, I believe it was like circulatory or circular.
Circular, yeah.
Circular, circular. There you go, simple. And I think you mentioned what Kyle Wiens talked about with me on a podcast. Do you know Kyle Wiens by any chance? Does his name ring a bell?
I don't.
He is the person who started iFixit. And they really started out as a wiki or kind of like everybody can share their own how to fix something because of all this right to repair situation.
Right to repair, yeah.
That kind of touches on this circular thing you're talking about in parts pairing, which you talked about tangentially earlier. I'm just curious your thoughts on this. A power supply having to be thrown out as a result of a new CPU, which is a ton of waste.
You've got things, and I put the pin in with the one thing, I think to bring it back might be to discuss your thoughts on Oxide and what they're doing to sort of give you a full integrated, as best they can guarantee of all the stack working together in concert versus things not working software-wise or hardware-wise.
So this idea of circular, this idea of parts pairing, and this idea of a full stack thing you buy versus parts in the middle.
Yeah. Well, I mean, the circularity aspect, I think, is critical. Right now, there's not a big incentive, actually, to make it better, to fix it.
What does it mean to be circular or circularity? What does that mean to you?
Well, I think what it would mean, I think Apple, let's use the example there, when they take your old iPhone, they normally don't sell it to some random company. They do for a certain percentage. The vast majority, from my understanding, I've talked to the techs. I'm like, where do these iPhones go? They're like, we take it apart. We make new iPhones with it.
And they're building iPhones so that they can take them back because they know they're getting them back. Obviously, that has a great side effect that they remove a bunch of the secondary market and they control that. But they get to take all the materials and reuse them, which probably saves them a whole lot of money in their supply chain, as well as meeting some of their sustainability goals.
It's expensive to dig up all that stuff and process it and do all those things. And so I would see circular and computers would be where there would be a benefit to, whether it was, say, Intel on the chip side or Dell on the computer side, wanting to have their thing back. I think Cisco does it pretty well with its switches. They control their secondary market really, really well.
And my understanding is over 90% of their products are disassembled and remade. They really focus on a circular supply chain, which is good for the earth and good for economics, which is awesome.
It's better than like sitting on a shelf, collecting dust, having no use or going into a literal landfill where someone doesn't dispose of it properly. Maybe it has a battery in there, blows up a garbage truck. We've heard stories like that, you know, where basically if it doesn't have a removable battery, you're just renting the thing anyways. Yeah.
Like I'm upset with my AirPods right now because they don't last as long and there's nothing I can do about it. All I can do is buy another set, you know, for the new retail price, you know, whatever that is. And the current version I have is basically worthless, but there you go.
Yeah, I mean, so I think the trash side of this is really, we can't ignore it. I mean, this is hundreds of thousands or millions of computers a month coming out of data centers. Where are those going? They are generally not recycled.
So where do they go?
I mean, they go to landfills and like, you know, scrap and in not great ways, in very destructive ways, which is really unfortunate considering my understanding is like 70% or so of the carbon impact. I heard this from the, there was a company several years ago built around taking OCP gear and giving it a second life.
Dean Nelson has done a bunch of this within the infrastructure masons on just taking hyperscale equipment and trying to give it two or three more years worth of life after that upgrade cycle. But my understanding was something like 70% of the carbon impact of a computer over its entire life is making the computer 20% is running it the whole time.
And 10% is what they can recycle from it is the energy impact. And so like that to me says, wow, the best thing you could possibly do is not make another computer or at least 80% of it. You don't have to make new parts on. Right. And so, yeah, I think that there's a, there's an alignment issue and opportunity where, uh,
The industry would want that to be more circular, where in data centers, we would have collection points where Intel, NVIDIA, or Dell or HP wanted their things back so that they could upgrade them and change the parts and do the things. Or even better, you would design the data center differently so that it was upgradable by default.
It would make multi-generation systems instead of buy a new kit and build a new kit. What Brian's doing at Oxide is super cool. You like that? Yeah, I mean, I love the rack scale concept there. Super timely and applicable. And he and his team have done just such a great job at building open, right?
You know, like literally building a whole new computer versus just skipping a whole bunch of the hard stuff around firmware and bootloaders and related, which is awesome. Awesome. And I presume makes them super confident and running awesome computers for people. And I know that they've been able to offer a really high quality experience with these oxide racks because of that.
