Conquering Fear and Understanding AI's Potential with George Spafford, Vice President Analyst at Gartner

(Josh at 00:00:00) Today, we're talking to George, Vice President Analyst at Gartner, about the real world value of AI and what George learned from living abroad. You're listening to Joel Beasley, Modern CTO.

(Joel Beasley at 00:00:18) What's happening, George?

(George at 00:00:20) Hey, Joel. I'm doing good, sir. How are you?

(Joel Beasley at 00:00:22) My hair is crazy because it's raining here. So going back and forth between my building, I'm getting soaked.

(George at 00:00:30) Cool. You know, Josh and I were talking. He needs a raise. I don't know what you're paying him, but you gotta bump it up at least 20%.

(Joel Beasley at 00:00:37) Josh needs a raise?

(George at 00:00:39) Josh needs a raise.

(Joel Beasley at 00:00:40) We just did raises, like, last month or two months ago. Something soon. Yeah.

(George at 00:00:45) He's turning red, Josh. Josh, man. I'm teasing you.

(Josh at 00:00:48) I had nothing to do with any of this.

(Joel Beasley at 00:00:51) Yeah. This is funny. I like it. I am very comfortable talking about money. Let's talk about money.

(Joel Beasley at 00:00:58) Some people aren't though, George. I've noticed that I'll be in a group. I have kids, small kids, so we'll be out at events and stuff. And I realize that I kind of have to read the room a little bit better when talking about money with the different crowds because some crowds are, like, not about it. It's like, quiet. Money's a thing we don't talk about, and so I'm learning.

(George at 00:01:18) Well, it's interesting. You look at cultures. My wife's Filipino. And Philippines, you'd walk up to somebody. Hey. How are you doing? How much do you make? You know what I mean? It's very early on in the conversation because they're ranking, you know, in terms of where you're placed. So a lot of Filipinos, when they come to the US, they gotta learn, alright. We don't ask that question. We don't make jokes about somebody's weight or ask how heavy they are or something like that. You know? So there are these cultural differences.

(Joel Beasley at 00:01:44) Wait. They do that in the Philippines? That's awesome. I wanna go visit.

(George at 00:01:46) Yeah. Yeah. Yeah. Oh, well, I went to school in Japan when I was in college. And so I went back after, like, twenty years and saw my host mother. She looks at me up and down. She goes, you've not aged at all, but you're fat.

(Joel Beasley at 00:02:02) That's so funny. Well, this is it, George. We hang out and we talk. I wanna start really by talking about, you know, you went to Japan. You had a host family. You've traveled. I think you've said you have a wife from the Philippines. You tend to have lots of exposure to different cultures and travel. How did that get started? At what age did that start for you?

(George at 00:02:21) Well, you know, back in probably the eighties or something like that, that's when the Japan fad hit the United States, you know, and you had all these TV shows, completely fictitious, outlandish stuff, you know, about samurai, about ninjas, about all this stuff. I'm like, wow, that's really different. So when I went to Michigan State University, I enrolled in Asian studies and it's like, wow, okay, this is completely different than anything I've ever seen. You know what I mean? At that point in my life, the architecture and everything else. So it really interested me. And I talked to my mom and dad. I said, you know, given how Japan and everybody's writing about the Japanese miracle and the Japanese economy, you know, which was very prevalent in literature and culture at that time. I said, I think it'd be good if I go and study in Japan. So I applied and I went there and it was fascinating, you know, and some of the things also is that it causes you to question or to realize stuff that you hadn't really necessarily understood or thought about when you're in your own culture. So it gave me a really good chance. And then the other thing is my host family was fantastic. You know, it's really cool. And my host father, he worked for Nippon Life Insurance and he had a number of big corporate clients. And I was very interested in going and talking to these clients about what they did with technology and how they brought people up to speed and quality and whatever, and kind of just this unfocused all over the place. Cause it's not like I'm one of those guys that said, oh, you know what? I want to do this, Joel. No, I'm like, well, I think I want to go. I don't know where I want to go. So I went in, I had a chance to talk to these companies and see what they're doing and stuff like that. And a lot of that stuff that I saw, it just kind of sat in the back and germinated along with everything else and my undergrad degree and what have you, and started kind of combining the IT and the interest in people and culture and business. It just kind of all started coming together, if that makes sense.

