Jill:
[0:00] Welcome to Destination Linux, the sunnydale of open-source slaying, where every line of code stakes the heart of proprietary evil. I’m your Buffy Summers this week, Jill, wielding the scythe of Linux with my scooby gang of tech warriors, ready to banish the vampire of closed source. Michael is our angel, brooding, code-forging vampire with a soul who can patch a server farm at some point when he gets around to it, probably tomorrow. His creed, if it’s not open source, it’s a curse that binds the soul. Ryan is our Xander Harris, loyal, hardware-hammering sidekick, patching servers with duct tape, grit, and always ready to equip his way through a kernel panic. This week’s mission, we’re staking out an interview with the head of engineering for OpenShift AI, Sherard Griffin, diving into his work on OpenShift AI and open source innovation. So grab your crossbow, sync your repos, and remember, in open source, we slay restrictions and save the world, one commit at a time. Let’s patrol. This is Destination Linux.

Michael:
[1:28] So we met Sherard Griffin at the Red Hat Summit, and we had such a great discussion with him that we had to have him on the show. Sherard is the head of engineering for OpenShift AI at Red Hat. Sherard, welcome to Destination Linux. Thanks.

Sherard:
[1:43] Oh, thank you so much for having me. We had a great conversation last time. So excited to continue it.

Michael:
[1:49] Yeah, absolutely.

Ryan:
[1:50] That’s a great discussion. We shut down the ball field when we were there together. Literally, the place closed down. We were just talking about it before the show. None of us realized it. We were still talking. And that’s how good of a conversation we had.

Michael:
[2:02] Well, the ball field, here’s a more important just ball field. We were actually at a Red Sox game and then forgot the game was happening because it was such a great discussion.

Ryan:
[2:10] Yeah, I don’t even have won. No, I have no idea.

Sherard:
[2:14] I was paying attention slightly. I was multitasking because my team are the Mets. And remember the Mets are playing the Sox. You know, man. Yeah, I was there. I was there in New York during the 86 World Series. I got slightly paid attention, but we had an awesome time. Yeah, I’ve never shut down the stadium.

Michael:
[2:29] Who won then?

Ryan:
[2:30] I didn’t know.

Sherard:
[2:31] Oh, the Mets. The Mets won.

Michael:
[2:33] There we go.

Jill:
[2:34] Michael and Ryan didn’t even know when I wasn’t there and I knew.

Ryan:
[2:38] Oh, wow. Well, all the people who are Boston fans, you can send comments about Sherard going to destinationlinux.net/comments and tell them why Boston is better.

Sherard:
[2:48] I do love Boston, though.

Ryan:
[2:49] There you go.

Michael:
[2:51] So we wanted to have you come on the show because we had such a great conversation and we had so many more questions we didn’t get to.

Michael:
[2:58] So let’s start off with your origin story. What sparked your passion for technology?

Sherard:
[3:04] Oh, I like you already, Michael. You make me sound like a superhero. The origin story.

Ryan:
[3:09] You are to us.

Michael:
[3:10] You are.

Sherard:
[3:10] I love you guys. I love you guys.

Jill:
[3:12] You’re awesome. You’re a red hat ninja.

Sherard:
[3:15] Yeah, my beginnings in technology really started. I was fortunate enough to start at a very young age in technology. In fact, my mom was always instilling education. My dad was always instilling technology. The early years of my life, up until I was about eight years old, I grew up in New York. And we lived in the projects in Brooklyn, the government housing in Brooklyn. My dad, he had a job as part of the Manhattan Transit Authority. He used to clean buses. My mother worked for the banking, but my dad, he would have the night shift. And then during the day, he would spend time teaching me how to code. So my first experience, Probably many of us that look our age with the fair amount of gray hair looking at you, Ryan.

Ryan:
[4:00] Yeah, a little bit.

Sherard:
[4:01] Experience with Commodore 64 was the first thing I got my hands on. And my dad taught me basic on it. And we programmed tic-tac-toe. It was the first time I got experience there. From there, ended up, you know, Commodore 64 128. Ended up going to Amiga. You know, started messing around with the Amiga and started learning technology there. But that’s really how it all began. And then the video game phase hit, right?

Ryan:
[4:28] Oh, yeah.

Sherard:
[4:29] I was like, game, I got to figure out how this thing works. And so, you know, I did a lot of stuff in that space. But yeah, very early, it was all about just basic and learning how to code.

Ryan:
[4:38] So, you know, a lot of parents are probably thinking right now, because I know I’ve gone through this with my kids, of how did your dad get you to be interested in programming? Was it something you were just naturally you wanted to do? Or was it something where it was like, son, you’re going to sit here and learn this type of thing? Like, how did he instill that want or did you naturally have it?

Sherard:
[5:00] Oh, I did not naturally have it. We spent a lot of time. I was fortunate enough to spend a lot of time in the early years with my dad because he worked nights and then he spent part of the day with me teaching things like chess and whatnot. And so my dad always had a passion for learning. He always had a passion for technology. So even though he was cleaning buses, he would write his own operating systems on a Tandy. He would teach himself how to code with different languages, just always had that thirst for technology. Uh, and so, you know, he basically said, Hey, you know, gonna spend some time with you. I’m going to sit you down next to you and to me, show you how to program. Now, what got me slightly interested was that there was a contest. This is back when there were physical magazines. You guys remember those days?

Ryan:
[5:47] Oh yeah.

Sherard:
[5:47] Physical magazines.

Ryan:
[5:48] Good old days.

Sherard:
[5:49] And not only that, mail ordered physical magazines. And so I can’t remember the name of the magazine, but there was a contest where, Hey, if you write, uh, uh, you know, code and it’s pretty good in this case, a video game, uh, then we’ll publish it. And so like, that was the goal. He was like, Hey son, that was like maybe six at the time. Hey, we got a chance to be in this magazine. Now you got my attention. I can be in a magazine. I can show my friends. So that’s when we wrote Tic Tac Toe. Funny thing about this, guys. I actually, maybe six months ago, uncovered in my closet the original printout of that basic code. Tic Tac Toe with my dad’s notes on it, you know, by Sherard Griffin and Bryant Griffin. And I actually uncovered that. I was like, holy crap. Now, you listeners may be wondering, did we ever get in the magazine? No, we did not get in the magazine. I waited issue after issue. Is the code in there? Is the code in there? No. One of the downfalls of that tic-tac-toe is you’re programming tic-tac-toe at an early age. It’s not so fun if you make the artificial intelligence so good that you never win. So it was only ever, the computer player, it was the only one that ever win the game. At best, you could tie it. But that was what kind of got me hooked. You know, I knew, oh, if I actually write code, it could end up in a magazine or all right. Now you got my attention. Let’s sit down and do this. So, you know, that’s kind of the beginning phases of it.

Michael:
[7:16] When I was in an origin story, I mean, that is 100 percent accurate because that is awesome.

Ryan:
[7:22] I’m thinking this is a redemption moment, though. If you’re willing, you don’t have to. I know I’m putting you on the spot here, but we can talk about it after. I would love to get a picture of that code. We will publish it in this video Here so that people in 190 countries which is far More reaching than that magazine Will be able to see the original code Of Sherard and his father I would love if we could get a picture of them You.

Sherard:
[7:46] All are my people I will go tomorrow and have a picture With me and my dad With that.

Ryan:
[7:52] Code We’re going to have redemption on it I love.

Sherard:
[7:58] That And it also tells It’s a telling tale of how much we don’t clean out our junk in our.

Ryan:
[8:05] Closet.

Sherard:
[8:07] I’ve been carrying this code since I was six.

Michael:
[8:10] That’s not necessarily you’re not cleaning out your closet. That’s more of like being a collectionist. Jill has tons of stuff, right?

Ryan:
[8:20] Now Jill needs to clean out her closet or stuffies, as you can see.

Michael:
[8:24] I guess Jill needs to get rid of some stuff.

Jill:
[8:27] Yeah, Sherard, I have over 700 computers now in my collection.

Sherard:
[8:32] Wow. So you probably have every version of everything I’ve ever had.

Jill:
[8:37] Yeah, yeah.

Sherard:
[8:40] Going back to, you remember when you had to flip the switch to turn the hard drive on? You got one of those?

Jill:
[8:45] Oh, yeah. I got several of those. I have a VAX computer terminal from the 60s.

Sherard:
[8:54] Wow.

Jill:
[8:55] That’s amazing. In the 70s, of course, the Apple IIs and the IBMs and everything in between. And all the computers you mentioned, I have an Amiga 500.