But I think there's also this opportunity for them to maybe think, and I don't know because I've not bought an oxide rack, but I'd love to. I'd love to have a reason to, which is to change the economic model too and allow people to subscribe to them, which I think would be super cool and really give them that competitive edge.
I'm sure they're thinking about it because Brian's super smart and thinks ahead about 15 steps. Yeah.
make the computer make the rack better and then now that you have disruptively scalable and efficient economics on making computers you know figure out how to run them better for people you can go listen to this podcast at full length brian shares the story with me but the catalyst to do oxide officially was born out of angst of support from dell that they had an extreme issue that i'm going to paraphrase because the podcast is better and brian tells the story better
But there was a major issue that they were just like basically band-aiding over and they couldn't explain. And it was because of interoperability between the hardware they chose to put within their machines and their systems, whether it was software or hardware. And Brian was smart enough because he's back in the ZFS and D-Trace days like Brian's a smart guy.
He's not just the entrepreneur in the moment working at Sun trying to solve this problem. it was a whole different thing that they were dealing with. And he's like, yeah, I'm reading between this line. And he had a venture capitalist who was like, you know, I will literally fund anything. This is in quotes. I will literally fund anything you put in front of me.
And so he was like, okay, maybe this is the time I need to go big. And so that was the catalyst. This angst of this interoperability between the stack that is like, here's one server, here's the compute, here's the storage, and none of them are made to work necessarily in concert. One blames the other. Dell hides this thing and his story is his story. So take that with what you will.
And I'm sure it's fully true because Brian's a truth teller.
I mean, I know this well because I think the entire value proposition of Packet was built around this problem, which was how to normalize this super fragmented, disjointed system integration world and make it just work. Like how to turn on a bare metal computer a million times and get a success ratio of multiple nines, right? And the answer was...
mainly not in physical hardware issues the answer was mainly in that bios doesn't work with that uefi with that nick with the firmware on the nick that you don't even know how to flash right with that hard drive on the nvme something or rather on something you know it's like it was a A little matrix of compatibility issues that you had to do. And so we solved it. I don't know, solved it.
We worked on it and improved it through almost chaos style, like run thousands and thousands of boots and reinstalls and processes a day to find out these bugs. Right. But yeah, it's super frustrating. And you would get into like, oh, well, it's the Mellanox card. It's like, okay, well, it's Mellanox Connect X5. They're like, no, no, no. It's a Dell branded Mellanox. That's totally different.
Like, wait, it's still Connect X5. They're like, no, no, no, no. It's a different one. Like, oh. So. Yeah. It's just a computer. Well, I tell my wife, I was like, she's like, what do you do every day? I was like, just making computers turn on and off. That's all we do.
That's all we do.
That's all we do.
What's up, friends? I'm here with a new friend of ours over at Assembly AI, founder and CEO Dylan Fox. Dylan, tell me about Universal One. This is the newest, most powerful speech AI model to date. You released this recently. Tell me more.
So Universal One is our flagship industry leading model for speech to text and various other speech understanding tasks. So it's about a year long effort that really is the culmination of like the years that we've spent building infrastructure and tooling at assembly to even train large scale speech AI models.
It was trained on about 12 and a half million hours of voice data, multilingual, super wide range of domains and sources of audio data. So it's super robust model.
We're seeing developers use it for extremely high accuracy, low cost, super fast speech to text and speech understanding tasks within their products, within automations, within workflows that they're building at their companies or within their products.
Very cool. So Dylan, one thing I love is this playground you have. You can go there, assemblyai.com slash playground, and you can just play around with all the things that is assembly. Is this the recommended path? Is this the try before you buy experience? people do?
Yeah, so our Playground is a GUI experience over the API that's free. You can just go to it on our website, assemblyai.com slash Playground. You drop in an audio file, you can talk to the Playground. And it's a way to, in a no-code environment, interact with our models, interact with our API to see what our models and what our API can do without having to write any code.
Then once you see what the models can do and you're ready to start building with the API, you can quickly transition to the API docs. Start writing code, start integrating our SDKs into your code to start leveraging our models and all our tech via our SDKs instead.
Okay. Constantly updated speech AI models at your fingertips. Well, at your API fingertips, that is. A good next step is to go to their playground. You can test out their models for free right there in the browser. Or you can get started with a $50 credit at assemblyai.com slash practical AI. Again, that's assemblyai.com slash practical AI. And also by our friends over at Wix.