(Joel Beasley at 00:04:23) Yeah. And what was your first real gig?

(George at 00:04:26) Okay. So out of college, George needed a degree. Now this was, let's see. When did I graduate? '90. All right. And at that time I had two job offers, one with AT&T, one with IBM and for my graduation gift, my parents said, okay, we'll fund you to go back to Japan and see your host family and also go down and see some friends in Singapore. So I did that and I came back and, you know, this is right in the middle of really bad economic times. Both job offers had been rescinded. So my first real job out of college was working for RadioShack, trying to sell computers when they were very closed, very overpriced, whatever architecture and trying to tell them that the reason I wasn't selling this was because it was a horrible architecture just didn't work. But from there though, I went in and I worked for a printing company. And the cool thing is when you're in a two person IT shop that's growing rapidly, you got your fingers in everything, programming, pulling cable through the ceiling, wiring stuff up. And the joke was if it had a power cord in it, we were responsible for it. So I got exposed to a lot of different things. I would feel very fortunate for my boss, Jonathan Fisk, who's probably one of the most patient guys on the planet because he tolerated me in my antics and would be like, okay, George, let's go. Alright. Come over here and do this instead. Alright? And it was very cool. So that's where I kind of started out my corporate IT gig, if you will.

(Joel Beasley at 00:05:55) Nice. I like it. Everything with a power cord. Like, you're a chief electron officer.

(George at 00:06:01) Dude, I had to take part. You know those big, well, I don't know if you ever seen them. In offices, they have these big, giant industrial shredders. Well, ours was jammed. It's like, okay. Do we pay a technician to come out? No. George has screwdrivers. All right. We'll go and get George. And so I had got this big industrial shredder all set apart, going in to find out where one of those big binder clips had gone in and jammed something. You know, I take it all apart, lube it, put it back together and whatever, but all kinds of fun stories like that that you might not have encountered if you'd worked for a big siloed, you know, IT organization. But, yeah, anything with a power cord.

(Joel Beasley at 00:06:37) I'm glad you're open, and we're talking about this diverse experience set that you have. And the way you speak about it, you're proud of it and you're aware of it. And I think that there's a lot of people out there that wish they could be open about it versus, you know, jobs typically want you to be very narrow and focused and specific. But it seems like you've realized it was a superpower. You embraced it and leveraged it. Tell me more about that. So you have this awareness of yourself that you like these different things. How did you weave that into a career?

(George at 00:07:09) Okay. So part of it is just by personality. I'm notoriously nonlinear inside of Gartner. You know, if you think about the movie Up, you know, where the dog sees the squirrels, like, squirrel and everything, like, stops, that's me, you know? So I'm just intrigued by many different things. And I've been lucky because I've had a number of different careers that allowed me to build on different building blocks. So when I'd look at a career, I'd look at both number one, cultural fit. Number two, what can I learn? All right. And if it wasn't something that I was going to learn something at, I'd be like, well, I don't know if this is really a good idea. So I had a lot of jobs where I learned things. Sometimes there are good things. Sometimes there are bad things. Sometimes in hindsight where I looked at it and said, wow, I should never have done that. All right. And you really have to always take a step back and look at it and ask yourself what I'd call the four whats. Okay? What worked? What didn't work? What should I do next time? What shouldn't I do next time? And apply that to what it is that you're doing and just keep broadening yourself out. If you allow an employer to dictate your career path, I think that's gonna be kind of your demise. I think you have to say, alright. What do I wanna be when I grow up? Now bear in mind, I'm turning 56. I don't know. I thought I was gonna be at Gartner like two years, maybe three years max, get Gartner on my resume and be, hey, hasta la vista baby. But instead I'm still here thirteen years later because I work with some really cool people and I get to talk to folks like yourselves and Josh and CIOs, CTOs all over the place. I get to hear stories of what worked, what didn't work, whatever. And I'm still here because it's a really cool job. So coming back around, I think the burden is on your shoulders to say, okay, I don't wanna be one and only one thing. I want to explore. I want whatever. And then figure out how you're gonna do that, both formally and informally. Some of it may be on your employer's time. Some of it might be on your off hours. Like, I started, stopped, failed a few times, software businesses where, you know, I was writing software, like out of the study. Right. You know, and all kinds of weird stuff, you know, whatever we did or we'd write it, I'd throw it out there and see what happened. But that taught me a lot about coding and about business, etcetera. But my point is, yeah, it's on your shoulders. Don't let an employer dictate to you or limit you. You need to think about, okay, what interests me and just keep building it. If you go down a path and you're like, I don't like this path, then choose another one. But you just keep trying. You just keep learning. When you stop learning, you're in a world of hurt. Any thoughts or comments on that, Joel?