Sherard:
[9:06] That’s how the list goes on. Of course. Wow.

Michael:
[9:09] There’s a running joke on the show about like a random computer pops up in the conversation and we just say, hey, Jill, do you have it? And then she says, yes. And then we, of course you do, Jill. And so that’s just a running joke that we have. And one time it blew my mind because I was like, hey, is there a chance that Jill, you have like a mainframe or something? And of course she does.

Jill:
[9:32] Yeah.

Sherard:
[9:32] What? That’s insane. Wow. Wow. That’s impressive. I am new heights, new heights. But pretty soon, 700.

Jill:
[9:44] Yeah.

Sherard:
[9:44] I’m in the AI space. It’s like you’re going to need image recognition to keep up with what’s in your library, right? You know, can someone send you a picture and say, hey, do you have, because good Lord, that’s got to be pretty hard to keep up with everything that’s in your inventory.

Jill:
[9:57] I have it cataloged and I do bring out the machines every once in a while to fire them up and make sure they’re okay. You know, most of my machines work. There are some that don’t, but most do. And so you got to fire them up and, you know, give them tender, loving care and occasionally replace parts.

Sherard:
[10:17] Wow. That’s amazing. I love it. There’s a community around that too.

Jill:
[10:21] Yes.

Ryan:
[10:21] Oh, yeah.

Sherard:
[10:22] Fabricating parts for old machines.

Jill:
[10:23] Absolutely. What key experiences or projects solidified your belief in the power of open source software to drive innovation?

Sherard:
[10:34] Yeah, that’s a great one. Boy, oh boy. I would say I didn’t really fully understand or embrace open source until later on in my career. And I probably I probably shouldn’t say this, but I’m going to say it anyway. Um uh our family start a sentence our family you know we got a little bootleg in history all right so uh so i grew up in a world where we had an msdn license and you know yeah of course they tell you hey enterprise use only enterprise means you know my own personal startup doesn’t it i don’t know we had kind of blur the lines but uh you know so software was always freely available to me freely available to me. So I didn’t really appreciate the value of open source until later on in my career. So Jill, give you a little bit of background.

Sherard:
[11:29] I, when I first started into my professional career, I’d always been around solution architecting and trying to do things with data. I’ve always sold myself on, you know, being, you know, very good at data early in my career. And the reason for that, I remember coming out of university, I needed to get a job. And so there was one of two routes that I was going to go. I was either going to go the video gaming route or I was going to go the data route. And I figured, man, I could make a good amount of money if I go the data route. So the video game side would be fun, but I don’t know if I can make as much money. And so I focused on what I can do from a data perspective. That was right around, you know, right before and right around the time that Hadoop started taking off. And that was really the first time I started wrapping my head around open source. It was really centered around Spark and everything that Spark can do.

Sherard:
[12:19] And so I worked for a company back then that they had pretty much a proprietary competitor to Spark. But then I saw the rapid innovation of what Spark was offering. And so it was challenging. Here’s a company that has proprietary technology that can do similar things, but the open source community is growing so much faster. And not only that, I could see the code, I can play around with it, I can make modifications to it in my own distribution of it, which I started doing, adding new features, new capabilities. And that ended up being the genesis of my fascination. So much so that it actually drove me a huge, huge factor of driving me to Red Hat. Whereas a lot of people, maybe a lot of your listeners, they fell in love with open source back in the Linux days. And maybe even a lot of the associates that ended up at Red Hat Engineering came for what we did with Linux and what we did with OpenShift. Mine was actually around data and what the open source tooling, the Hadoop ecosystem allowed you to do with data, how you could all of a sudden crunch all of this information and do really fascinating things. That was when I really got a good understanding of it. And that was actually why I ended up at Red Hat, was to figure out a way to, through open source tooling, be able to process massive amounts of data. So that journey followed me all the way through until I landed at Red Hat.

Jill:
[13:43] Wow. You’re essentially creating a library for the data. Yeah, yeah.

Sherard:
[13:49] And being able to solve problems using solve data problems, which is if you think about this from like 10, 15 years ago, you had so much proprietary, so many proprietary BI tools, right? Everything was proprietary power BI. You had all these things that were just very heavy on the proprietary technology. And then here comes this disruptor with open source tooling, Spark and all these in Hive Metastore and Kafka and all these other things. You can cobble together these open source projects and have comparable even Apache Superset, all these things. Now you have comparable business intelligence capabilities using open source. So it just blew my mind. I couldn’t believe that this was available and you could actually roll that into your own products and make money off of things that are out there. Oh man, that was a game changer for how I saw software.

Jill:
[14:43] Wow. It’s almost like you’re taking a MySQL database, you know, and growing it in the cloud.

Sherard:
[14:52] Yeah, yeah, for sure. For sure.

Ryan:
[14:54] Very cool. What’s amazing about that too is, you know, anybody our age who says they haven’t had a moment of maybe borrowing software or utilizing a license that they weren’t quite is lying. Because yeah, that’s what we had to, that’s kind of how things were distributed.

Michael:
[15:08] I have never done this ever in my, Napster.

Ryan:
[15:13] And so, but what’s interesting about that is a lot of times it’s people, and we’re going to get into this later on in the interview, but it’s people who are trying to also get into careers in this. And you’re like, Hey, I need to, if I want to get a career in here, I need to learn this enterprise software. But then you go and you look for a license for it and it’s $10,000, $50,000, $100,000. But that’s what’s amazing about open source. All of a sudden, that super powerful tool that’s proprietary, that’s $50,000 a license, you can use without a single dollar. And now you have people that digital divide closes and they can go learn that stuff regardless of their income level. And that is super powerful.

Sherard:
[15:49] Ryan, you hit the nail on the head. I talk about this all the time. There are a lot of places I could be when it comes to employment. There are a lot of different projects, different communities I can get involved in. I do what I do and I’m passionate about it because of that digital divide. I see open source. It has to be the future. There has to be no other choice. And that’s because if you look at the income, the socioeconomics, what’s happening in the world and the access to technology versus people who don’t have access, it’s not shrinking. You would have thought by now, you know, we all, you know, you all are probably as big a Star Trek fans as I am.

Ryan:
[16:28] Oh, yes.

Sherard:
[16:29] Yeah, you have that utopia, right? Where there’s no more poverty. You think we’d be getting closer to that. It’s actually the opposite, right? And there’s always a challenge. We always have to gut check ourselves as to why is that happening. So for me, especially being an AI, that’s my chance of at least contributing to getting us closer to that, closing out that digital divide by making sure that the tools we do in AI, which to me is going to absolutely be the future of technology, at least the tools that I’m participating in are readily available for anyone of any background.

Ryan:
[17:03] Love it. Now, you kind of touched on this, but I still think there’s a little more story here of how did you discover Red Hat as a company and what convinced you this was the right place for your career?

Sherard:
[17:15] Yeah. So, the funny thing about how I discovered Red Hat as a company. So, first off, I live in Raleigh, North Carolina. And we have that big tower.

Michael:
[17:25] That’s a good way to find out.

Sherard:
[17:26] Yeah, that’s a big, good way to find out. We have that big tower downtown Raleigh. And it has, you know, just like your hat, Ryan, it’s got that big logo. You’re driving into downtown, you’re driving out of downtown, you can’t avoid it, you see it. I never thought twice about that. And in fact, my first exposure to Red Hat was actually at North Carolina State University where I graduated from there. I graduated in 2003.

Sherard:
[17:51] Red Hat was actually still on NC State’s campus, and that was where their headquarters was located. And I used to go over, if you’re not familiar with North Carolina State University, there’s an area where a brand, at the time when I was a student, it was a brand new technology area called Centennial Campus. And that was where they were moving all of the engineering, all of the AI, everything was moving over to that area. And Red Hat had an office there. And I’m like, huh, Red Hat, don’t they do Linux? Like, why are they at NC State? I never really thought about it. And then when I got the call, I was actually working at another company. A buddy of mine called me up and said, hey, I got a job for you at Red Hat. And I’m like, why would I go to a Linux company? I do data. That’s the last thing. I was like, no, no, no, no. Let me explain to you what we actually do and what our portfolio looks like. And let me tell you why I need someone who knows data to come to Red Hat. And if you think about all of the build systems. That’s what we do. We package software and being able to analyze the packages, being able to say, when is a build going to fail with leading indicators? That could save us time for getting Fedora out the door, for getting Ansible out the door.