I've got just 30 seconds to tell you about Wix Studio, the web platform for freelancers, agencies, and enterprises. So here are a few things you can do in 30 seconds or less on Studio. Number one, integrate experience. Extend and write custom scripts in a VS Code based IDE. Two, leverage zero setup dev, test, and production environments. Three, ship faster with an AI code assistant.
And four, work with Wix headless APIs on any tech stack. Wix Studio is for devs who build websites, sell apps, go headless, or manage clients. Well, my time is up, but the list keeps going on. Step into Wix Studio and see for yourself. Go to wix.com slash studio. Once again, wix.com slash studio. Do you have a crystal ball?
Not handy.
No? Could you assume you do? I'm going to ask you to predict the future to best you can.
We can go into a pretend world?
Well, I think that you've got a lens that Jared and I can appreciate and conversate with you about and dig in. But I've got to imagine between how we can cycle out at the data center level, this new age, this innovative tech coming out,
these models, this thought on the rack stack kind of scenario where you get the oxide model, which to my knowledge, they're really the only player in this town, really. So if their idea is successful, then they have a very large head start and a deeply entrenched moat. There's a lot that it would take to bow them. Where is this going? Where is this cycling out GPUs, CPUs, chip ores, hardware?
Where's it going?
Yeah.
What is the future?
I don't have my crystal ball, but I do attempt to do what I refer to as the 10-year thing, which is I just try and think out 10 years, which is super scary because it's like 2035 almost. that sounds really scary to me but you know you just try and like stretch your brain and be like all right well 10 years ago was 2014 15 okay so you know kubernetes didn't exist and
Software was hard and VMware was ruling the world. And, you know, it was pretty different. GPUs, that was just for gamers. And, you know, data centers were a bad business, right? And like, you know, all this stuff. And so I think 10 years from now, probably pretty different. So how do I stretch my brain? to think about that and imagine.
And frankly, that's one of the things I love about not working right now, kind of taking some time off because you have nothing else to fill your head. So you do a lot of daydreaming and thinking about whatever. And so as it pertains to infrastructure, I daydream about a few things.
One is that I daydream that our digital infrastructure is like a core pillar, if not the catalyst for our energy transition. Instead of a consumer of energy and resources, it's a catalyst and transitionary, almost like policy for that.
So I'd really love, I think the amount of money, dollars, interest, and the type of customers, creditworthy customers who consume the vast majority of digital infrastructure means we have this massive opportunity. I think we've started to see it like what Microsoft did with Three Mile Island a couple of weeks ago with Constellation Energy, what Google's doing with SMRs, the amount of
what Evol's doing over at ECL on hydrogen-based data centers, these can be baseline load for extremely expensive long-term energy transition investments that we can't make otherwise. They'll just take so long because we won't have a clear customer. We won't have a clear demand driver. It won't be as urgent. And I love the idea of this phase of
power hungry you know intensive energy driven applications driving positive change in our society versus just adding to the problem of using all the power using all the grids using all the stuff right driving up the price so that's one thing i daydream about because i certainly would like to live in that and be part of that versus be like
the data centers that sucked up all the energy so that we couldn't transition other parts of our economy to carbon neutral or zero carbon. So that's one. The other is that I think we'll see a radically different OEM. I think OEMs are Sorry, I love the Dells and HPs and Ciscos. I think the OEM model has to change. That is going to be so radically different.
Where technology was kind of brokered through folks, which was really bell curve driven. It was how to get x86 PCs and Windows laptops to a million or two million or three million businesses. That's just a different model now. And I think there's only a few thousand really... intense, large scale infrastructure consumers of which the top 10 consume like 60 or 70% of the market already.
So it's like super concentrated. And so I think the OEM model is going to like change dramatically. I think because of that, the chip model will change. And I would, I believe this is again, we're in imaginary Zach mode here. I think the chip companies are going to be in direct competition with their cloud customers. It's going to get knives out.
That makes sense to me.