(Joel Beasley at 00:09:57) Yeah. Yeah. Yeah. So let's go to the opposite end of the spectrum of somebody who they do this naturally. Right? They bounce around a lot. So if we take one extreme and say someone who's bouncing around so much, let's use my sister. My sister.

(George at 00:10:13) Shout out to jewelry. We're so sorry.

(Joel Beasley at 00:10:15) Shout out to her. Every week, she's expressing her newfound love thing she's going to do with her life. Right? And she doesn't stay with anything. It's photography and music. It's just keeps going, and she never sticks with anything. And she's, you know, a talented person. Right? Anything she touches, she's very creative, but she just never sticks with something long enough for it to mature to anything. We saw this happening when she was 18. She's 33 now, and it hasn't stopped. And she just never, she gets an inch deep in a mile wide. If you're one of those types of people, how do you develop the discipline or the focus to, you know, at least pick your project, but then see it through?

(George at 00:11:01) Yeah. Yeah. Well, you know, I guess this is something where you have to, again, ask yourself, you know, what does it do that I wanna do? One of the things I tell my nephews and nieces is, you know, don't let a job define who you are. A job is what you do to make the money that you need to do what you want to do. So I tell your niece, look, you gotta settle on something. You know, give it a year, give it two years, whatever it may be, but view it as a way to get the money to fund the stuff you do outside of work. You wanna dabble in all kinds of things? Awesome. Rock on. But you need to have that job. You need to have that money coming in and you're gonna have trouble. And we all know that as an employer sees you jumping job to job to job to job, that's a big warning flag for the employer. You know, if they're seeing it, I think they say less than a year or so, my dad personally is going from job to job. That's a red flag. So I'd tell her, I said, look, find something. View it as a way to make the money that you want to fund what it is that you wanna do outside of work. Give it a go, give it at least a year, and then move on from there. But some people just aren't wired that way, unfortunately.

(Joel Beasley at 00:12:09) And I like what you said about the year thing. That's actually one of the very first disqualifiers we have on applicants. We're less than 20 people. We're not a big company. But when we look for sales or whatever position, if there are, I'd say half the resumes maybe are people staying no longer than a year and a month or a year and two months or something of that nature. And for me, you know, being an entrepreneur and like, it takes a year to come in and get your projects and really understand what's happening at the business. And then in your second year, you can not only do the job that you're paid for, but you can figure out how to do more to push the business forward farther because you've got your core excuse for a paycheck down, and then you can grow the business farther. So, yeah, that's what we definitely look for. And I'm unapologetic about it because they will stay a year. If you keep hiring those people, they're not gonna break their pattern.

(George at 00:13:07) Yeah. Yeah. Well, you know, what's really interesting is when I got interviewed by Gartner. First off, one of the recruiters reached out to me way back when and said, so I see you can write. Yeah. You ever heard of Gartner? Well, yeah. You ever thought about working at Gartner? Maybe. I don't know. But what was interesting was it took several months before I was hired and I talked to all these different people and Joe Baloch, who was the vice president over everything at the time or whatever the title was, Joe is a very smart guy, respected him. He's interviewing me and he's kind of pushing, you know, on areas that I had a hard time remembering to look to see if I get frazzled and stuff. But he asked, so what do you think about Gartner so far? I said, well, you folks sure take a long time to hire somebody. And he didn't miss a beat, Joel. He goes, well, we wanna make sure we hire the right people. Alright. Now I have no problem talking. I shut up at high speeds because he was absolutely right. You know, you look at whatever business we're doing, it's built on people. Having the right people is extremely important. Okay. I think sometimes we get caught up in looking at the technical skills alone and we don't judge the whole package enough. If you're trying to build a high performing team, you need people who can work on a team. You need people who are smart, who are trustworthy, who are motivated, who are self starters. You know, if you're having to hand hold them constantly, and I always call them the followers, you leave the room, all work stops. You enter the room, all work starts, you know, you leave the room, all work stops. You can't afford people like that. You need to make sure you hire the right people. And I think sometimes we're in such a rush that we just go down the technical checklist of, okay, what are your skills? But we don't think about that fit and it's so critical.