Sherard:
[19:01] What’s the root cause analysis of why a build failed? Maybe someone had a bad PR somewhere, maybe it failed a networking test, or maybe it was a blip in the system. All these kinds of things, because building software is Red Hat’s bread and butter and security around that. Oh, now it started to make sense. We needed to actually, you know, analyze all of that massive amount of data that we were collecting from every single PR, every single build that we run to see how we can get software out the door faster. So that actually intrigued me, the fact that they had a data problem that needed to be solved. And then the catch was, but I need you to use open source to do it. Okay, now I’m putting my big boy pants on. That’s a challenge. That’s a challenge. I was like, okay, that’s the type of challenge I’m all ears for. I’m happy to roll up my sleeves and figure out how we get this done.

Michael:
[19:49] That’s awesome. Yeah. And speaking of challenges, what’s the most exciting challenge you’ve faced in your career?

Sherard:
[19:58] Oh, hmm. The most exciting challenge I’ve faced. I’m trying to avoid the obvious around AI because we all know, man, before ChatGPT, I had hair, but we all want to avoid the obvious. I’d say, you know, one of the most exciting challenges I faced in my career was not just convincing Red Hat that they’re…

Sherard:
[20:28] That you could actually use open source to process data. But it was convincing customers of Red Hat and convincing the open source community that Kubernetes and Linux is foundational for data and AI, right? And that was not easy to do. If you go back 10 years and you look at what happened in Kubernetes, Kubernetes was very driven around applications. It was not seen as a platform for data. And we did so much work to change that dynamic. I remember the first time we met with users of open source technologies like Kubernetes and we met with customers.

Sherard:
[21:07] They always wondered, why am I talking to Red Hat about data? Why am I talking to Red Hat about AI? Why would I even think about Kubernetes running data on Kubernetes? Why would I think about running AI on Kubernetes? And so back in those days, it was kind of like, you know, we had to keep showing the concept and we had to be able to show, hey, there is a place for this. You move forward now, Kubernetes is the default way. If you look at what all of the industry is gravitated towards, Kubernetes and Linux, if you look at scaling out AI to massive workloads, you look at thousand, you know, GPU farms with thousands and thousands of GPUs. If you look at inferencing, where you’re serving up models and those models are eating up 8, 16, 64 GPUs, having a terabyte of memory, all these things have really changed how people have looked at infrastructure. And that was probably the biggest thing was convincing people, I’m not crazy. Trust me, this is going to be a thing. You actually do want to use Kubernetes for more than just applications. You can do this for data as well.

Ryan:
[22:16] When you talk about convincing, and I know it’s probably a long process, but what was that, if you could think of the thing that you think helped move people from, no, this is not how I think of this product, to, okay, I’m interested. What was the, was there a magical shift?

Sherard:
[22:33] Oh, gosh, I got to do this in a way that won’t upset certain fans of certain technology. Um, if you look back, gosh, I can’t remember exactly when it was, maybe seven, eight years ago, there was a merger of two heavy hitters in, in Hadoop space, right? Should I, should I name the names or do we kind of know who I’m talking about?

Ryan:
[22:56] You can go from there. Yeah.

Sherard:
[22:57] Okay. All right. So, you know, a merger of two heavy hitters, uh, and that changed the dynamic of, uh, what exactly you were going to be able to do in those, in, in those communities, in the technologies. And there was a heavy, at the time, there was a heavy reliance on running big data, running AI outside of Kubernetes clusters, right? And this is a lot of the Hudoop ecosystem. Hudoop ecosystem became dominated by, you know, one single vendor. And that kind of changed the dynamic of what customers, what users of those platforms were looking for. They were looking for alternatives.

Sherard:
[23:32] This is when OpenShift started gaining adoption. Kubernetes started gaining more adoption. And then it became a conversation where people weren’t necessarily looking for another alternative until that merger happened. And when that merger/acquisition happened, all of a sudden they wanted to look at look at other options. It’s similar to what’s going on. If you look at the the virtual machine landscape today, right? I don’t have to describe too much of that, but we know what’s going on with the virtual machine landscape. People were happy until something happened and it’s like, oh, well, let me look for alternatives. And so that’s similar, something similar happened. I think that to me was a pivotal moment where we were able to show, hey, we can actually, so we first started, I first started my career at Red Hat, showing Hadoop running on Kubernetes. And that way it was a nice transition to say, you know what, if you’re not happy with what’s going on in that ecosystem, you’ve already invested in Kubernetes or you’re thinking about investing in Kubernetes, then I can, you know, we can show your workloads running on this. And then it quickly pivoted to, hey, let’s show your data running on this to let me show you how you can use AI to run on this. And that’s when we ended up starting the Open Data Hub community. And we started, you know, it’s called Open Data Hub. For those of you, I’m going to shameless plug, opendatahub.org. If you look at what we did with Open Data Hub, it’s called Open Data Hub because we focused on data problems and running data on open source infrastructure like Kubernetes.

Sherard:
[25:01] But we quickly pivoted to AI. We just got too lazy to change the name. So I just stayed as open data up.

Ryan:
[25:07] It’s a good name, though, because it’s got the open in it. You know what I love about that story is that I talk to people a lot about planting seeds. A lot of times people want, you know, whether they’re creating something, whether they’re wanting to sell themselves in something or an idea, they want to be able to just do it. And all of a sudden, everyone goes, yay, and celebrates and they win. But what your story is, is, hey, I’ve been working on this thing for a long time. We planted the seeds. We had the foundation for it. And then when this thing happened, we were ready to take advantage of that shift that happened in the market. And I think that’s an important thing for people to remember that things don’t always happen overnight. But if you get that stuff prepared, you’re planting the seeds. Eventually, if you hit it at the right time, you’re going to be able to turn that into something special. Yeah, exactly. It just comes later.

Sherard:
[26:00] Yeah, that’s exactly right, Ryan. There is a little bit of luck that’s involved there. And then we saw this with even Red Hat Enterprise Linux. There’s a little bit of luck that happened there. You look at KubeVert, right? There’s a lot of luck that happened with KubeVert. And then you look at what happened with Kubernetes and AI. A lot of luck there. So you’re exactly right. Like it’s not always, hey, you got a great idea and you’re going to go disrupt the market. Yeah, let’s go disrupt the market. You know, that’s that’s not always what happens. Sometimes it just takes quite a bit of industry luck to be ready for that.

Ryan:
[26:37] Yep. Now we got to be ready for Windows recall because that mess needs to be ready to capitalize on. Just that product. You don’t have to respond to that. All right, Jill, your turn.

Jill:
[26:46] Yeah.

Sandfly:
[26:47] As Linux users, we know that security is non-negotiable, but with threats getting smarter, Your security tools need to keep pace without dragging your system down. Traditional agents, they slow you down, leave blind spots, and it’s time for a smarter approach. That’s why Destination Linux is proud to be sponsored by Sandfly Security, a revolutionary agentless platform designed for Linux. It’d be nice to have Sandfly on mobile too. Yeah, that’d be cool. Because really, that agentless idea is really a smart way to look at a phone device as well because it could look for any anomalies happening within the device that would say, hey, that app you’ve got, there’s something wrong with it because it’s acting differently. We’ll send that over to Sandfly. Anyways, go to destinationlinux.net/sandfly and see how Sandfly can transform your security strategy. Sandfly doesn’t detect and respond or just detect and respond. It revolutionizes security with SSH key tracking, password auditing, drift detection, covering threats from every angle. Whether your systems are in the cloud, on-premise, embedded devices, Sandfly ensures they’re all secure without the headache of an agent-based solution. Listen to what Salinda Lakamge from Tait Communications has to say. Sandfly addresses a serious security gap in the industry. Intrusion detection on Linux platforms. The best part of Sandfly is its agentless nature, making it versatile application for any Linux environment.

Sandfly:
[28:13] Experience security that’s not just effective, but gives you peace of mind. No agents, no downtime, just cutting edge protection. That’s right. And also by being a listener and part of the community for the Destination Linux podcast, you can get a discount. But on the home edition, you’ll get 50% off by using the coupon code DESTINATION. So if you want to get the home edition for Sandfly Security and get a discount, well, you get it by being a part of this community, DESTINATION at checkout. And of course, you can just dive into the future of Linux security right now at DestinationLinux.net/sandfly. That’s DestinationLinux.net/sandfly.

Jill:
[28:51] So, and speaking of AI, you know, it’s very exciting, but of course people have trepidation and a lot of people are worried it’s going to be another technology that creates an even larger digital divide. How is Red Hat preparing to make their AI technologies accessible to the open source community? Yeah.