And because of that, we're going to basically see these as national interests. It's geopolitics. It's not just business anymore. And we're going to figure out how to protect that or break things up or more regulation and more government involvement is going to happen in this part of the industry. And because of that, we're going to see this kind of duopoly kind of competition between
chip companies and cloud companies because are they really any different right now um right who's the second largest producer of gpus is google that's just for them though um and so you know graviton and their cpu and their nick business at aws these are very very big initiatives so i think chip companies and oem is going to change dramatically my hope is that that's more aligned to a
kind of responsible long-term resource use. You can see sustainability is driving heavy in my thinking, but also hopefully allows better access to tech. Like that's one thing I miss.
Like when I got in the internets in 2001 after Juilliard, you could just walk to your, walk, you could just kind of go over to your neighborhood ISP and get a quarter rack of colo and a port on a router and like start mucking around on the internet. And that was cutting edge technology. You were using the same stuff that the big guys were, you know, just like whatever.
That is just not possible now. Look at this stuff that like open AI is doing. You just can't plunk down a billion dollars to just be in the game, just to start, just to learn, just to play. And I think that that would be really awesome as we maybe see a shift in the chips and the OEMs and the clouds, there becomes better access. Because I think we need the innovation.
I think we need thousands of new companies that don't do it the old way. And I'm very worried of... a megalopoly or whatever you want to call it of like the mega seven, do all the tech and you shall, what did you say? And you shall like it.
You will own nothing and you will like it.
You will own nothing and you will, you will, you will like it. I think that that's just not going to be good. You know, I think we're in need to see more messy innovation. That's going to be more oxides, but look how hard that was a required, like that had a huge amount of like the VC saying, I will do anything plus the right Brian at the right time to create what one, uh,
new computer company in the last 20 years, probably. That's not enough.
Yeah. Well, on that topic, what do you think about the Department of Justice's upcoming potentially breaking up of Google? I mean, it seems like it's plausible.
I don't know what's going to happen or not, but... I'm not going to get into election politics and how that plays in, but it doesn't seem like either candidate is a huge fan of megatech. So I think you're right. I think that there's... just this run that we've had of, you know, kind of advertising-based monopoly.
I think I got a quote or I read an article, I could be wrong, about one of the initial Google trials that led up to this. And they asked the jury, like, you know, which search engine do you use or would you use if not Google? And I think, like, seven of the jurors like said, there's another search engine.
Right.
You know, like there's another, I didn't know there was one. Right. And that was just like, well, that's probably not going to fly.
So. Right.
I look forward to it. I'm actually really curious if, you know, and, what, you know, privacy oriented models can work. And certainly as we see AI become a little bit more, you know, search was, got creepy over the past couple of years is everything followed you. And like the results are like, I mean, I don't know how many sponsored results you see on your Google search, but it's like 20.
Before it gets to anything, you know, that isn't an ad these days. So I think the younger generation, I've talked to my kids about it. They just don't believe any of it. They're like, that's not real. So I think that that plus AI where it starts ingraining itself with personalized responses and data is going to freak people out.
And they're going to, you know, both governments and people will switch towards more privacy oriented models, which I'm a fan of. So I think like, yeah. Government, like the internet, is a big part of our lives, and that means policy's coming.
Yeah, 100%. I agree with that. And I think people already are freaked out, so much so that so many people are convinced that Instagram is listening to them, like audibly turning on a microphone. And I've argued against that. I don't think that's the case. But I've stopped arguing it because it's so well-targeted. Right.
That it kind of feels that way.
So quick that I'm kind of like, you know what? I'm starting to become a believer. Maybe they really are. Cause it's so personal and timely. Like the timeliness is what gets you. It's like right after I talked about it, I didn't have to search and I'll tell people we have, but you're connected to this other person and they were probably searching it because you talked to them about it. You know?
So I was trying to help understand like how the, all the tracking works and stuff. But at the end of the day, it's like, it's gotten very targeted to where I'm like, Yeah, maybe they just turned the microphone on after all.
I don't know.
I can't explain this.
I'm going to get your tinfoil hat for you, buddy.
Yeah, exactly.
But yeah, I agree. That's a little creepy. And so I guess if I were to zoom it out, my prediction is the internet is going to be a lot less flat. It's going to get a lot more siloed because there's going to be more and more regulation that keeps data within places that... makes things act different ways. It starts enacting different rules and policies.
Like our little, do you approve on the cookie tracking, allow all annoyance is the beginning of our upcoming hell. And I hope, I find it with myself, I'd rather pay. I'd rather pay 10 bucks to have the thing that I don't need to. A friend of mine is developing a mapping thing called Over a story, I think.
