This episode features an interview between Matt Trifiro and Rex St. John, founder of Taroko Technology. Rex has spent the last decade building developer and innovator ecosystems at Intel, Arm, and NVIDIA. He recently resigned to work full time in Web3 to organize communities and developer programs. While at NVIDIA, Rex built the global software ecosystem for NVIDIA Jetson. At Arm, he built global developer relations programs for IoT and Edge. While working for Intel, Rex was the lead developer on many projects and built the Global developer ecosystem and GTM for Edge AI, Robotics, Drones, IoT, Industrial, and Embedded Systems. On his YouTube channel, Rex shares knowledge, analysis and insights about the future of hardware, software and crypto technology. He received his education from the Isenberg School of Management, UMass Amherst and is on the advisory boards for Silicon Valley Robotics and Dronecode.
This episode features an interview between Matt Trifiro and Rex St. John, founder of Taroko Technology. Rex has spent the last decade building developer and innovator ecosystems at Intel, Arm, and NVIDIA. He recently resigned to work full time in Web3 to organize communities and developer programs. While at NVIDIA, Rex built the global software ecosystem for NVIDIA Jetson. At Arm, he built global developer relations programs for IoT and Edge. While working for Intel, Rex was the lead developer on many projects and built the Global developer ecosystem and GTM for Edge AI, Robotics, Drones, IoT, Industrial, and Embedded Systems. On his YouTube channel, Rex shares knowledge, analysis and insights about the future of hardware, software and crypto technology. He received his education from the Isenberg School of Management, UMass Amherst and is on the advisory boards for Silicon Valley Robotics and Dronecode.
In this episode, Rex tells us how he got his start in video game technology and then decided to learn to program and code on his own. He turned that into a career of coding, software development consulting, and technical evangelism. Rex provides his views on the evolution and adoption of Edge computing and IoT, including how kubernetes is helping to drive transformation. He also delves into the future of what he calls the ‘Machine Economy’ - where hardware, cryptocurrency, blockchain, and the metaverse may work in tandem to ensure continued efficiencies and support for business operations and security. Rex also raises the importance of improving the environmental impact of cryptocurrency and blockchain, and advancements in democracy, social engagement, and community.
---------
“I do think people get tired of hearing about IOT and the conversation has to change. And then ultimately, it's driven by the workloads and the profitability. If people aren't investing, then the innovations are going to move slower. And then if the workloads aren't there, then it's not a justified innovation. So, I think right now it feels very much like edge computing is on the rise.”
“Shifting everything from manual allocation to the biggest thing I think is just shifting all compute resources to an open economy, where you can list anything on the network and get paid to put that on. I think that's going to have a very profound and massive impact on the computer industry at the very least.”
---------
(02:29) Getting Started in Technology
(04:00) Evolution of Edge Computing and IoT
(06:45) Electrical IMP and Particle
(08:22) Internet Connectivity and Rise of Edge Computing
(10:00) Latest Trends in IoT and Kubernetes
(13:30) Kubernetes Driving Transformation
(16:00) Orchestration
(18:10) Machines Operating Machines
(18:55) Principles of Web3 and Security
(21:00) Intersection of Security with Web3, IoT, and Edge
(22:30) Cryptocurrency, Helium Mining, and Device Networks
(24:00) Demand for Hardware and maintaining efficiencies
(26:20) Market Based Scheduling
(28:40) Metaverse Rental GPU Market
(29:40) Proof of Work and Proof of Stake
(32:00) Improving Environmental Impact from Cryptocurrency and Blockchain
(35:00) Machine Economy and Carbon Credits
(36:36) Carbon Accountability
(39:25) Doughnut Index / Doughnut Fund
(43:30) Views on Regulation
(46:45) Technology Breakthroughs
(47:25) Innovations in Democracy, Social Engagement, and Community
--------
Over the Edge is brought to you by Dell Technologies to unlock the potential of your infrastructure with edge solutions. From hardware and software to data and operations, across your entire multi-cloud environment, we’re here to help you simplify your edge so you can generate more value. Learn more by visiting DellTechnologies.com/SimplifyYourEdge for more information or click on the link in the show notes.
--------
Rex’s Book Recommendations:
[00:00:00] Matt Trifiro: Hi, this is Matt host of over the edge. The only podcast focused on teaching you about edge computing, the grid and the future of the internet. On this show, I interview experts and practitioners with deep knowledge and expertise in digital infrastructure and the software and technologies that support it we'll even throw in a little metaverse web three and cryptocurrency to keep it on trend.
Join us each episode for a mind, expanding romp through the vast technological and business landscape that is quickly to finding our new digital.
[00:00:28] Narrator 1: Hello and welcome to Over the Edge.
Today’s episode features an interview between Matt Trifiro and Rex St. John, Founder of Taroko Technology
Rex has spent the last decade building developer and innovator ecosystems at Intel, Arm, and NVIDIA. He recently resigned to work full time in Web3 to organize communities and developer programs. Rex runs a YouTube channel where he shares knowledge, analysis and insights about the future of hardware, software and crypto technology.
In this episode, Rex provides his views on the evolution and adoption of Edge computing and IoT, including how kubernetes is helping to drive transformation. He also delves into the future of what he calls the ‘Machine Economy’ - where hardware, cryptocurrency, blockchain, and the metaverse work in tandem to ensure continued efficiencies and support for business operations and security. Rex also raises the importance of improving the environmental impact of cryptocurrency and blockchain, and advancements in democracy, social engagement, and community.
But before we get into it, here’s a brief word from our sponsors…
[00:01:36] Narrator 2: Over the Edge is brought to you by Dell Technologies to unlock the potential of your infrastructure with edge solutions. From hardware and software to data and operations, across your entire multi-cloud environment, we’re here to help you simplify your edge so you can generate more value. Learn more by visiting Dell.com for more information or click on the link in the show notes.