Sherard:
[29:13] Oh, that’s a great, great question. We spend a lot of effort doing this and there’s varying degrees of answering that, Jill. There’s both the, actually, it’s kind of three-pronged. Part of it is the partnerships that we have with academic institutes and what we’re doing with students and research work that’s going on in that space. You know, part of it is if you just purely look at the technology and what’s going on there. And then the other part of it is more on the social good side of things. And so I’ll kind of speak to all of that on the academic side. You know, to me, it’s critically important that we’re getting students engaged. I remember, you know, my first experience with Unix was actually at a university. Right. So when I was at NC State, you know, I think it was like I can’t remember the class. It was like CS 101 or something like that, or 105. And it was like, they sat me down and was like, here’s Unix. Now go submit your homework on it. I was like, what? And I got to figure out how the heck to do that. And that was my first experience with it. So it’s important that we start from that level. Red Hat has a great partnership with a number of universities. One of them being Boston University. We’ve worked really closely with them. I’ve personally worked with them over the past few years.

Sherard:
[30:33] We’ve collaborated with them on a project called the Mass Open Cloud, which is basically a public research cloud. And you can look it up, Mass Open Cloud. And it’s designed for students, researchers to get access to this. And you have a lot of corporations who have donated hardware. They’ve donated software. They donated time and effort to provide a research cloud for the academic institutes. So giving them access to these tools at an early stage, you even have OpenShift AI. We have that available for the Mass Open Cloud. And it’s easy for researchers and students to use these technologies.

Sherard:
[31:11] The other side of it, and then we also have a strong partnership with Shaw University. Shaw University is a historically black university that is in downtown Raleigh. I actually can look at it out of my office window in the tower. And we spent the last number of years actually working with that faculty, helping to design courses. I personally have taught Git courses and talked to them about AI and trying to make sure that they have access to all these tools. I told them, look, the new resume is GitHub, right? Your new resume is what you’ve done in GitHub. And, you know, people can see your contributions. And so, just kind of working with students and showing them not that you can just get access to this, but I think more importantly, Jill, is showing them that they can contribute. Even if it’s I’m correcting documentation, even if it’s I’m just going to modify a markdown file, or I’m going to possibly try to do a PR on a very simple issue. It’s so important for them to know that you don’t have to wait on the sidelines and hope to get involved. You can kind of just roll up your sleeves and as a student contribute to something meaningful. So that’s that’s the other part of it.

Sherard:
[32:21] On the technology. Yeah, it’s really cool. And on the technology side, you know, I think everybody knows this, you know, part of being Red Hat at Red Hat means that all this all the software we work on is open source. And that means we really work with a lot of the communities. We work with a lot of our partners to make sure that the best of breed technology around AI is available open source. If you look at what we’ve done with Kubeflow, for those of you who aren’t familiar, Kubeflow is the leading community for orchestrating AI workloads. That is the Red Hat is the the biggest contributor of that. If you look at VLLM, which is a community centered around serving machine learning models, Red Hat’s the biggest corporate community contributor of that project. You look at KSERV, which is also centered around model serving and inferencing. That’s one of the leading communities centered around model serving orchestration. Red Hat’s the leading contributor there. And so, you know, it’s, and then the partner side of things, we work with partners like NVIDIA, like AMD, like Intel to make sure that their technology is available in all of these different places as well. You know, NVIDIA has done a lot of contributions to VLLM. AMD, VLLM is their primary way of serving up models on their GPUs. We’ve done a lot of collaboration with the Kubernetes ecosystem, working with Google.

Sherard:
[33:46] So there’s a lot there. There’s a lot to unpack. But it’s our way. We’re continuing to push the envelope, but doing it in a way where we’re also making sure we’re inviting others, others who may not have thought about OpenShift as the default, inviting them to make their technology available as well.

Ryan:
[34:03] You know, all of that is just absolutely amazing. I love all of the focus on, you know, and we’ve seen it. So, yeah. When we went to the Red Hat Summit, one of the things in our videos when we were doing kind of our, you know, closing of the event, our thoughts on the event is when we’d go to these booths of these major vendors, Microsoft, NVIDIA, AMD, and we would talk to them. They weren’t talking to us about their proprietary products. They have them. They were talking to us about their open source projects at the summit, which was really cool to see.

Ryan:
[34:34] And one of the things we both specifically mentioned in our videos. I do have an interesting question, not one we necessarily have to answer today, but I wonder how we’re talking about the young kids, which is always important. But I know a lot of the older generation is very worried. There’s a lot of people talking about AI replacing jobs down the road. And there’s older generations now people have to work longer than ever before. And they’re really worried this technology is going to leave them behind. And I think there’s probably something that, you know, the bigger community and AI community needs to do in looking at that as well as an opportunity to maybe find some ways to get people more involved in that technology. So they’re not left behind as an older generation, you know?

Sherard:
[35:19] Yeah. Yeah, that’s a very, very good point. I can speak to like my philosophy on that. One of the things that I’m currently engaged in, at least internally at Red Hat, is technical skills and the Technical Skills Academy where we’re rethinking and we’re rebuilding a curriculum centered around upskilling our internal talent. And so working with various people across different parts of the company to figure out how to do that. I think that has to be something that almost open source is open source as well, right? It’s not just, it’s not the software, but it’s the education around that, right? Like, what does it look like to ensure that the workforce is not only prepared for AI, but they know how to use it as part of their everyday work life? Um, so that is one where, uh, you know, we haven’t really seen anyone solve that specifically and you’re right, Ryan, a lot of focus, it’s a huge problem. And a lot of people do focus on like the college level, right? Like how do we get them to be relevant in the workforce? But there’s probably a lot more, I actually, I know there’s a lot more that needs to be done with people that are already in the industry.

Ryan:
[36:28] Yeah, no, that’s, that’s awesome. I love that you’re already putting some thought

Ryan:
[36:31] to it and upscaling employees there at Red Hat. You mentioned focusing on AI strategies that work in the real world. What gap do you see between AI hype and the practical enterprise implementation? And, you know, Linus Torvalds is famous for saying like 10% of it is reality and 90% of it is hype. Where do you fall in there?

Sherard:
[36:54] Yeah, even only 10% is reality, that 10% is going to become the 90%, right? So let’s get ready for that. But I think one of the things that I see, and if you look at where things are headed, It’s all about trust and confidence, right? So, remember the days, we’re going to date ourselves again, right? Because I like you guys. And I like telling people, I’ve earned these gray hairs, right?

Ryan:
[37:18] That’s right.

Sherard:
[37:19] It’s a badge of honor. But remember the days when you didn’t quite trust what Google gave you back in the search engine, right? Maybe you would even go to page two, right? You go to page two and say, hey, let me double check and make sure. Now, we don’t even go past half the first page, right? As a matter of fact, you just get a summary of what your problem is now at the top.

Michael:
[37:40] Sometimes you get a response from Google and you’re like, that’s not right.

Ryan:
[37:44] But you’re right. You had to search before.

Michael:
[37:48] I remember when I was going to make it page 10 sometimes.

Sherard:
[37:52] Yeah, exactly. But we’ve gotten so comfortable. And then I look at my kids. My kids have grown up with the Alexas in their room, right? And they’ve grown up with Gemini on their phones. So now, look, using Google is so antiquated for them. Like they don’t even need to see page two, three, four, five. They just, you know, talk to an AI bot and it gives them a response. So I think part of this, the reason I’ve mentioned this is I think part of this is just like where we are in terms of adopting technology, the more you grow up with it as an inherent part of your ecosystem, the more you’re just going to be adept to trusting it. Like for those of us who grew up with the internet, we never really doubted it. But, you know, maybe our parents, maybe our grandparents, they didn’t quite know what to think of it. Right. And they didn’t know what it could actually be used for. But we just grew up, you know, having an AltaVista account and, you know, creating all these websites, you know, GeoCities websites. You know, that was just what we did. That was our jam.

Ryan:
[38:49] Bring them back. Bring them back.

Michael:
[38:50] And the times where you debate which one, GeoCities or AngelFire, which one do you go with?

Ryan:
[38:54] Yes.

Sherard:
[38:55] Exactly. Yeah. How far can I go with this free account? Right. So, but, you know, so like to me. You know, the part about this that’s critical and what I talk to customers and users of AI about is we’re at the inflection point. And I think this year is a critical year where we need to be able to have confidence in AI. And I think companies who are looking at deploying AI, what they’re really looking at is they’re not looking to just deploy AI. They need to have confidence in what that AI is solving for. And so this is where I think a lot has to happen around security. We need to know that these models don’t have vulnerabilities. And I always equate it to CVEs, container vulnerabilities and the container scans that Red Hat does today. We do a lot of scans to make sure that containers are safe to use, make sure that container images are safe to use for open source communities.