Basically, it's paid mapping because he's so upset about being targeted while he's trying to get a route. It's trying to give you ads on where to stop. He's like, I don't want that. I just want to have... And so maybe we're on to that. Maybe a couple more subscriptions for Adam that come with no ads.
The one that pissed me off the other day is I think I pay for subscription for like every single different – because my kids watch soccer and the other one watches NFL. So you end up paying for every streaming thing whether you want to or not. Totally. And one of them, I don't know, it was Paramount or something. Like we pay for it. And then it gives us ads when we hit the pause button.
And I was like, oh, that's a step too far, man. Like I hit the pause button and then it was like, well, you're Boz. Let me just advertise to you. And I was like, oh. I pay you already. Stop.
They don't care. Squeeze every dollar out if possible. So this antitrust, more regulation, potentially more governmental control, like pair that with your –
future look at these energy investments, because it seems like, I mean, if Google and Microsoft and these companies are investing in these new or even old forms of clean energy, nuclear, et cetera, I mean, that stuff is going to get also like, do we want that kind of a megalopoly as well as like these seven companies own all the energy?
Well, I think that they might not own it, but they're going to cause it to occur as customers. I think they're very different businesses. Some are high margin, high growth. The other ones are used to being regulated, non-growth businesses. I think the big shift we're having right now is that energy is gone from basically we reach peak load.
in the 90s and it's been slowly decreasing or like growing very minimal to it's actually growing by a few percentage points a year which is crazy and we're not used to it so you know we might see some energy businesses go private and turn into like growth businesses basically right i mean you can see constellation which is the nuclear operator doing that if they figure out that they can build
10 or 20 more new nuclear plants, which hasn't happened since the 60s, it'd be a growth business. So I think we'll probably just see a lot more. Well, we've already seen it, like even in Ashburn, Virginia, which is the largest data center market in the world.
Just the local planning board saying, hey, I'm not going to approve this building permit because you're going to bring in 500 diesel generators or you're going to use our water. You're going to use 50 million gallons of water a month. We don't want that, actually. That's a municipal thing that we need. We want no diesel generators next to our school and we need the water for our other stuff.
I think these things are mixing only because it's just become a bigger part of our lives. And it's a critical part of a lot. It's not just a nice to have. It's a must have, I think, in most ways. So that just means more involvement between public policy and private companies.
And maybe that results in just straight up regulation, or maybe that just results in better partnership or, you know, that being on the talking points of our next podcast. You know, local election is how we're going to support or not support these types of initiatives.
And I think like that's become the lingua franca of digital infrastructure is like way more available than it was a couple of years ago, where even though I've been in this industry for 25 years, my mom didn't know what a data center was. And now she knows all about data centers and chips. And it's not for me.
How did she learn it?
I mean, the news. Chipsack, this is delayed, coffee conversation, what's AI? It's all in the ether now, so it's much more aware. And so now people are like, oh, data centers. I mean, you hear data centers like that word probably a thousand times more than you did three years ago or five years ago.
in the news and publications like every single you know article comes up in some business thing there's something related to this and so i think it's just becoming more part of the public consciousness and because of that you'll see more stuff whether that's rules or you know, yes, I want it or no, I don't want it. So it's coming.
I think we've actually done pretty well for critical infrastructure, you know, data centers and the relates I've been relatively unregulated. Now it's just coming. It's like a big, big part of the industry, big part of the economy. So it's coming.
So you're not working right now. You're daydreaming. You're playing instruments.
Playing in the gigs for free.
Right, right, right. You know what I mean. Traditional work. Do you know what's next? Are you investing? Are you starting something new? Don't go 10 years. Go like two years and then go personal in your life. What's it going to look like, you know?
Yeah, so for those who know me well, no, I don't sit still. It's not a great quality I have. So Jacob and I, my brother, we've done a bunch of investments over the past 10 years through a little family office we called Humans of the Internet. We like to back humans who work around and with the Internet, as well as a couple of specific social causes.
And so we've invested in 50 companies, which has been great. So we spent a lot of time supporting the founders. They're all very, very early. Some of them have grown to be not early, but we like to support the early ones because that's where we feel we have some experience and some value to add.
And then, yeah, I'm just swimming around working on a couple of things, both in open source, some standards around liquid cooling through the Linux Foundation, as well as exploring some open source software ideas around service provider stuff that I wish has existed. So I've been working on some of that, mainly as hobbyist stuff at this point. But, yeah, we'll see.