[00:01:56] Narrator 1: And now, please enjoy this interview between Rex St. John, Founder of Taroko Technology, and your host, Matt Trifiro.
[00:02:05] Matt Trifiro: Hey Rex, how are you doing today? Pretty good. I'm awesome. You know, before we get into talking all things edge and crypto and web three, I want to make sure that our audience knows that even though you are a full-time employee of Nvidia, you are on my show presenting your personal opinions.
Is that correct? My opinions. Okay. Well, awesome. I like your opinions and we'll go right into that. So why don't we just start with how you got started in tech.
[00:02:32] Rex St. John: I flew out to Seattle and about 2006 to work at my brother's company. And he had started one of the early video game microtransaction based companies called wild tangent.
And they had figured out a way to convert advertising and turn it into a form of token that developers and players could use to monetize and play games. So I worked in video games for a little while, and then how's that. [00:03:00] That went under during the 2007 implosion and every engineer at the company had a job in a week.
And when I saw that, I decided that I was going to learn the program. So I spent about 12 months in a coffee shop with used college textbooks, learning, coding, and interviewing. And then I ultimately scaled that into a career in software development consulting, and then ultimately technical
[00:03:24] Matt Trifiro: events. That's pretty cool.
Did you have any sort of an engineering or technical background prior to the game?
[00:03:31] Rex St. John: Just marketing? I took one computer science class, but you know, my brothers are very into engineering, so they, they told me what to do and what the study. So I just followed their advice.
[00:03:41] Matt Trifiro: Yeah, that's been interesting. I mean, I, I started my career as a programmer and then realized that there were people that were much better at it than me.
And so I went into marketing, so it's like we crossed paths. When we think about edge computing, IOT comes up a lot because the devices, the things are actually at the, at the edge. Can you tell [00:04:00] me how you see edge computing and IOT evolving in particular? I know you have an interest in, in web three and crypto payment systems as possibly being a, a way to monetize some edge and also make it easier to consume edge.
Can you tell us a little about your thoughts on that?
[00:04:19] Rex St. John: So I kind of have this very simple binary model that I called device up in cloud down. So throughout my career, you know, I began in the Silicon space at Intel and I was part of Intel developer relations division. And I was part of Intel new technology.
And at the time it felt like a lot of developers were very happy where I've got an embedded device and now I'm able to connect this to some form of internet or connectivity, whether that be a LoRa or Bluetooth or whatever, or ant plus, and they're happy to take their embedded device and sense of signals someplace.
And that was kind of like their mindset. So that's been, that's [00:05:00] been like kind of the, the focus of. Over the last decade. And then lately it's become much more sophisticated as this edge computing conversation has come in. So I view edge computing as a, almost like a cloud phenomenon pushing down. So IOT device up and then edge feels like infrastructure or cloud down.
So with all the, the, the tools and practices developed by folks like Netflix and Facebook and Amazon. They've got some very sophisticated methods for how to spin up and manage fleets and very large clusters of devices. So they, they created a lot of software techniques for how to do that. And now they're trying to push those techniques down to the edge.
So they're trying to move that, that computation power from the cloud down closer to the. And then these two worlds are kind of combining, and there've been a lot of really nice frameworks put together, especially by LFH by yourself that explained this in great detail in the cell, what the IOT is kind of like the device edge.
And then like the edge is [00:06:00] like, you know, there's a million different taxonomies there. So that's, that's the basics of it.
[00:06:04] Matt Trifiro: I, yeah, although it's interesting racks is, so I guilty as charged. I spent the last five years trying to clarify and educate the world on edge, and now I'm starting to. That we're going to stop talking about it.
Eventually. It's just going to be all part of the internet. It's just, can the workload run, can the network and the compute deliver the quality service for the workload? Yes or no. And it doesn't matter where it is. It matters it can perform. So, so that's, that's a great framework. So why do you say. Uh, what does that, what
[00:06:36] Rex St. John: does that mean for me?
That means I've got some embedded device and I'm just happy to get this thing connected to theater that and sending a signal. So like when I began, we very happy to have things like the electric imp and electric imp was ultimately acquired by. And electric imp and devices like particle and some of the,
[00:06:56] Matt Trifiro: what are these?
I don't think my listeners will necessarily know. What does it mean? Sort of
[00:06:59] Rex St. John: like a maker [00:07:00] devices. These are like little micro controllers, like no Linux and with electric amp, some of the biggest challenges that we're faced with these devices is just how do you make it simple and easy to set it up provision and begin sending them?
So the electric end people had some extremely clever methods of like, you take your device and it's got an led and a flashes and you hold your phone and you record it. And then it automatically pairs it. So they just had these really nice mechanisms for how to get these things connect to the internet.
So that, for me, that was the. The early IOT was like, how do you, how do you actually get this stuff and make it usable and easy for developers to get some value out of it in a really short period of time? Yeah.
[00:07:38] Matt Trifiro: I recently interviewed Rob Tilly, who most recently ran IOT at Ericsson and he said, look, at the end of the day, 90% of the value of IOT is replacing the guy with the.
Yeah. You know, somebody got to go out to an oil rig and take a measurement and get back in his truck and drive back to somewhere. We're just automating and making that [00:08:00] faster, which I, you know, I think is a, it really speaks to, you know, we've been talking about IOT, I mean almost two decades now. Right?
Yeah. And I think every, every year was going to be the year of IOT. Yeah. And maybe we actually are in, in the year of IOT. So where are we in this arc of connecting things to the internet? I mean, are, are we still mostly replacing the guy with the clipboard or are we doing more interesting things?
[00:08:28] Rex St. John: It seems like we did spend a lot of time with just trying to get things connected and being happy that it's connected to the cloud.
And then once we'd done that, Then it kind of felt like, all right, we did that. Is that actually going to be the thing that drives massive volume and like all this business and all this profitability and like, is this going to create some sort of, you know, new apple and there's some extremely successful businesses that started and scaled in that?