Sherard:
[39:49] We do this in the upstream projects. We have security scans as well as some of the downstreams. That hasn’t really replicated itself when it comes to models. And so there has to be a lot of effort to see, is this model safe to use? Are there known backdoor vulnerabilities? Are there known security gaps and leaks in this? So I think that’s one of the things that the industry really has to address. The second thing it has to address, it has to be these open source guardrails, these open source models used for guardrails, they have to get more accurate.

Sherard:
[40:19] Right now, the proprietary ones are still leading the charge and there’s still a level of effort that has to happen for the open source communities to get a little bit better there. But I think that’s naturally going to happen. If you look back a year or maybe even like two years ago, it was heavily proprietary models were destroying open source models, right? It wasn’t even close on the performance. It wasn’t even close on the accuracy. Now you look at where we are, the script has been flipped, right? The open source models are heavily in favor with a lot of industries, with a lot of users. You can now go to Hugging Face. The timing on that was perfect. And you can grab a model and do something meaningful, actually solve a real problem. So I think the guardrails part of this and the AI safety part is just a little bit lagging on the open source side. It’s still led by proprietary. But I think, honestly, once you get to that level of confidence with the security, with the guardrails, I think people’s hesitation…

Sherard:
[41:21] Is going to, that’s going to go down quite a bit. Now you’ll still have that generational divide, right? You know, I could still talk to my grandmother and she still doesn’t trust a computer. That’s okay, grandma. That’s okay. And I think our kids are going to be looking at it like, man, you still don’t trust AI. Yeah. It’s just, it’s just generational.

Michael:
[41:38] Well, I think it’s no kids. We don’t trust the companies behind the AI.

Ryan:
[41:42] Some of the companies. Yeah, for sure. You know, it’s interesting because I remember in the workforce when the browser, the internet became a big thing. I’m that old. And then I remember people thinking I was a wizard because of the way I could utilize the prompts in a Google search and find the things, the answers to the questions. It blew their mind. And I think the same thing I tell people with this whole thing with AI is that you really got to, even if you don’t like it, Even if you don’t trust it, you’ve got to learn it and understand it and know it because I agree with you. This is going to be a part of our future, whether you like it or not.

Michael:
[42:21] And the genie is out of the bottle.

Ryan:
[42:23] The genie is out of the bottle.

Sherard:
[42:25] It’s not going out of the bag. And I like a lot of things you want to put in.

Ryan:
[42:29] Yeah. I like a lot of what AI is bringing. You know, there’s amazing things that this is going to be able to do. There’s amazing things that’s already done. And I agree with you. I think the real question is, what are the guardrails we’re going to put in place? How are we going to protect people’s jobs or get people moved into other industries? And these are real problems that we’re going to have to solve. But I’m so happy that open source is not behind and that we’re not sitting there fighting to catch up. And, you know, we’ve got 20 years before we ever get to the, we’re not there anymore. We are now on the cutting edge with these other models. You know, you’ve got DeepSeek and Kimmy and all these other things too that are coming out that are just like blowing people’s minds. And so there’s just so much to be excited for. But you’re right. We’ve got to be careful in how fast some of this stuff are going. How responsible we roll some of the stuff out, but it’s very cool.

Sherard:
[43:21] Yeah. And I was thinking about this earlier today. Some people ask me, you know, why does Red Hat care about open source and AI and why are we so committed to the game? And I think about, like, if you look about, you know, what, 30 years ago, when Red Hat first got into the Linux game, it was like, you know, we were fighting for something. It’s like, oh man, we’re going to battle. It’s like, we’re William Wallace, right? And it’s like, you’ll never take our freedom. We’re gonna go fight i’m i actually have a trip to scotland next week and so i’m like getting my mind ready for that yeah we have like back in the days you have war paint like linux yeah we’re gonna go take on the world and then linux one yes oh crap what do we do now we’re like we’re fighting for open.

Michael:
[44:08] Source and then now open source one you’re like.

Sherard:
[44:10] What’s next what are we doing and so i was like oh wait this ai this ai is proprietary no that’s not the only that’s not the reason why we got in it, but going back to what you said before, Ryan, about timing of things, the timing is right. There was a huge lean to proprietary AI. This is when, especially with the foundation models that had come out, it was always defaulting to proprietary foundation models and proprietary applications that leveraged it. And that was another opportunity for us to see another linux moment this is going to be this is going to be such a pivot this is going to be such a transformation it has to be open source and our you know our leadership uh our engineers we felt that in our hearts we cannot let proprietary be the way the default way that ai is consumed it has to be led by open source open source and that goes for the models as well as the technology.

Sherard:
[45:06] That’s backing that. And what happens in a world where closed source proprietary AI wins out? And then starting to look at who’s the best company to help fight this battle. It became obvious why we needed to really step into the space. And at least from an open source perspective, we have so many passionate people who work on open source on a daily basis. Man, they were ready to go. You know, they saw this as, you know, OK, let’s put the war paint back on. All right, William Wallace, yeah, you’ll never take our AI freedom. Just, you know, going after this. And it’s been it’s been fantastic. Now, I don’t know what happens when open source AI wins out. I’m probably like, all right, quantum. Ah, you’ll never take on quantum free time. We’ll see.

Ryan:
[45:54] It is really important. We talk about the digital divide. During COVID, there are statistics out there of how many kids had no access. I mean, there’s people right now in the United States who have no real access to the internet. Think about how difficult it is for them to do their homework. Now you look at AI. Again, another thing, if people say they don’t, they’re probably lying. People are using AI for their college work. They’re using it not to just pirate or make stuff for them, but they’re using it to enhance their work to levels that you couldn’t do before AI easily. That would take hours of research now you can do in minutes with AI. If kids don’t have access to that because it’s behind a paywall they can’t afford, Then now they are again, we have a whole nother generation of people who are going to be behind on it. So Red Hat and other companies making this open source is vital that we don’t go backwards back again, more backwards than we’ve been before. We’re still backwards with the internet. That’s still a big problem, but access to the internet, but even those who have the access, they don’t need to have an additional paywall just to be able to utilize something. So they’re not behind all the other kids that have that.

Michael:
[47:01] There’s so many things that are related to that kind of situation.

Michael:
[47:05] But we’ve talked about some cool projects you’ve been involved in. And I was wondering, are there any cool projects right now? Like some big projects that you can talk about that you’re currently working on?

Ryan:
[47:15] Secrets. We want secrets.

Michael:
[47:17] Yeah.

Sherard:
[47:17] Secrets. Secrets. Let’s see.

Michael:
[47:20] Reveal the curtain.

Sherard:
[47:21] Yeah. So I’m trying to think. Let’s see. I’m trying to think.

Ryan:
[47:27] Without getting into too much trouble.

Sherard:
[47:29] Yeah.

Michael:
[47:30] Without getting into too much trouble.

Ryan:
[47:31] We want them to get in trouble, just not too much trouble, just a little bit of trouble.

Michael:
[47:35] Fair enough.

Sherard:
[47:36] So I’m going to go back to some of the things that we talked about at Red Hat Summit.

Sherard:
[47:42] One of the coolest things that I think we’re looking at, and you guys have talked about this on one of your podcasts, is bringing AI to the IT admin, right? And giving that person a supercharge, right? You know, being able, look, these guys, man, they’re working to the bone. They’ve got thousands of nodes multiple clusters to maintain it’s getting increasingly complex it’s not enough to have playbook or automation to run these things they need help and so like some of the cool stuff that you know i’ve been involved in and some of the a lot of my peers have been involved in is is the light speed portfolio and just looking at ways in which you can not just take um you know kubernetes not just take uh ansible not just take linux but like looking at all of those things. So we’re also looking at that from a, this is what’s interesting to me and some of the things that we’re working on and we’re starting to look at is the things that you guys have talked about in terms of rail light speed. And, you know, for those of you who haven’t listened to that episode yet, I’d highly advise you to. Uh, uh, uh, but, you know, part of what was mentioned there is just being able to do things like, uh, talk in plain English or plain language to your Linux console and it go do interesting things for you. Now we’re starting to look at ways in which we could apply that to an AI platform. And now you have AI for AI, right? And so like, this is getting meta guys. So stay with me, stay with me.

Ryan:
[49:09] Yeah. Inception here.