(Joel Beasley at 00:14:51) And then what do you, what's your specific role at Gartner?

(George at 00:14:54) Sure. I'm a VP Analyst in Gartner, which means I talk to people and I share stories. Now what it is, is I'm in the IT Leaders group. We have different groups in Gartner. We have a peer practitioner who goes out and designing craft studies that they're going to do.

(George at 00:15:09) And then they named the companies that they're working with and what they found out as part of the study. We have IT leaders, which is my group, which is looking at strategies and trends that are going on in the industry and whatnot. And then we also have, like, for example, Gartner for Technology Professionals is really getting down at the practitioner level. And these folks are just so smart. You know, whenever I talk to them in terms of how to bring it all together, then we also have groups like consulting, events.

(George at 00:15:36) We have advisors, but in the ITL role, I have this chance to talk to all these different people about what they're doing. And again, what works, what doesn't work. And what keeps me awake at night is how do we get people to learn? How do we get people to change? And when I talk to organizations these days, talk to leaders, their big concerns also revolve around this, along with, hey, how do we get the right people? How do we upskill them? What do we need to do? The technology can be challenging at times, but it's the people part that can be really hard.

(Joel Beasley at 00:16:07) And not to beat the AI horse to death because obviously it comes up a lot, but you have a fairly unique perspective because a lot of times I'm interviewing people that are, like, you know, the head technology person at their company, but you get to interact with a lot of different tech leaders, and you get to talk about the future a lot. So I think it's valuable to hear from you. So, obviously, people's always a hard thing. Right? We're making these tools that are compressing time, making, you know, what used to take a week take a day. How, what's the impact of that if you extend that out, like, infinitely?

(George at 00:16:43) Boy, that's, like, the hardest question you've thrown at me. The short answer is nobody really knows. You know what I mean? For the longest time, AI languished in the land of sci-fi. You know, we'd see these things like HAL 9000 or we'd see the robots in Star Wars or whatever. And it's like, oh, wow. That's amazing. But we were so far off from it. You're still to the point now where you're getting, I guess I'd call them gist or informed answers back, you know, in terms of things when you're looking at the generative AI. But when you start looking at things where the sheer volume of data and the speed and the complexity exceed what a person can crunch, for me, that's where things get really interesting, really fast.

(George at 00:17:26) One of our clients is a big payment processor. They're generating 1.4 petabytes a day of operational log data. Now I'm old enough to remember when a petabyte was a theoretical number, you know, it's like, okay, kilo, mega, giga, you know, tera, what comes next? Peta. Oh, we know, and everybody got, like, all excited. But here we are, a group generating that much data each day just of operational stuff. All right. The day in the past of somebody going through a log file line by line manually to find something is gone. All right. You don't have the time. There's just way too much data.

(George at 00:18:04) So using the machine learning, the AI capabilities to make better use of that data, to predict what we need to do, to understand the risk, to say, all right, what are the impacts going to be if we put something out there? Or, for example, to understand, you know, we've got incidents going on. What might be causing them? Making recommendations? There's a lot of interesting stuff that's going on.

(George at 00:18:31) Now I was hearing a thing the other day and it kind of gave me pause to think when you have something that's very repetitive and you can go after it, you know, it's very straightforward for AI right now, but where you have stuff that's high touch. All right. Meaning there's a lot of human element to it, at this point, there's nothing that's quite touching it. So right now I'm jaded, I'm a skeptic. I've got more gray hair than what you can see in the camera. For those of you who don't know, Joel can see me in a camera right now.