And then some of them had phenomenal exits, but now it feels like [00:09:00] the winners of that year have been chosen. You know, you had electric GameSpot by Twilio and then particles and the Belina's and, and, and folks like that. And now we're in this new era, which is Kubernetes has come along Kubernetes and cloud native.
And this is changing the conversation in, in some really interesting ways and creating a whole new swath of innovative. And then there's further generations past that, but you know, I, I do think people get tired. People get tired of hearing about IOT and the conversation has to change. And then ultimately, as you said, it's driven by the workloads and the profitability.
If people aren't investing, then the innovations going to move slower. And then the work, if the workloads aren't there, then it's, it's not a justified innovation. So I think right now it feels very much like edge computing is, is on the road. So let's,
[00:09:48] Matt Trifiro: let's talk about that. What, what are you seeing today that's meaningful and interesting to you that you see as fundamentally different potentially starting this acceleration that you didn't see in, in your [00:10:00] prior experience with
[00:10:01] Rex St. John: IOT?
I would say after the last two years or so, I've talked to like several hundred. Cloud native edge, native open source edge, AI edge IOT, telco edge, talking to all these different companies and startups and, you know, AI companies, ML ops company. So I've gotten a pretty broad view of the current state of the market.
And my current read of it is that Kubernetes has just absolutely blown up and. The industry had embraced containers as a, as a good way to manage software. And that took a little bit of a lift for people to every, everybody to rebuild their, their software and infrastructure on containers. Now, Kubernetes is coming along and adding additional layers of complexity, and this is causing people to have to restructure things and reconsider how they're building their platforms.
Reconsider who like what skill sets they need and how to structure. So this advent of Kubernetes is causing a lot of innovation in the, in the form of these turnkey Kubernetes [00:11:00] orchestration platforms. And then there's also been some side effects such as people questioning whether or not Kubernetes is the right tool for the job for a lot of different situations.
And when they're not, it's too heavyweight. So known as. Uh, subsequent innovation and how to slim Kubernetes down or replace it with something that's much more streamlined that, that provides the value of Kubernetes without the Kubernetes. So that's, that's kind of the stage we're in now where it's just like, it feels to me like we've gone from all right.
I'm happy. I just have my microcontrollers sending data. And now it's like, actually the thing that's sending data is a full computer and it's connected to that. For computers, maybe it's organized into a cluster of full computers. So I always use the example of. Jetson mate border from seed, which is a cluster of Nvidia Jetsons.
So this type of device is fundamentally different from like just a electric amp, sending some MQTT data. This thing's running full Lennox is doing GPU acceleration. They used to have [00:12:00] containers, needs to have Kubernetes east to have networks. So
[00:12:03] Matt Trifiro: this, this isn't a video podcast, so I'm going to describe this.
So it, Rex was holding up in literally the Palm of his hand. So it's about the size of, and it's four by four inches and it had three, I guess the in Jetsons is the three Jetsons with the heat, with the heat sinks on it. So it's a small device. I mean, it's something you could fit in a, in a, a small Tupperware box.
And, and how, how powerful
[00:12:22] Rex St. John: is that thing? Pretty powerful. I think if you're at a tad up all the chorus name, I couldn't tell you off the top of it. But this is a quite a powerful device, which once you begin to clustering it. So this is more of a edge server than it is. So it's more similar to what we have in the data center in a way than what we might've previously had at the edge.
So Kubernetes is enabling. And I just see the industry is having to process what that means right now. And currently the first phase of it seems to be okay, we're just need to be happy that we're able to plug this thing in, put Kubernetes on it, have high availability [00:13:00] and, and like just have it up and running as part of our, as an extension of our infrastructure.
But I think it's going to evolve to include AI in the next maybe year, year, or two as that, those fundamental building blocks go with. Yeah,
[00:13:13] Matt Trifiro: that's cool. So let's, let's, let's put a pin in this for a second and let's, let's double click down into some of this, this basic terminology. So you mentioned Kubernetes.
Most people know what Kubernetes is. It's a, it's a cluster orchestration system, but other than the fact that it's popular and utilizes containers, which I think everybody understands. What is it about Kubernetes that you think is putting in a position to drive a transform.
[00:13:37] Rex St. John: So let me rewind this. So one of the most interesting conversations I had when I was at Intel is I was talking to the CTO of an automated robotic forklift startup that raised like $200 million.
And I was trying to pitch him the Intel Juul, which is like this 300, $400 module. And he's like, how much does that. Oh, it was $400. I'll buy five. Like I don't care. Like [00:14:00] I'll put, like, I'll decorate my forklift with, with computers. It's a $200,000 forklift. What I care about is that my engineer doesn't have to spend any more time learning how to use it.
And it doesn't interrupt my engineer because that's a $400 device. But if it takes my engineer 40 hours, then like the cost of that engineering. It's extremely higher than the cost of adding a new device. So that, that, that human element of like having that program or having that person that has to go out and be on site is, is far more expensive than a lot of this hardware.
So I think that the opportunity with Kubernetes is to add additional resiliency. Like I met like this device. Four different four different modules clustered together. And one fails. Kubernetes has a, has a, has a practice called high availability. So it can recover. The two cores that are still running can recover the other, the other computer that went down.
Or take on some of this workload. So [00:15:00] when you start doing the math on, like what it cost to replace the device, if one of those modules goes down and you lose the use of that device for three weeks, maybe it's better just to put, build a little edge, edge cluster like that and let in and be happy with one of them.
And have the resiliency that Kubernetes offers in that you can have those high availability clusters at the edge for that kind of use case. That is
[00:15:23] Matt Trifiro: the standard of data center. You just assume everything's going to fail. Yes. And you build software to go around it and that's, and it's, and it's very much more cost efficient to, from the standpoint of the historical way of providing that level of availability, it was redundant, mechanical.