Sherard:
[49:10] And so you think about all of the fragmentation when it comes to runtimes for this. You have VLLM, you have SGLang, you have OpenVINO, you have Triton, you have all these things. And then you start to add in the complexity about the infrastructure you’re deploying into you. Maybe you have AMD, maybe you have NVIDIA, whatever that is. And then you have the open source models or the closed source models. You have Llama, you have Mixtral, the permutations and combinations of being able to find the most accurate thing, the thing that’s going to solve your problem. So let’s say you’re doing computer vision. Let’s say you’re doing a chatbot, whatever that use case is.

Sherard:
[49:52] The trial and error part of this, it’s pretty high up there, right? The permutations are just insane. And so, some of the things we’re looking at is how do we give you a platform where I kind of equate it to like the I feel lucky button in Google where it’s like, you know what, I’ve got a use case and I feel lucky and just have AI figure out what’s the best way to leverage your infrastructure, what’s the best way to leverage your models, give you the cost evaluation and say, okay, this is the most effective, the least expensive way to roll out this model in terms of what you’re doing. So some cool stuff going on in that space. The other thing that we’re doing is really looking at how do we help with the infrastructure side for customers and users of these platforms, Kubernetes in particular, to be able to get access to different types of hardware. Because if you look at the struggles right now, a lot of what’s in these drivers, a lot of what’s in the Linux kernel, it’s geared towards certain distributions. And so we’ve done things with image mode for RHEL and working with our partners to make it easier to kind of hot swap different infrastructure or different hardware GPUs without having to completely rebuild your product, your application, your container images, making it super easy. So, you know, doing some stuff in that space personally, actually not personally, but another cool project, this is totally different.

Sherard:
[51:17] But another cool project that we’re doing is we’ve actually been working on, you know, a lot of times we also try to do socially good projects. And one of the things that we’re trying to do is to help cultures use AI to not just record their history, but also see ways in which AI could be used to enrich and enhance stories. And so, you know, one of the things that we kind of, you know, a buddy of mine, he started this project. It’s called Griot and Grits. he started this project and then he called me and got involved and I started getting Red Hat more involved and we’re working with a various number of institutes on this as well is to be able to take people’s oral history you know things that maybe you recorded your grandfather maybe you recorded your uncle talking about your family history and then giving them access to AI so Ryan this goes back to allowing people free access to technology giving them access to AI to be able to take those recordings of their family history use AI to even enrich and enhance the story of someone’s talking about maybe the Korean War, then use AI and generative AI to show what the Korean War may have looked like, provide context for people who may not understand the impact of that, and being able to tell a more complete story that their family could then carry on for generations. So, Red Hat’s heavily involved in that. Got a lot of cool stuff that we’re working on.

Ryan:
[52:36] You know, what I love about this is…

Michael:
[52:38] That’s an understatement.

Ryan:
[52:40] Advice I give a lot of people who are dealing with a loved one, like a father or grandfather thing that’s passing away is record an interview with them, right?

Sherard:
[52:49] Yeah.

Ryan:
[52:49] Record an interview with them and you’ll be fascinated at the stories people will be willing to tell and you’ll learn from people. Now you could take those stories, right? And integrate some of that into AI to enhance that with visuals and things to add to it so that their story lives on forever. That’s so beautiful. I love it. I want it now. When can I have it? That’s amazing.

Sherard:
[53:09] Yeah, we’re currently working on it. It’s super cool that the team, you know, we had a great group of volunteers from Red Hat, from various companies that have been working on this project. Really, really cool. And guess what? You know, it’s an open source project. My buddy of mine, he doesn’t have a background in open source. He’s like, dude, man, I’m not sure about this. Like I can make a lot of money. I was like, no, this is something it’s got to be open source. Like these are the types of things that’s, you know, if you want my involvement, that’s the that’s that’s that’s the line in the sand. Like this has to be open source. It has to be accessible to everyone. So, you know, super excited about that.

Jill:
[53:46] Awesome. And Sherard, we’ve talked about the AI guardrails and why it’s so important that they need to be open source. So how do you see Linux’s role in the AI revolution? And what makes Linux particularly well suited for AI development and deployment?

Sherard:
[54:08] Yeah. Oh, Jill, this is a, this is a good question. This is a good question. Um, this goes back to my, my William, and I know you guys are movie fans. So that’s why I threw that in.

Ryan:
[54:19] I’m a huge Braveheart fan, so you’re just speaking right to my heart right here.

Sherard:
[54:21] This goes back to my William Wallace reference. Like, what’s interesting is Linux is at the foundation of AI innovation, right? Like, who would have thought 20, 30 years ago? Who would have, like, honestly, guys, who would have thought that Linux would be at the heart of AI? And you look at all of the major vendors, you know, what they’re doing, you know, a lot. And almost all of it is based on Linux, right? And why is that? It’s the most secure platform for enterprises. You can’t argue that. It is highly secure and it’s highly secure because it’s open source. It’s making itself available for scrutiny. And so when you start to look at the layers here, and this is where I think Linux has the opportunity, I see a world in which we take that exact same playbook. AI is not necessarily all that secure yet for the reasons I mentioned before. And we have to take that exact same approach. And so, you know, last time, Michael and Ryan, when we talked.

Sherard:
[55:24] I equated what’s going on with AI and what’s going on, especially in VLLM, as the new kernel, right? And so, you look at what VLLM is doing and what that community is all about. It’s about giving the best performance possible and the best access possible to the hardware that it’s sitting on, right? It sounds like a very familiar story, right? You have GPUs, you have memory, you have things that have to be orchestrated and optimized from an AI model perspective to get access to that hardware. Very similar to what was happening in the Linux kernel. And so when you look at that playbook, everything that had to make the Linux kernel win out, we’re seeing this repeat over, history is repeating itself. And now when you look at something like VLLM, which is very, you know, that is Linux, right? You look at VLLM and what it’s built on top of, it’s a Linux container at the end of the day. And now we need to take it one step above. Like, what are the things that we have to add to that VLLM container, that AI, that new AI kernel that we’re talking about to make it secure, to make it safe to use, to make it easily digestible? So I think it’s a very, very, it’s not just the fact that Linux is at the foundation of all of this AI innovation, and it’s the default operating system that people are going to because they trust it in their enterprise environments. But it’s also, in my opinion, it’s absolutely.

Sherard:
[56:50] You can take that analogy of what happened with Linux and the kernel and how it won out and you can directly correlate that to what’s going on with the Linux kernel, the AI kernel for what we’re doing today. Yeah.

Jill:
[57:04] Oh, awesome. And I’ve always thought about it. And also, because with Linux, you have not only the community input, but that system of checks and balances. And that’s one of the reasons that Linux has taken off. So, so well, and, you know, is we rule the world, really, in infrastructure.

Sherard:
[57:27] Yeah, gosh, it’s not easy being king, right?

Jill:
[57:30] Yeah.

Ryan:
[57:32] I love it.

Sherard:
[57:33] It’s not easy being king.

Michael:
[57:34] Love it.

Sherard:
[57:34] But no, you’re 100% right, Jill.

Ryan:
[57:37] Sherard, let me ask you this. You know, Lightspeed, obviously Red Hat has, for people wanting to play with Lightspeed, has some ability to do some free licenses and things that you can get. But there’s Fedora out there. Are we going to see Lightspeed in Fedora at some point?

Sherard:
[57:54] Oh, that’s interesting.

Ryan:
[57:55] I’m putting you on the spot. I know.

Sherard:
[57:56] Oh, man. Let me call my buddies there. I would hope so. So I will bring that back to the team, Ryan. I think that’s a very, very valid point. I would have to, well, I’m not going to say anything, but I will certainly bring that back to the team.

Ryan:
[58:12] That’s enough for me to know that it will be a discussion point. We brought it up as well to the folks at the ballgame that worked on Fedora specifically. I know they were very excited about the possibility. And here’s why I think it could be really important. I’m just going to make my little sales pitch here. Is today you’ve got a bunch of kids who go on to windows and they get familiar with bing right and the in the chat and they don’t even maybe even know that chat gpt is what’s powering that behind the scenes they’re looking at microsoft and they’ve got this virtual assistant and they click that button and it pops up and they ask it questions and as you said that’s their new search engine and we don’t really have that in linux right now so you know the the new student gets into linux they install fedora what an amazing experience that is today we have so many new listeners. How many new listeners did we get recently, Michael? 33% increase?

Michael:
[59:01] 33% increase, but we basically got subscribers to the point of like, we… Doubled our subscriber base.

Ryan:
[59:08] But what’s interesting about this doubling, Sherard, is this is a lot of young people. We have a lot of young people who are coming in. And so, you know, they’re trying Linux for the first time. I want them to have that button. I want them to have that experience there.