I love to build, so I'll definitely build something. Just got to figure out where. My answer is almost always I say yes to everything. I'm like, that looks interesting. I should work on that. And before you know it, I'm really busy.
Tell me more about this liquid cooling stuff.
Yeah, so we have a... There's an organization, part of Linux Foundation, called the SSIA, where... we created a – think of it like a USB standard for single and two-phase liquid cooling between server manufacturers and chip guys like OEMs and data center operators. So that way we could have the right size coupler and start to things like color-coded.
So that way you had a computer, you slid it into a rack, and you would be able to plug – liquid loop into it. And the idea here was that if we could standardize that interface, as well as some of the more complicated stuff around fluid, like flow rates and pressures, that the industry could more holistically adopt liquid cooling, which is dramatically more efficient in terms of removing heat and
and using less energy in data centers. And it seems almost obvious, except for computer manufacturers release computers every one to two years, data centers change every 20 to 30 years. They're on very different timeframes, so there needed to be more of an open standard.
So we've been working on that, working along with the OCP crew, as well as different organizations and the Green Software Foundation.
which is also in Linux Foundation, to just try and figure out how this standard, which is published under what's called the JDF, Joint Development Foundation, a little bit different than open source software, which you can actually put in GitHub and put like a BSD license or Apache license on. You can't do that for just words on drawings on a spec.
And so we worked with Linux Foundation on the JDF, which has been awesome. And, you know, I'm not, I wish we were more successful, meaning I wish it was more common, but I can see, yeah, like a ton of the work we've done has been adopted and consumed in practice or in some sort of principle, like the new NVIDIA stuff looks basically like what we did, which is great.
They're taking inspiration from that. OCP's new liquid cooling side for its hyperscale customers looks very similar. So I think the supply chain is starting to really gel around this. My hope is that Cisco, Dell, HPE, and the longer tail of the data center industry can kind of align.
towards a standard versus trying to think that being proprietary differentiates themselves which it doesn't it just makes it harder for everybody so i'm hopeful that we'll get there but i'm not you know doesn't move fast sure community-based organizations where everybody gets to have a say they move slow you know but that's okay
Is there a possibility of, I think back to the idea of habit stacking, Jerry, where the habit might be liquid coal, a server rack in a data center, and the need for clean water. Is there a way to combine interests? Now that you're sort of, let's face it, you're at a position in life where you can pause, so you must be financially well off enough to pause.
Does part of your daydreaming think about... not just capitalistic innovation opportunities, but legacy things. You've got kids. You obviously seem like a very caring person, as I know you. Is there room for combining these efforts to power a data center and clean water? Buy a waterfall. I don't know. Is there a combination of things? Not buy, literally B-Y, but B-Y. Buy a waterfall.
Put a data center near there.
Located adjacent to.
Yeah, exactly. Something where... I don't know. Is that a thing? I'm here in Texas and... Lake Travis is not a lake anymore. It's dirt and like little islands out there. There's like a little bit of water in there. I ride at this place. I ride mountain bikes and it's by slaughter Creek. Slaughter Creek is not a Creek anymore.
And it's, you know, it's one thing that I don't think it's like climate change kind of thing, but I think it's just really hot and, And there's a lot of people in this area and the aquifer gets less replenished. I mean, is there ways we can level up humanity in these ways and liquid cool, but also find ways to keep the water, produce the water, cloud sea? Do you think about things like that big?
Absolutely. I mean, that's why I wouldn't build a data center right now. It's because where is my time and energy best spent when I can – think outside the box a little bit, right?
Move about the cabin, as my brother likes to say, which has been super useful because I'm not stuck within an existing organization that has to answer to its shareholders this quarter and do this and that, but I still have touch points and trust within it. So areas that I've been focused on there, and I totally agree with you, Adam, that we can take these areas.
I've always believed in building businesses was about building a bigger tent, right? Right. It's like, how do you create more opportunity? That's good for society. It's not just about how do you create something that's valuable? How do you create opportunity that people can?
I mean, the biggest joy I've had out of building several companies has not been the money I've made, although that hasn't hurt. It's been that, you know, a thousand people said, wow, I got to have careers that started here. Right. Or like, I grew up my family because I worked at this company. That is awesome to me. So I like the bigger tent policy of creation.