Yeah, and software high availability is much more efficient. So, so I think I hear you saying that that one of the advantages resilience, but also the ability for a developer to use the same tool when writing their server software that they use to deploy that the elements that run on the device and that, that in this, into a single workflow.[00:16:00]
And then I think I heard one other thing, which will be a nice transition, I think, into some of the web three conversations, which is orchestration. Yeah. Can you explain.
[00:16:09] Rex St. John: So I mentioned containers caught on, in a big way. Everyone knows that. And then they caught on so much that we began having too many containers.
There were a lot of containers. And I think the folks began to realize that just managing and maintaining and keeping these containers healthy was becoming a burdensome thing. So they desired this automated system that just kind of took care of it. And that's where Kubernetes came from and the edge computing space you don't necessarily.
Often care exactly where the workload is running. If I've got some edge cluster that's in some, some autonomous retail store, that's running a bank of cameras to watch everything everybody is doing. And then I have a co located data data center. That's a mile away with Kubernetes. If, if that local cluster fails in some way, then you can kind of fall back to [00:17:00] that next best backs cluster.
And Kubernetes enables like kind of use cases like that. So when you treat everything is just an extension of the data. That's what you get versus like trying to draw a box around your, your retail store and being like, well, this is a different system. And then my data centers is a different. It's like, actually you want those to be the same system so that you can have a gradient balancing.
[00:17:21] Matt Trifiro: Yeah, yeah, yeah. That's, that's really interesting. And it's orchestration. The word is actually a great word, right? Because you can imagine an orchestra with a hundred different musicians, each of the different instrument and a conductor, which would be the schedule, or I suppose. And the conductor has all the quote unquote workloads, which are the notes that need to be played and points at the different musicians and tells them when to play what note, and if the violinists pass out the first violinists, you've got to say that without getting to more of it here.
And so orchestration is a really good thing. I think the other thing that people miss, not, not because it's, because I think it's subtle, but it's really true. We're moving from a [00:18:00] world. Humans operating machines to machines operating machines and all this happens to it has to happen at machine speed. So even without human intervention, right.
Because it's going to happen in potentially hundreds of millions of locations. Yes. Many, many, many times per second or millisecond or microsecond. Yes. And so you need these algorithm approaches. Okay. So, so let's imagine a system where we have. Thousands of computers, some of them on the devices or on the Jetson boards or in a data center, in a micro modular data center near near the facility in a regional data center in a core data center.
And they're all running Kubernetes. So that's a world which I see that emerging. How does, how does web three finish to this? And before you answer that question, let's try to define web three. Can you, can you. Provide a high-level definition
[00:18:52] Rex St. John: of web three for boy. I mean like there, there are some much more qualified people to answer that principals.
[00:19:00] Yeah. Basically what's happened is, is we've decided to take this transaction layer and build it on these shared system. It's kind of the, about the migration from single nodes being in charge of their own security to this sort of. Crowd-based security where everybody is in charge of, of maintaining the shared security, the system.
So I'll explain what I mean. So like, let's say you have one bank and you only have one security guard. And if the security guard takes a nap or the bank gets broken in around the back where the security guards not looking, then that system is, is less secure. Then a system where you take all the money and you put it into a glass fault and you surrounded with 10,000 security guards.
So those 10,000 security guards can all, like, you can't take any money out of that vault without, without somebody noticing. So what they've done is they've, they're kind [00:20:00] of trading privacy. In exchange for a different type of security. So that by, by turning transparency up to 100%, just letting everybody see how much money there is and where the money is and what locations the money is.
They're able to create a, a much more serious level of. By making that kind of trade-off. Whereas before they're like, well, let's kind of like bury our money in the backyard and hope nobody, nobody digs there. So this is the opposite. That's like security through obscurity and these blockchain based systems.
And these proofs proof of stake systems rely on large crowds of people to, to create this kind of collective.
[00:20:41] Matt Trifiro: Right. And, and I, I certainly understand how that applies to the cryptocurrency world and other uses of, of, of general ledger technology. But how, how does that intersect with web three and how does that intersect with IOT and edge computing and so on?
[00:20:58] Rex St. John: No web three, [00:21:00] which is somewhat of a little bit of a review. Of of crypto to, you know, make it more mainstream. Th th they're taking the, the wallet, or they're kind of like rebuilding everything around this wallet structure where it's like, I'm going to use this wallet as my place, where I store all of my assets.
I'm going to use it to log in. I mean, use it to. I mean, you use it to sign things like sign documents. So the wallet is kind of like becoming the new center of user identity and it's owned by the user and nobody can mess with it. So by taking the wallet out of the hands of everybody, all the big companies and every.
They just kind of rebasing everything around the individual and their own data and their own money and everything is going to be owned by owned and operated by them. So that's, that's sort of the big switch in web three, and that's going to have a, that, that switch is going to have massive ramp ramifications for IOT and edge computing.
And I can already see it happening how's that? Well, every device is [00:22:00] likely to operate its own wallet, and I'm already seeing this, this type of switch take place. I'll take an example of just people are familiar with mining. Like I buy helium. And I'm going to set that up and I'm going to put it in my window.
Let's explain what helium is. Super interesting. Helium is a, it's a, it's a global network of, of people that are buying these gateways and they set the gateway up with an antenna and a window. I have like seven of them and when they plug it in, it connects it and it begins providing IOT connectivity to other devices around it.
And it begins my. The this cryptocurrency called helium or ancient hint is what it's called. So you're actually earning money with this device that has its own wallet in a way. So like you're providing that connectivity in exchange for, for earning these tokens. So, so the
[00:22:52] Matt Trifiro: idea is if we can get lots and lots of, lots of people investing in these helium devices like you are, because they're going to make a little coin on the backend [00:23:00] that helium the company could potentially create a global net.
That people could pay helium to connect their IOT devices to through the shared infrastructure that you've helped pay for.