Sherard:
[59:22] That wow moment.

Ryan:
[59:23] That wow moment where they can, they’re not looking at, well, when I’m Windows, I’ve got access directly to click a button and I could start messing with AI. Whereas, you know, Linux, I’ve got to sit there and try to find an app to install. And there’s stuff, there’s apps you can install to use some of the open…

Sherard:
[59:36] But that’s not the default experience.

Ryan:
[59:38] They’re not the default experience.

Sherard:
[59:40] There you go, short.

Ryan:
[59:40] That’s my sales pitch from there.

Sherard:
[59:42] I love it.

Ryan:
[59:42] We’ll move on. All right.

Sherard:
[59:44] I love it.

Ryan:
[59:44] For those interested in working in the field of open source, what are some practical starting points that would help them land a job? And this is important because we have a weird economy. We’ll just go with the word weird. And we have a weird job market. What would be the advice you would give some folks?

Michael:
[1:00:01] That’s the most understatement of the year.

Ryan:
[1:00:04] We’ll just go with weird. Yeah.

Sherard:
[1:00:05] Yeah. It’s, you all know this, it’s highly competitive, right? You know, we have companies, you know, some of the top AI companies in the world doing layoffs, which means there’s a saturation of talent out there. That’s just, that’s just the reality, right? I don’t want to ignore the reality, but it’s also never been easier. Like I said before, we grew up in a time where you had to have an MSDN license, you know, legally to be able to do some things. So it’s never been easier for people to kind of roll up their sleeves and get access to this technology. I would say number one thing has to be the correct mindset.

Sherard:
[1:00:45] You have to know, you have to be confident in your skills and you have to be ready to just contribute, right? And there’s a level of vulnerability when you get in an open source community. You’re naturally opening yourself up to be judged. And that’s not an easy thing for anybody who hasn’t done this before. So if you’re getting involved in an open source community, I think that’s the best way to get your career started. I said this before, the GitHub repo is your new resume. And the more you can show that you’ve actually done meaningful contributions. In fact, a lot of times, my teams, we’re looking for contributors of communities to bring into Red Hat. That’s one of the primary things that we see is like, hey, if this person is a really strong contributor in a space that we care about, that’s a great opportunity for us to bring them in and have employment with Red Hat. And so I think there’s a level of not being afraid to contribute.

Sherard:
[1:01:43] Yes, it’s going to make you vulnerable, but it’s also going to give you the ability to dive right into a technology. It’s going to give you the ability to collaborate and most importantly, build relationships with the open source communities out there. There’s a ton of them available. I mentioned it before, like VLLM, SGLang, you have what’s going on with Kubeflow, MLflow, all of these technologies. It’s fantastic to get even Linux. There’s a lot of stuff that’s going on with Linux on the kernel side to optimize AI workloads running on the infrastructure. And then, oh, let’s not forget PyTorch. PyTorch is phenomenal. If you really love that low level development and you really want to get your hands, you know, your feet dirty, whatever the term is, you know, get your hands dirty and I don’t know, whatever.

Michael:
[1:02:31] Feet wet.

Sherard:
[1:02:32] Yeah, feet wet, whatever. Yeah, I don’t know how I got on hands and feet. But anyway, you know, working in a community like PyTorch, it’s great. And so I think that’s the number one thing I would say is don’t be afraid. You’re going to get judged, but the community, you can find the right community that’s very supportive, get your foot in the door there. And then, like I said, it’s highly competitive. You may have to look at startups. You may have to look at a smaller company that’s maybe series A, series B, looking to grow. You come in as a skilled developer and it may not be as competitive, but everybody’s looking at AI now, right? The one benefit getting into the industry, if you’re focusing on AI or if you’re focusing on learning AI, it’s never been the right time because this is probably one of the few times in technological history where every company from every domain and from every background is looking at this with scrutiny. So it’s a great opportunity to get in the door. And I think what we alluded to before, The cat’s out of the bag. What did you say, Michael? How’d you phrase it?

Michael:
[1:03:38] The genie’s out of the bottle.

Sherard:
[1:03:39] The genie’s out of the bottle. There’s kind of no going back. So embrace this. Learn it. There’s so many tools. You can use Claude. You can use Cursor. You can use Gemini. A lot of free tools to play around with to see how you could leverage AI in your job. We have teams. I just got a demo of this today where my team used Claude to be able to create a UI mock-up and it dramatically accelerated their ability to even see if a new feature they were working on was even plausible. And within an hour, where it would have taken them days or even weeks to see if a new feature was even something that you could even do, they were able to showcase this within a couple of hours. And this is the power of AI. So I think, you know, the more we embrace it and the more people just really just, you know, or you have a mindset of curiosity around it, I think you can find that path to success.

Ryan:
[1:04:37] Love it. Great advice.

Michael:
[1:04:39] Yeah, that’s awesome. And speaking of advice, I’m very curious. On your LinkedIn profile, you mentioned you had a passion for teaching underserved youth about careers in software technology or software industry more specifically. And I was curious because our audience, we talked about it. We have a lot of younger people in our audience and they’re all eager for guidance. So what’s your top tip for them to break into the software industry career?

Sherard:
[1:05:03] Oh, be persistent. Be persistent. This is not an industry for a timid mind. All right. You know, you have to leverage your connections.

Sherard:
[1:05:14] Back when I was younger, me, I have an older brother. He’s about four and a half years older and super, super talented. Um uh he’s been in the film industry and uh you know does a lot of stuff uh and before he kind of got into the film industry he would always tell me his talent would get him where he needed to go and i was like dude no it’s connections as well you can be the most talented person in the world but if you don’t build those connections if you don’t work on those connections and sometimes those connections it takes years for them to mature into something meaningful uh you know That’s a key part of this. And so a lot of what I try to tell people is don’t just focus on your skills, focus on your networking. Go to these conferences. There’s the open source conferences. There’s Kubeflow. You have to invest in your network just as much as you’re investing in technology. And I even, when working with young kids, it’s kind of the same thing. I just talked to a group of middle school students who were doing some AI projects at a STEM camp that one of my friends runs. And you know you know it was the same thing it’s like look.

Sherard:
[1:06:21] You guys are so talented. And there were some high school students in there as well. And they did a project. I said, look, I expect each one of you to contact me for an internship opportunity. I’m leaving it up to you. I’m not going to offer you anything, but I expect you to come to me. And so we have to put it in the mindset of these young people that these opportunities aren’t just going to drop on your doorstep. You got to work for it. You actually have to go out there. You have to be persistent.

Michael:
[1:06:45] You hear that, Ryan?

Ryan:
[1:06:48] Listen, I think that’s amazing advice. And I also think the power of open source for the young folks, and you mentioned this earlier, is that you can start contributing to a project working with some of the greatest engineers and software coders in the world. You can literally write side by side with them without an interview, without anything. You can go contribute to an open source project right now. Like you said, your GitHub becomes your resume, and that’s how you can build your network with these individuals, and you can make amazing connections there. And it can start off with the littlest thing. You helped put a semicolon somewhere it was missing. You helped with some documentation, and it can turn into something so much bigger. Because where else can you go for free and work with the greatest writers of code in the world for nothing?

Michael:
[1:07:36] Actually what other type of industry could you find the best of the best and then work alongside them and there’ll be receptive and you don’t even have to go through like a process you just send them your work and they look at it and if it’s good they keep it and you know or they might give you assistance and all that sort of stuff like name another industry that has anything like that yeah.

Sherard:
[1:07:59] You you pretty much can’t and and we have you know you’ve probably heard the phrase especially since you guys are open source guys, you got to be willing to chop wood and carry water. You got to be willing to, a lot of times communities, for example, I was just in a community meeting earlier today where it was all around, how can we help with the CICD, the continuous integration, continuous deployment part of LLMD. LLMD is a new distributed VLM inference engine that we’re working on. So you look at VLLM and being able to distribute to thousands of nodes. And one of the things that community is struggling with is the chop wood carry water thing. Like how do you know, nobody wants to really focus on CICD, but it’s a great opportunity for someone to get in. Like you said, Ryan, focus on documentation. It’s the things that the developers, it’s, you know, it’s maybe as you get more senior in your career, it’s like, ah, man, I really don’t feel like writing this documentation. I really don’t feel like writing the CICD automation. That’s a great opportunity to step in because that’ll get you connected with people. And guess what? They’re going to be hugely thankful that someone chopped that wood and carried that water for that community.