But where I was referring to earlier related to aligning civic infrastructure, like energy, with the... Awesome capitalistic movement. If you can harness that stallion called hundreds of billions of dollars going into digital infrastructure right now and use it towards the right thing, well, shoot, that's amazing. That can benefit society in way bigger ways than...
I got chat GPT or the AI can answer my thingamajiggers or target my ads better, right? It can say, wow, that helped decarbonize a whole bunch of our grid. Or on the water side, I'm invested in a hydrogen-based data center business, which I believe to be a foundation. I just got back from the Atacama Desert in Chile, which has this project too.
It's like has the highest solar radiation percentage annually of any place on earth. Why? Because it's so high, it's so dry, and it's so sunny. And it's only a couple hundred miles from the Pacific Ocean. They have this plan to create basically the Saudi Arabia of hydrogen there with this like huge amount of energy they can produce that they can turn into a portable fuel.
That when you burn hydrogen, what do you create? Water. And it's pretty amazing. We built a test or not we, but you've all in the ECL team built the test site over in Mountain View. They have a one megawatt hydrogen-based data center. It produces hundreds of gallons a day of water. That's the byproduct, right?
Versus a normal data center, which uses thousands of gallons of water a day through evaporative cooling. Sometimes millions. I think there is really an opportunity to take existing technology, whether those are like battery storage technologies, for example, been exploring a heat-based, like a thermal long storage solution. These are awesome opportunities.
Your data centers create heat and also use a bunch of energy, but they can store energy, which is really important. So I think that there's some ways that we can start thinking, and my challenge is how to activate that between, The let's call it sub 50 people who build and operate all the data centers in the world, the capital that flows around it.
And some of these weird innovation things that are so hard for that group to see because it's not part of their big scaled organization and how they've always done it. Right. Like there's almost what was the thing? I was discussing somebody who was building a new data center in a far off place. And I was like, why are you building it with all these standard rows and no storage space?
And I was like, you realize nobody's ever going to go there. It's in South Africa. It's like 16 hours on a plane ride from Paris. You know, like nobody's going to the data center. Why are you not building it for robots? And like, you know, five lights out people who are going to rack and stack everything. And they're like, well, we always build it this way. Why would we change that?
We built like hundreds of data centers like this. Why would we change the design just for South Africa? And I think like Theron represents where I'm most interested in intercepting is, could I take some of my relationships and experience and, gather some of these new potential opportunities and help normalize them. So that way we can build a hundred new hydrogen-based data centers.
I know Yuval's building a gigawatt facility in Texas. You're going to produce a lot of water. The challenge is right now he's going to use kind of gray and blue hydrogen. But he'll probably be the largest customer for some of the new hydrogen plants coming out of the solar and wind energy in Texas, which is awesome.
The oil and gas industry in Texas can start to produce hydrogen from all the excess renewables that sit. You know, in the western Texas side and like that's awesome because there's going to be a demand driver called a big ass data center who says, give me your hydrogen and I'll pay you for it versus like a government subsidy or some sort of like other thing that would take a long time.
So yeah, that's where I'm kind of swimming, happy to help, really anxious to be able to make a positive impact there. And I think that this moment right now that we're going through with a lot of change and opportunity in the digital infrastructure space represents like a unique spot. blessed to be sitting here. Yeah. Able to do some of that.
Very cool. Yeah. Cool talking through it too. I think the, when you mentioned Bernie hydrogen to create water, I thought about Mark Watney in the movie, the Martian, but also the book, the Martian, he blew himself up cause he had his ratios wrong. And when he lit it on fire, it was too much of one and not enough of the other. And it created a bomb basically.
Like it was just a small bomb, but it really messed things up.
But no, no H bombs, please.
Yeah. Yeah.
Absolutely.
Yeah. Just, just the energy of the water.
Yeah. That's a good formula.
Well, even like to zoom out a little bit, like there was parts of the conversation where I almost interjected about this capitalistic push of this subscription model, this idea that the way they can fund this innovation, which I get the innovation needs to be there, Intel and others, but they can only do that innovation if they can sustain themselves through some version of subscription. Yeah.
But then you peel back the layer that it seems like the draw to the end, not the means, is this idea of creating enough capital to push the thing forward. And then they're just in the cycle, like you had just said, this is an example of creating a new data center. This is how we've done it. This is how we do it. And so you almost kind of get this angst against capitalism, right?