[00:23:11] Rex St. John: So they're up to 500,000 devices, and I think the tokens valued into the almost tens of billions. It's still sub 10 billion, I think so they're, they're expanding their model into 5g and probably other there'll be other models in the future as well.
[00:23:27] Matt Trifiro: Yeah, that's, that's super interesting. And, and how. How do you see that translating to workloads? You know, I recently watched your video where you're wandering around the look like Doug firs and somewhere in Seattle or near what Seattle probably. And you were talking about the ability to use a token or a cryptocurrency to, to pay for a workload on some device.
You don't. Oh, I may not even know where it
[00:23:59] Rex St. John: is. [00:24:00] We have to actually rebase the entire conversation about what the problem is that we're trying to solve. Yeah, that'd be great. Yeah. So there's a few problems that we're trying to solve here. And like the first is the fact that we're constrained in the amount of hardware, resources, connectivity, energy, security, storage that we have available.
And we want to increase the efficiency of those systems because it feels like, or it seems like. The demand for the hardware is far outpacing the supply of the hardware. And that's, uh, I mean, kind of on a global basis. And like people are having to stop producing cars because they can't get the semiconductors.
So, and, and then meanwhile, I think there's fundamental physical barriers, which I'm not a physicist, but there's, there's limitations to how much more effective and powerful you can make that hardware. So in the future, there's going to be a major shift, which is going from how do we make more powerful hardware all the time.
How do we make maximum use of the hardware that we're able to produce to meet the [00:25:00] meet the overwhelming demand? So we talked earlier about Kubernetes and like you mentioned, the orchestration and the scheduling, and you also mentioned that we need to move towards a move away from individual humans, making these decisions about this machine should run this work.
On next Tuesday for this long, you know, that's not gonna work. And the other thing that I feel like is not going to work is, is some of the scheduling that has been used historically, which is some arbitrary scheduling of like Linux kernel. We need to dedicate the CPU to this workload because it's been, it's even worse than
[00:25:37] Matt Trifiro: that.
Sometimes there's an it guy who's who has got masking tape on the rack and he says, okay, that workload is going to run on that VM in this machine.
[00:25:45] Rex St. John: Yeah. I think we're going to move to a world of what I, what I'm calling market-based scheduling, which is the people are going to be. On the available compute resources, storage energy, and there it's going to [00:26:00] be backed by the money that they have.
And they're going to say, Hey, I want to use that GPU because I need to run to this job and I've got enough money to. And I'm going to outbid everybody else and I'm going to run it for two hours. And then that, that transaction is going to go out to the internet somewhere. And somebody in their region is going to say, well, I'll sell you that com I'll sell you that capacity.
Maybe I'm aquaponics or digital Realty or American tower. And I'm participating in this network of market-based allocations and I'm going to sell you that. So I think that's where things are getting. And, and, and helium is like the first hint of where that's going like that. I think that's going to come pretty soon.
[00:26:41] Matt Trifiro: That's interesting. Yeah. And I think to me, there's a, a really obvious reason this has to work this way, and that is because it's going to have to happen at machine speed. And so you're going to have to have machine speed negotiations, which means you're going to have to attach the price. You're willing to pay to the workload and rely on some system.
I mean, this happened. In a limited basis [00:27:00] with AWS spot instances, for example, that is an attempt for Amazon to sell unutilized inventory on a, on a market basis. And I can, I can see that being exactly how it works. You think about, like you were saying, the, the developer, right? The developer doesn't care where it runs and as long as it provides the security.
At the cost at the carbon footprint, at the latency, at the, whatever, the quality of service you put in the manifest and an intelligent orchestrator should be able to assemble all the fungible assets and schedule the container or the pod on the appropriate asset or tell you, you need to bid more money or it can't be done.
[00:27:44] Rex St. John: That's exactly what I think is going to happen. I don't, and I can't find anything wrong with that. I see it happening already. And, uh, a few weeks ago, these guys owed toy. They've got something called the render token. Now what they're doing is [00:28:00] so like they raised, I believe they raised a bunch of money. I could be, I don't want to comment on.
And likewise, like in the same week I saw this guy and he was taking pictures of, he had like all these Ethereum mining, GPU racks, he had this entire warehouse and he's like, I have to shut all these down because mining Ethereum isn't profitable anymore. And I'm like, well, maybe we're not, you know, as proof of work algorithms are exited, Ethereum is the second largest blockchain.
They're going to move to proof of stake. You're going to end up with all these miners and GPU's just lying around and what's a better workload than the metaverse to come in and say, here's some render tokens. I'm going to borrow those, those GPS you aren't using anymore. And I think there's going to be the proof of work market as it declined.
Has kick-started this metaverse rental rental GPU market that, that I see coming in the future. And the, this is my opinion. This is not my employer's opinion. This is [00:29:00]
[00:29:00] Matt Trifiro: for those who don't know. Can you explain the difference between proof of work and proof of. Yeah,
[00:29:05] Rex St. John: excuse me. I mean, proof of work is you basically just create a complicated equation with a lot of horsepower problem.
And then, so you can prove that Y and then that's used to secure Bitcoin, and there's been a whole lot of discussions and press there. And proof of stake moves towards this model where it's like, it relies on a lot of people running, what are called validator nodes. So it's like I can run a very cheap device.
It could be, my laptop could be a raspberry PI and through the magic of cryptographic mathematics, I can validate or verify. The the, the central blockchain and it's it's security without having all this overhead, not having big computers. So they're expecting there to be this world of, you know, all the, so I don't get too far in it, but it's a very lightweight, much, much, much, much more lighter than.
Way to secure the blockchain then. So
[00:29:59] Matt Trifiro: [00:30:00] if we're able to use our computers, our unutilized assets more efficiently and depending on,
[00:30:07] Rex St. John: and on that point before you go further, please. So the, the, I just double check my math, but I saw the most recent data is that 30% of data center hardware is underutilized or it's.