(Joel Beasley at 00:19:02) We've got a filter on him. He's got—

(George at 00:19:04) He's got an ash filter. Oh, Joel, you don't look anything like what I thought. But, uh, you take all that AI washing away. I mean, people are putting AI on everything. You know? And we've seen this before. I've been in the industry long enough that, you know, oh, hey. You know? We'll put ITSM or ITIL on it. It'll solve world peace. People will buy our software, our consulting services, or whatever. Oh, we'll put DevOps on it. It'll, they'll buy our software. They'll do our consulting services or whatever. Cloud. Oh my god. Solves world peace. And now here we are at AI. Oh, let's solve world peace. We'll do this. We'll do that. Do another thing. I look at it. I'm like, well, okay. There's some really interesting things coming out of it. There will continue to be some really interesting things coming out of it. But like any other tool, there are going to be use cases where it makes sense and use cases where it doesn't. All right? And I think time will tell how much we can really mimic the human condition and have the empathy and stuff like that. That's not there, but crunching massive amounts of data and trying to make better use of it and improving the interactions between systems and whatever. For me, that's pretty intriguing. And like every day, we have these groups in Gartner, you know, they're always posting, you know, what's going on with AI and what are they hearing about? What are they seeing, et cetera. And the evolution is stunning. Some of it's complete baloney. You just know it's smoke, mirrors, and PowerPoint. And then, you know, some of it's real, and it's amazing. You know? So we'll see. Time will tell what it can actually pull off. But I think, at least speaking as an analyst, I probably got a few more years left in the gig before I have to worry about it too much.

(Joel Beasley at 00:20:42) Oh, I 100% agree. Like, it takes, if you go back, I'll back up. I was very surprised on this show when I was talking to people that still have mainframes. Like, that mainframes are still a huge, huge deal, and they have these old technologies. I remember I had one conversation about five years ago where I think we edited out the name of it, but it was a power plant who had software that no one was alive that knew how to operate it. So they had to hire this company to come in and reverse engineer the power plant software so that they could provide, you know, the electrical services to people.

(George at 00:21:21) That's not an unusual story, Joel, of people and AI and learning and change and what have you. And you're telling me about the old stuff mainframes.

(Joel Beasley at 00:21:33) Yeah. Yeah. And so it takes a long time. Thank you for helping me get back there. It takes a long time to once technology exists for it to just permeate throughout humanity. So to your point, you have, I'm sure you have many years left as an analyst. Right? Because while this technology today, you could say act like, you know, Freud and it will counsel you in the tone of Freud, right, with those principles that are written in those books. And it can mimic these things and do these things. But that through business and refinement and application and deployment to society, to not only be deployed to society, but become a norm within society. That takes human time for it to disseminate throughout us and conversations and all of that. So I don't think it's like an immediate thing. At the same time, it is change that is likely going to accelerate faster than any of the previous change because, like, have you used Midjourney? How many hours have you logged on Midjourney, if any?

(George at 00:22:36) My daughter renders me all kinds of cool stuff out of Midjourney. I have an account there, and some of the stuff is just stunning.

(Joel Beasley at 00:22:43) Yeah. Have you ever used it in the line of work, like, trying to create Facebook posts or assets? I'll tell you what. Like, we use GPT. GPT saves us about a $100K a year just in man hours because we make our show but we make about 20 other shows for other companies. Like, we produce other podcasts. We do all the prep and interviews and questions and everything. So we have a Slack channel dedicated to how to use GPT to do this and we have one resident expert, you know, who helps teach the other producers all the tips and tricks, and we share knowledge. But GPT has had a huge influence in our business in the past seven months. It's made it easier for us to scale. And then we just started integrating in the past couple weeks Midjourney. And so before we'd have a designer, have design meeting, talk about what we want in the ad, and do all of this, we've boiled that down to the marketing director not even needing to talk to a designer, just issuing several commands to Midjourney and getting it 90% of the way there. And then we just go to the designer and be like, hey. Can you just touch this up? It's so fast. You could literally say that it could cut half of the designers, the need for half of the designers out today. But that's not happened today because it's going to take time for me to talk about that. It's going to take time for people to share. It's going to take time for those tools to get better and people to build tools on top of Midjourney.