Ryan:
[1:09:11] Yeah. I love it.

Jill:
[1:09:12] And Sherard, I was going to tell you, I am a 30-year-old retired professor. Oh, awesome. Yeah. Computer animation and motion graphics and game design. And I introduced open source to my college.

Sherard:
[1:09:27] Linux, of course. That is amazing. That is amazing.

Jill:
[1:09:30] And so what you said resonated with me because that’s what I teach my students. And also you have to have a really good work ethic and have passion for what you want to do. Yes.

Sherard:
[1:09:42] That is amazing. That is amazing. So you’ll appreciate this. My brother, he used to work for Industrial Light and Magic. So that’s, that was, you know, we went to NC State together and straight out of NC State, he started working for Industrial Light and Magic doing special effects, visual effects. So Star Wars, Star Trek, Chronicles of Narnia, Transformers, Harry Potter, you name it, the Avengers, the Eternals, he’s done it all.

Jill:
[1:10:12] Very cool. I’ve worked for many studios as well.

Sherard:
[1:10:16] That is amazing.

Jill:
[1:10:17] Yeah.

Sherard:
[1:10:17] That is very cool.

Ryan:
[1:10:18] So, Sherard, that was the Kitty Glove interview. Okay.

Michael:
[1:10:23] Yep.

Ryan:
[1:10:23] So, what you may not know if you’ve not read up on our interviews is we also have something called the lightning round. It’s called the lightning round.

Michael:
[1:10:32] It’s intense.

Ryan:
[1:10:33] Yeah. The lightning round is where the Kitty Gloves come off and the real questions begin. This is where your career will either be made or it will sink. Everything you’ve worked for comes down.

Sherard:
[1:10:44] Am I going to be on the register after this?

Ryan:
[1:10:46] Yeah, this is big time. So the key here is we’re going to ask a question and it’s up to you to answer it as fast as you can. First thing that comes to your mind in there. And there’s lots of wrong answers. So you can get them wrong 100% here. And I’m going to take the first question and then Jill and Michael and we’ll go in that rotation there. So the first question, are you ready for the lightning round to begin?

Sherard:
[1:11:12] Oh, I am not, but let’s go.

Ryan:
[1:11:14] All right. What movie represents the future of AI the best? Terminator, 2001, A Space Odyssey, or WALL-E?

Sherard:
[1:11:22] WALL-E.

Ryan:
[1:11:23] Good.

Jill:
[1:11:25] What was the first programming language you learned, and do you still use it today?

Sherard:
[1:11:31] Basic. No.

Michael:
[1:11:36] So what’s a hobby that you do that has nothing to do with technology?

Sherard:
[1:11:40] I write music.

Ryan:
[1:11:42] Nice. Your favorite movie?

Sherard:
[1:11:46] Oh, my favorite movie. Indiana Jones and the Last Crusade.

Ryan:
[1:11:50] Very good.

Michael:
[1:11:50] Nice.

Jill:
[1:11:52] Hard Cheetos or Cheeto Puffs?

Sherard:
[1:11:56] You know, you guys talked about this in a previous episode. Are these outside of their minds? It’s got to be white cheddar cheese puffs, like the Cheeto Puffs.

Ryan:
[1:12:07] The white cheddar. Not just any white cheddar.

Sherard:
[1:12:09] But the simply white cheddar, right?

Jill:
[1:12:12] I like the organic ones.

Ryan:
[1:12:14] He just threw the whole question for a loop. I don’t even know what to do now with that.

Michael:
[1:12:18] I’ve never even had those. I don’t know.

Jill:
[1:12:20] Oh, they’re delicious.

Michael:
[1:12:20] I don’t know.

Ryan:
[1:12:21] I just got a kernel panic. I don’t know what to do.

Sherard:
[1:12:24] Get yourself some simply Cheetos that are the white cheddar puffs. You thank me later. You’re going to be talking about this on a future episode.

Ryan:
[1:12:31] All right.

Michael:
[1:12:32] We’re going to have to check those out and come back to you on the next episode.

Jill:
[1:12:36] Those are my favorite. Because I like White Shed or anything.

Ryan:
[1:12:40] Well, you’ve been holding us back. You’ve been acting like you’re a hard-to-do person. And all of a sudden, we find out he’s a secret puff person.

Jill:
[1:12:46] What is that?

Sherard:
[1:12:47] Hey, man.

Ryan:
[1:12:48] Me and jill me and jill unbelievable well gerard your interview absolutely incredible to have you on the show i knew this would be amazing and uh you did not disappoint we knew it’d be amazing from the the game that we were at and we had amazing conversations in here i think so many people are going to be inspired by this thank you so much for coming on the show and we have to have it’s not a want we have to have you back on the show yeah oh yeah future if you’d be even if.

Michael:
[1:13:15] You You don’t want to come. You have to. Of course.

Sherard:
[1:13:17] No, this is a blast. You guys got me out of my office doing something. This is exciting. This is a great break to my frenetic day. I can’t thank you enough for stopping me as I was going to that party and say, hey, can we talk for a bit? It’s like, I’ll be late for a party, but man, I’ll chat with these guys. This has been a blast. You guys are phenomenal. You’re great people. You have a great audience. I love what you do. The topics you talk about are just so critical. So thank you for what you do. And thanks for inviting me.

Ryan:
[1:13:49] Thanks, Gerard. We’re going to have you on. We’re going to have you back real soon on this.

Jill:
[1:13:53] Yeah.

Sherard:
[1:13:53] Let’s do it. Let’s do it.

Michael:
[1:13:55] Can’t wait.

Jill:
[1:13:56] Let’s make it monthly.

Ryan:
[1:13:58] Joe’s like booking calendars already.

Jill:
[1:14:01] I’m going to schedule them.

Michael:
[1:14:03] I don’t even just booking calendars. Like putting it as a perpetual, it’s going to happen over and forever.

Jill:
[1:14:08] Yeah.

Ryan:
[1:14:09] Love it. So I don’t know about the rest of you, but that is going to go down in the books as one of my favorite interviews of all time. Gerard, thank you for coming on the show. And a huge thank you to each and every one of you for supporting us by listening or watching, however you do it. We love your faces. And you can come join us in our Discord and continue the conversation, all of the amazing topics that we brought up by going to tuxdigital.com/discord. And there you can make tons of open source friends. and we talked about networking, that’s where you can start right there in our Discord, start your networking journey. Right now.

Michael:
[1:14:45] That’s right. And also you can check out our forum at tuxdigital.com/forum to have deeper conversations and, you know, maybe if you need some tech support or want to help others with some tech support, check that out as well. And if you want to support the show, you can go to tuxdigital.com/membership and you can become a patron. You get access to so many cool perks like unedited episodes of the show, being able to watch the show live, getting merch discounts for the store and oh you didn’t know about the store well go to tuxdigital.com/store and you can get all the cool stuff we have we have so much great stuff we have uh stuff that’s really cool for jill there’s some stuff that is like related to what things i’ve said there’s also the ryan’s okay i guess stuff ryan’s okay uh it’s fantastic so if you’re curious or if If you want to go in solidarity with me and tell the world Ryan is okay, I guess.

Ryan:
[1:15:41] You know what, Michael? That’s our top selling piece of merchandise right now. And I want to tell you, it’s a great troll. Congratulations. Payback will be bad. Payback will be bad.

Michael:
[1:15:53] Wait, wait, wait, wait, wait. This is my payback to you for the weak thumbs and all the other stuff.

Ryan:
[1:15:59] But listen.

Michael:
[1:16:00] The Michael AI stuff and all that.

Ryan:
[1:16:02] To take the top-selling merch is Ryan is OK, I guess, in giant letters. You win right now. You won the battle, but not the war. It will continue.

Michael:
[1:16:12] Oh, we’ll see. We’ll see about that.

Jill:
[1:16:15] And make sure to check out all the other cool shows here on Text Digital. That’s right. We have an entire network of shows to fill your whole week with geeky goodness. Head to textdigital.com to keep those Linux penguins marching. And everybody have a great week and remember that the journey itself is just as important as the destination.

Ryan:
[1:16:37] Thanks everyone love you all man we couldn’t have people like charard on the show if we didn’t have so many great listeners you know to the show for real that’s true that’s true so thank you so much support this show and have shown it so much love over the years and the fact that keeps growing is why we can have amazing guests like that like what an interview, what an interview I mean it’s such.

Michael:
[1:17:02] It was so awesome to have him on and also thank you for everybody who’s a part of the community and we’ll see you next week.

Ryan:
[1:17:09] Next week everyone.