Most of the time. And then 90%, in some cases, 90% of energy in the data center is waiting. Uh, in some, some circumstances. So that's what we're dealing with. Like you're dealing with a lot of idle hardware and a lot of wasted energy.
[00:30:36] Matt Trifiro: Yeah, that's true. And, and we know that you can drive utilization up really high with intelligent algorithms and design of financial analysis companies like two Sigma that they regularly run at 96, 90 7% utilization.
And so you're right. I mean, I, I, I'm looking around my office and I think. Uh, easily have two dozen devices that are doing nothing [00:31:00] except maybe updating an led. Yeah. Do we have enough connectivity
[00:31:03] Rex St. John: to support that? Uh, helium is expecting that they're going to be, be operating more 5g nodes than anybody else in the United States.
In the near future, they've just begun releasing 5g enabled hotspots. I I've ordered one. I'm on the list. I'm going to get two of those things. So it's, it's happening. I mean, this is not a fantastical scenario. Like it's, it's coming.
[00:31:25] Matt Trifiro: And so the, I mean the environmental potential environmental impact must be massive.
[00:31:31] Rex St. John: So here's where things start to get really interesting. So over the last six months or so, I've been, I've been watching web three is relatively, it really peaked in the last three months in terms of like buzz and everything. But what had happened under the, under the hood is. A number of people are looking in the industry are looking to figure out how to improve the environmental story of cryptocurrency and blockchain technology last [00:32:00] year, like Elon Musk and all these guys in the, in the media as well.
They pointed out a lot of shortcomings with proof of work. So that, that led to kind of, uh, a lot of the industry, you know, gaining, I think being tarnished in, in terms of its reputation. But at the same time that we're transitioning into proof of stake now across the board for a lot of the, the major blockchains.
So what these guys had done is there's a wall street journal article about a week or two ago, clean the Dow and to Ken per protocol, they've, they've set up a tokenized carbon market where there's these kind of certified. Uh, carbon credits, which had had this like trading volume, I think it was like a hundred million dollars of trading volume of these credits per year.
And when Klima and to Ken came along, I think that they introduced, they they've, they've built this thing called like a carbon black hole. Where their goal is to drive up the price of these carbon credits by tokenizing them [00:33:00] and allowing you to, to exchange them for their, their cryptocurrencies. And I think the volume of exchange has gone up to like a hundred million a week in some, in some service.
So, so it's just had a massive, real impact and it's gotten, gotten coverage by the wall street journal. And, uh, to con in particular, I've been having some conversations with them. They're introducing building blocks is what they call it like Lego bricks for representing carbon as, as like a fundamental unit of transaction.
And when, when you start thinking in that way, where you're like, okay, well, I can buy these carbon offsets and I can integrate them into my computing system and into my data center and into my IOT system. And in theory, I could, there isn't much reasonable. I couldn't have my device. If it's receiving payments for, for compute jobs, pay some of those payments to purchase carbon offsets, to, to account for its own impact.
So this world is coming and I think this is going to have a, this, this scenario that I'm outlining is going to have a major impact on the world of edge computing and [00:34:00] data centers. And I think we've yet to see how far it can go, but I think it's. And
[00:34:05] Matt Trifiro: what do you see the end result being?
[00:34:08] Rex St. John: I think the machines are so I've introduced this term.
It's not a new term, but it's, it's not really in favor of, I call it a machine economy. The machine buys buys itself. It justifies itself. It pays for its own insurance. It pays to replace itself. It pays for its own carbon offsets. And then there's there's money flowing in, operates on its own. Between itself and the machines around it and it exchanges value.
So this machine economy that I'm outlining, I think there's an open question about like what the currency is, and maybe, maybe it's literally energy points, you know, like maybe it's all tokens that represent units of energy
[00:34:49] Matt Trifiro: use and units of energy
[00:34:50] Rex St. John: saved, or maybe it's, maybe it's literally carbon credits, maybe everything is rebased or.
Carbon, uh, units of carbon credit as it's fundamental and [00:35:00] this entire machine economy that I'm, I'm talking about, that that pays for its own offset. That could, that could be well, if your
[00:35:06] Matt Trifiro: proof of work is sequestering carbon, that's not such a bad.
[00:35:10] Rex St. John: And then hopefully everything will go to, to proof of stake, you know?
And then, and then once that happens, like the proof of the footprint of proof of stake is a tiny fraction of proof of work. It's, it's quite efficient. So I think that future I'm hoping that future is, is what happens.
[00:35:26] Matt Trifiro: I've never heard that term machine economy. And I've never thought of the idea that a machine.
I mean, you, you, you have some of this to some extent, like in really indirect ways you have subsidized devices, right. Mean. But nothing that's quite as integral and yeah. Yeah. That's, that's pretty neat. That's pretty neat. I might have to have you back on in a year and we can see, we can see where this has
[00:35:47] Rex St. John: gone.
It goes much further, you know, like you've already got it in your house. If you've got a power wall from Tesla that's and then you've got a solar roof and then you've got a car. My brother, for example, he's [00:36:00] got power wall, solar roof, and he's earning money every, every month because he's selling the power back into the.
And then if you add a Tesla, you charge the Tesla from the battery stored in your Powerwall. So those are all using energy as an account, right? Right. It's dollars, but I don't see, I don't see a good reason why all of those systems and systems like them, wouldn't move to this carbon accounting system. And I think we're going to hear more and more about, I don't know what we'll call it, but I want to call it carbon accountable.
So we've all seen the charts about how many devices and how many microcontrollers and how many processors have been put into the world. It's just this exponential thing. It's like, you know, let's, let's put, let's put processors and everything, and that was the dream of IOT. It's like, we're going to put sensors and processors and everything, and that's well underway.
You know, I think, I think just based on the number of sales of microcontrollers and, you know, the profusion of that, that's how. But I think now that you've got sensors and everything, you know, they might as well also [00:37:00] be used for a car carbon accountability of those systems. So, you know, why wouldn't my entire house have a very strong and clear view of how much of its carbon footprint.