(George at 00:24:07) Yeah. Yeah. Well, you look at it. The marketing adoption curve is alive and well. You know, that's the one that said, okay. You've got the innovators who see something, and they grab onto it really easy like yourself. They see what these things are capable of. They don't need somebody to explain to them the value. They don't need to see somebody else what they did, et cetera. They understand the value and they jump right on it. Now, after the innovators, you get the early adopters. Early adopters are just kind of waiting to see if the innovators get the promised value from it, a little bit more conservative, and then they jump as well. Then when you start getting into the early majority, and this is where I kind of use the product adoption curve to parallel adoption inside an organization. These people I noticed need two things. Okay. So you've crossed the chasm. All right. If you want to get out the product adoption curve and make sure I'm doing the labels right, I never know for sure. But you get into the early majority, and these people need somebody to sit down and explain to them the value. You know, like, they hear Midjourney AI, and they're like, oh, isn't that something that does kinda, like, weird things for kids, you know, or something like that. And they're not thinking about how they could use it, like, in a marketing situation. Or they might not be going, oh, you know what? Why don't we go to ChatGPT and ask it, give us some structured question about somebody for their cloud strategy or for their whatever, and then get out, you know, this formulaic answer about what a strategy should entail or what are the questions that you can ask. But they need somebody to explain it to them. Now here's the catch. All right? So number one, we have them saying, all right, you have to talk to them in terms that they understand. Number two, they need to see the value. They need you to not just talk to them about something theoretically, but they need you to actually show it to them. And they need you ideally to show somebody else that's doing it and is getting value from it. So when you're watching the introduction of technology into an organization, it doesn't happen overnight. We actually will say big bang equals big mess. There's too many variables. There's a really, really interesting guy, at least I think he is, at University of Michigan, Dr. Mike Rother. And he said, you know, if Toyota is so great and so much has been written about Toyota, how come so few people can replicate Toyota? Everybody said, wow, that's a great answer. I said, wow, that's a great question. I thought to myself, wow, that's a great question. Because there is tons of books out there. I mean, even when I was going to Japan and stuff like that, there are all these books about Toyota and all this stuff. Well, what he found out was that the role of a leader in Toyota isn't to tell people what to do. It's to coach them through continual improvement. And the Kata that he said, the four-step Kata that he identified, a Kata is a Japanese martial arts pattern. Okay. It's something that you do repetitively to learn how to block, strike, whatever it may be against an imaginary opponent or opponents plural. So he said, all right, you're given a direction and you have to grasp the current condition. Now that's a really interesting way of wording things, grasp the current condition. You don't always know the current state perfectly. All right? And then, you know, we, in historically, what would we do? Oh, let's get in a room and have a group hug. We're going to envision the future state and it's going to be beautiful. It's going to be wonderful. And there's a bajillion variables we know nothing about, but we're going to imagine the future. And then what do we do? Oh, we're going to come up with a project plan and close the curve or close the gap. Right? Well, okay. Unknowns, unknowns. You're going to close the gap and you're in a complex adaptive system that while you're fiddling with it is changing. Nope, no spanky. You're not going to change. All right. And a lot of these transformation efforts fail. So instead, you have to look at it and you have to iterate through. And where I'm going with this whole story is I think when groups look at adopting technology and they look at how they go about doing change, they're fundamentally going about it in a wrong way because they don't adopt these concepts like AI overnight. Okay. You have to instead start small, focused, you have to learn, you have to improve and you have to show the value because the ratchet there is value. People have to understand and see the value. You know, if I keep promising you something's going to solve world peace and I never deliver, you're a smart man. Pretty soon you realize George, you talk a great talk, but you have nothing to show here. All right. I'm on to the next thing. That's the way life is. That's the way business is. But if I can show you the value and you look at it and you say, wow, the value of George's idea, concept, thing, whiz bang, AI, you name it, justifies the cost and risk that I, Joel, am going to encounter. And it's going to happen in the timeframes that are required, what's going to happen? You're going to want it. You're going to pull it in. All right? So when you're deploying these technologies, we have this tendency to want to go all in and go big bang, but there's just too many variables. We need to instead start small, focused, learn, improve, and show the value and grow our footprint. Can a big bang change work? Yeah. It's just riskier than I prefer. If I would really like to help an organization learn, improve and change, I take more of an iterative adoption approach. You know, there's an awful lot of lessons we can learn from history with the introduction of automation, of robots, of whatever that we could take and say, all right, let's not be them and apply that to AI as well. Last comment. You know, let's say you're a graphics designer. And I say, you know what? We're going to replace you with AI. Now, what I want you to do is show everybody how to use AI correctly so we can then fire you. Do you really think somebody's going to show you how to make themselves unemployed? But we do this kind of nonsense all the time, you know? And I think instead of we were saying, all right, how could we use AI in a very intelligent manner that will allow us to scale faster with the people that we have so that we can all have better and better outcomes. I think that's a heck of a lot better message than, oh, so if I get your AI tool, how many people can I fire? You know, it's just a screwed up recipe, but it's one we go after a lot. Sorry. I warned you. I'm nonlinear.