It's the driver of it. And I think you're the kind of person who aligns the capitalistic thoughts and innovation with the human necessary needs, not just to make a successful product. or project or a thing or an acquisition, but something that levels up humanity.
And I just, I just think that's a, it's an interesting thought experiment is like, you know, how can we align capitalism with humanism so that we are still here and not just to spend our money, but also to enjoy the lives we've been given.
Amen, man.
Yeah.
You said it very well. I came into working on the internet in a very, I'm going to call it altruistic time, the mid-90s, where it was the free software movement and also just connecting people.
And I was moved, actually, in the early 2000s, I went to United Nations IT, I don't know if it was ITF or something, but it was like the telecom side of the United Nations, and they hold like a meeting at ITF. At NYU or Columbia. And I went there and I learned that that was the only group within the United Nations where every country on earth participates.
And I was like, so we can talk to each other. And I just love that about the internet as well, which is this kind of like communication and is so powerful and being able to connect people. And I think we're just at one of those moments where we get to say, okay, what was the phrase somebody said? The brightest minds of our generation are working hard every day to make people click on faster.
And like, that's not very inspiring.
I so feel that, you know?
So how can the brightest minds work towards things that improve and benefit society? I think that's a better challenge. I don't think it's reasonable to say, Hey, you shouldn't use the thing. I'm not a Luddite. Right. So I'm like, instead, like, Hey, how can we do the thing in a way that's better? Right. For, for all of us. So.
We can dream. Yeah, maybe. We can dream, right? What else, Jared? Anything left for this, friends?
Do we run out of tape?
No, we got more tape. I got more tape for Plus Plus. We have a Plus Plus zone to our podcast. We're speaking about zones in the book. Yes. We have Plus Plus zone, which is... It's our performance zone. Yeah, there you go. Folks pay us... It's our members-only zone. They pay us to have a special feed that does not have ads. It's our subscription zone.
Oh, very cool.
And then some of those people actually want a feed of ads, so I'm just really...
confused yeah i don't get it are they like awesome super bowl style ads are they like you know i don't want to toot my own horn but we do a pretty good job of producing ads okay and it's not the read it's not my radio voice it's not it's our ability to share the story of the brands who want to connect with their audience and tell that story and try to do our best to do that that's probably the easiest way i can say it there we go
I think they're enjoyable. I like producing them. They're a lot of fun. And they're not just like, and this is brought to you by so-and-so. Go buy their thing at so-and-so.com. It's not like that. It's far more like, Zach, why did you do this? Why should our listeners care? And you tell them. And then we conversate about it. And it's like a segment of a podcast as an ad.
There we go.
How cool, right?
Well, I'm a big consumer of podcasts, so thank you for doing that. And now I need to know about the plus plus.
Oh, man. We'll take you into the zone here in just a second, if you don't mind.
Yeah, you're going to be in the zone. You're going to know all about the plus plus. All right. I guess we'll say goodbye. Bye, friends.
Thank you. See you later.
Party on, Changelog++ people. We've got an additional 10-plus minutes for you right after this outro. For everyone else, thanks so much for hanging with us. We love Changelog and Friends, and we hope you do too. Don't forget, we will be at All Things Open in a little over two weeks' time. Join us in Raleigh, North Carolina. We'll be hanging out in the hallway track.
hosting a panel on open source AI and more. If you're going or even just considering it, hop in Zulip and chat us up. You can do that by joining the totally free, totally rad changelog developer community at changelog.com slash community. Thanks once again to our partners at fly.io, to our beat freak, the mysterious Breakmaster Cylinder, and to our longtime sponsors at Sentry. We love Sentry.
You might, too. Use code CHANGELOG. Save $100 on the team plan. Boom. Next week on The Change Log, news on Monday, Shea Bannon, CTO of Elastic, on Wednesday, and our old friend Quincy Larson from Free Code Camp on Friday. Have a great weekend. Share The Change Log with your friends who might dig it. And let's talk again real soon. Hello, Plus Plus people. Welcome back. Zach, welcome to Plus Plus.
This is where we can tell us what you really think. Do more of the same, basically. But wait. Yeah, exactly.
So, twist ending. While we were talking about the iPhone and the ubiquity of... It's better.