Why wouldn't my car? Why wouldn't my computer? And then why wouldn't all those things be unified by some common. Yeah. And, you know,
[00:37:18] Matt Trifiro: I recently interviewed the CEO, Allie fan of it renew, and we were talking about there's other costs that aren't accounted for, like the disposal cost of a device when it's no longer useful.
And so you could see actually making sure all those things, all the ecosystem services, whether it's oxygen or platinum or carbon. True cost accounted for in the device and the operation of the device. So
[00:37:47] Rex St. John: I've written some articles on this topic. There's a book called doughnut economics. And the core theory of this book is modern economics.
Is very primitive in that it's like, well, [00:38:00] supply goes up and then demand goes down and you know, the price cost this and doughnut economics attempts to add in the secondary element, which is you also have to account for the damage or impact of that economic activity. So if gas prices fall and sales go up also, you know, sea turtles lifespans go down, you know, You have to embrace the full picture of all these elements coming together.
So I've, I've put forward some articles and discussions around something that I call a donut index or a donut fund. And I'll talk a little bit about exactly what this is. So I might need to first visit the topic of decentralized finance. So if you, if you've been ignoring blockchain and cryptocurrency and just ignoring it and hope it goes away, I've got bad news for you.
It's not going away. It's coming and it's getting bigger every day. And like, you can't run. So like that, like I'll just get that. So decentralized finance is kind of like [00:39:00] they've figured out how to do all these incredible new ways of, you know, earning interest and compounding interest and creating index funds and creating synthetic assets.
Like there's a, there's something called mirror protocol. And what this does is I can make a mirror asset out of Tesla. And the way it works is like you've got something called an observer that just watches the Tesla. And then it updates the underlying token. And then there's some, you know, this, some math around like proving that that's an accurate update.
So you can buy Tesla without ever. You can buy Tesla stock in the cryptocurrency space without ever owning Tesla stock. And like, I, that's not financial. Yeah, like this, people are buying this
[00:39:40] Matt Trifiro: stuff. It's like, it's somebody, if you buy a token, does it actually, does it actually buy into the,
[00:39:45] Rex St. John: you're just buying a mirror dash?
You know, what's really
[00:39:47] Matt Trifiro: interesting about this and I, I am, I am a no coiner, but I'm not a no, no crypto. And, and so I've only studied this a little bit, but one of the things that's really interesting about this. One of the [00:40:00] skeptics of cryptocurrency tend to say, well, it's a giant Ponzi scheme because at the end of the day, there's no intrinsic value in the underlying Bitcoin, for example, whereas.
It's actually has demand. There's a flat demand for gold and you know, but this is, there's a predictable, it's predictable, but it is when you need it. You can't, you can't build electronics without goal. I mean, it's like, it's, it's a, it's an essential element. And so even if the, the speculation market on gold goes down, the actual value of the gold will still retain.
And what's, what's different about. Bitcoin versus a unit of energy in my Tesla wall is that the Tesla was more like gold. Like there's actually a thing
[00:40:39] Rex St. John: there that could be, yeah. Why not? Yeah. The carpet. Yeah, the carbon or the energy. It's real. We can prove it's there. Yeah. So when
[00:40:45] Matt Trifiro: people say it's going to go away, I think you have to ask that distinction.
Well, the, the sort of synthetic things that don't maybe. Resolved back to an asset, even though people are willing to suspend their belief and trade it. Yeah. That's interesting.
[00:40:57] Rex St. John: Yeah. It's just missing legal filament. I mean, [00:41:00] somebody can come around and introduce one and then they just have to put a legal bridge that says, okay.
And you can trade this back for a test on site.
[00:41:08] Matt Trifiro: Well, it has to be backed by an entity that can do that. I mean, that's why the U S dollar backed by the U S government. It's like, well, okay. It's not, there's not a gold standard, but yes, the government.
[00:41:19] Rex St. John: So getting back to the donor funds, decentralized finance allows amazing stuff to happen.
So I think when you look at a device, like, let's say my PlayStation or this computer, I imagine a world where every device operates its own donut fund in a way where it, it has, like, it knows. This is how much it costs and have lice life and recover me. It's I'm going to set aside, you know, every time somebody pays my helium gateway for, Hey, thanks for letting me use the 5g.
And the helium gateway is going to auto portion 1%, cause it it'll be expect to live for five years and then it will invest in even compound that [00:42:00] 1%. And at the end of the five years, the helium gateway could say, Hey, I'm going to send this to reclaim device protocol. It's a hundred dollars and a guy in a truck or a robot will come and take me in and I'm going to pay for all the costs of reclaiming and stripping it down.
So I picture a world where every device operates a doughnut fund like that that's, that's funded through all the different mechanisms that. Working, it can recoup all these kinds of supply chain costs that are incurred.
[00:42:29] Matt Trifiro: So when you, when you look at all these fantastical ideas and you're starting to see the germination of them and some are gonna work out and some aren't and some of my pains are gonna be right, and some aren't and some of your.
The things that you're excited about this future that you imagine, if you were given a magic wand and say, you could, you could like topple down anything, any one thing that's slowing this down or blocking this, like w what would it be? What would the thing?
[00:42:55] Rex St. John: Oh, man, such a big question. I might have to, I might have to like, come back to that [00:43:00] in a little bit.
I might need to think about that one a little bit. Okay.
[00:43:03] Matt Trifiro: Well, how about, how about a few of the ones that just you see now? I mean, regulation probably.
[00:43:10] Rex St. John: I don't know, like, I think people have like the opposite view of regulation than, than how it actually tends to work. It seems to me that regulation, I'm not a lawyer often follows the industry because I mean, who, who writes the legislation and who can afford the lawyers to, to figure out what the legislation could be and who could afford to educate politicians and like train them on like, okay, actually, you don't want to do that.
Like, we've done. $10 million to go study this problem. And here's a better way. So if, if the industry isn't incentivized to look at a problem, the funding to, you know, like, I don't think doing legislation is a free activity. I mean, you've got hundreds of congresspeople and senators and staff and like, you know, like they, they, don't just, they don't just do drop what they're doing and do random.
They're going to have to [00:44:00] prioritize based on economics or impact or threats that are prioritized. So if, if the money is there and the workloads are there and the market is there, then I think that that begins to finance the, the ability to modify the regulation. And I think that regulation in the United States for cryptocurrency has evolved and will continue to evolve because of.
Do you turn on? Like, I'm not, I don't watch this too much, but like you look on Fox news. Tucker Carlson sits down with an hour with Michael Saylor for a dedicated interview. And when I saw that, I'm like, what is going on here? Like Tucker Carlson has an hour of time on network television to talk to the Bitcoin guy.
And it can only be driven by financial profit. It's like the, the Bitcoin lobby has made enough money where they can buy ads and, you know, like whatever. Have a spokesperson. And then like that, that say money is what's going towards probably updating [00:45:00] regulations. So I, and also, meanwhile, I think a lot of folks are looking at the cryptocurrency space and saying, well, maybe we should tax this.
We're going to find some new ways to tax this and enter an income off that. So that's going to attract a lot of attention. So I think the regulation will evolve. It's not gonna be. Like people have this black and white view. They're like, well, that's just going to be illegal forever. Did you pay taxes on that?
I'm like, well, you know, like they're not accounting for, you know, things change and things will. If, because this is, you know, there's a lot of things for the government to love about cryptocurrency. So we'll just put it that way. There's some things they don't like, but there's a lot of things that they, I think they do like, yeah, we will
[00:45:35] Matt Trifiro: like, yeah, that makes sense.
What about technologically? Like what, what, what, what do you, where do you see the areas of most free fruitful?
[00:45:44] Rex St. John: Yeah, I really do think that just shifting everything from manual allocation to like the biggest thing I think is just shifting all compute resources to an open economy, uh, where, where you can list anything on the network and [00:46:00] get paid to put that up there.
Because I think that's going to have a very profound and massive impact on the computer industry at the very least. I mean,
[00:46:09] Matt Trifiro: any market that I've looked. Increased liquidity means a
[00:46:15] Rex St. John: better, yeah. Everything almost so outside of computers and machines, like sure. I think that, I think we're about to see a period of innovation in democracy, social engagement and community itself.
Which is going to be fairly unprecedented. So this is going to get philosophical and abstract. And I'm sorry. I've
[00:46:35] Matt Trifiro: I'm, I'm encouraging it. So please. Yeah, this is really
[00:46:37] Rex St. John: interesting. All right. So there's three books that I recommend people read to understand what's about to happen. The first is called the third pillar.
The second is called. The capitalism with out capital and the third one is called sovereign citizen. So to summarize the first book, basically human society has been fundamentally structured around community and, and social and [00:47:00] religion, uh, for the last a hundred thousand years. And only in the last. 400 years have the market and the state come around.
We didn't even have nation states until, you know, like prior to 400 years ago, we didn't even have capitalism prior to that. So these two forces have pushed community and society to the sideline and taken over a lot of its functions. And the theory of this book, it's a nice book is that the community is going to come back.
And re-establish its dominance and that's going to require some technological changes first, but I believe that that shift is happening right now. And you can see a lot of. For all over the place. So I wanted to go for it. And that capitalism, without capital talks about the idea that we're moving to a world where the fundamental world of, of physical things is too expensive and time-consuming to invest in, or, or significantly lags the pace of innovation, of things that are intangible, such as [00:48:00] ecosystems, marketing relationships and, you know, things like that.
So, T to understand a concrete example of this. It's like, let's say you are, you, you work 80 hours and you were in a thousand. And you say, well, how do I maximize the value of this? A thousand dollars to me? And option a is you buy a lawnmower and a snow shovel and like some gas for your car. And then you get some utility out of that.
You're like, okay, that was kind of, that's kind of useful that I, I did that. And then the alternative is you go and sit in front of your computer and you buy, like, you sit on Eve online and you buy like a battle Creek. That seats 2000 people and you can play and go on adventures and it has a holodeck.
And, you know, you can in the holodeck with your $200 VR headset, you can literally have any experience you could ever imagine. Like what's the better use of the $1,000. Like, what are most people going to spend the $1,000 on if they get the choice? So it's going to be this intangible metaverse thing versus like the physical real world.
So that's. And [00:49:00] then sovereign citizen. I won't go into that book that gets political, but that's, I think that's, I think that's going to be so read those three books in and that will prepare you for what I think is about to happen.
[00:49:12] Matt Trifiro: That's that's amazing. This has been such a Penn optic and unexpected conversation.
So I really appreciate you spending time with me and my audience. If people want to read your work or watch your videos or reach out to you, how can they find you?
[00:49:28] Rex St. John: Oh, just on rec St. John and Twitter, and I've got a YouTube channel. I don't put that much stuff out there, but that's how I found your
[00:49:35] Matt Trifiro: interest in this.
And it
[00:49:36] Rex St. John: was, if anybody finds that stuff interesting, you know, I could record more, but I've been kind of busy, like with my third child newborn and stuff. So let me know, feel free to write.
[00:49:46] Matt Trifiro: We'll do all right, everybody, we'll put some of those links to the books and so on in the show notes. Thank you, Rex.
Really appreciate you having gone. I really appreciate it.
[00:49:55] Narrator 2: That does it for this episode of over the edge. If you're enjoying the show, please leave a rating and a [00:50:00] review and tell a friend over the edge is made possible through the generous sponsorship of our partners at Dell technologies. Simplify your edge so you can generate more value.
Learn more by visiting dell.com.