(Joel Beasley at 00:30:17) No. I love it, man. To your point, we'll jump around. My wife and I use squirrel as a verb. So I'll be like, am I squirreling right now?

(Joel Beasley at 00:30:24) She's like, yeah, you're squirreling. I'll be like, alright, cool. Typically, it involves me taking on new projects because I get real excited to take on new projects, and then I have to just cut them all off and be like, no, I gotta focus on the two or three that are gonna really change everything.

(Joel Beasley at 00:30:36) But yeah, you were talking about focus, change, value, and pushing that through organizations, and a lot of people getting it wrong. Right? And you gave the example. You made it super clear, the process. But I want you to bring it home a little bit deeper with a more concrete example. So do you have a concrete example of a transformation that was successful that started with a small focus, they tracked the value, and went through that whole curve?

(George at 00:31:08) Yeah. Well, if you look at a lot of agile transformations, they take an agile approach. Alright? So look at agile development, look at DevOps, etcetera.

(George at 00:31:18) The groups who go in and say, okay, we're going to iterate our deployment of agile, of DevOps, they're gonna have a lot better success than an organization who comes in, alright, we're gonna put together a good team of people, which you can absolutely do, we're gonna have them learn everything there is to know about agile, about DevOps, about fill in the blank, and then we're going to take and cookie cutter this out to everybody with all the variables that we're talking about, etcetera. So all you need to do is to look at the groups that are successfully adopting agile, and not just one or two, but I mean, they're actually going about it. They started small, focused, and they grew their footprint. I'm always really bad with names. Alright, I'll be very honest about it. So thinking of a company right off the top of my head, I'm just drawing a blank. But those would be a good example. Now let's just dovetail to one interesting segue. Look at open source.

(George at 00:32:19) Alright, fill in the blank, any open source effort you like, alright, whatever your favorite one is. Somebody comes up with a vision. What does that vision represent? Value. What does that value do? It attracts people or the collaborators. Then what can they do? Well, then they can crank out some code. They can get some stuff done. What are they doing? They're demonstrating value all the while. They're learning more about what is needed, what they need to do, etcetera. Alright, and they course correct. Alright, this is classic agile. Alright, now, along the line, they're improving their vision, they're improving their deliverables, they're getting more stuff out there.

(George at 00:32:55) What's that doing? It's attracting more users, it's attracting more collaborators. So what can they do? Well, again, they can get more work done. They can refine and expand their vision. So these giant open source efforts that are out there, again, fill in the blank of anyone you like, they didn't show up on the scene at scale overnight. They showed up very small, very focused, and the group started, they learned, they improved, they showed the value, and it grew through the attraction of value. So I think that's maybe an example where I can say broad brush, you look at most of those big efforts, they didn't just magically appear overnight. But look at agile, look at DevOps, those are very good examples.

(George at 00:33:36) So if you hop up on YouTube, there's tons of videos, tons of case studies. Dr. Rother has a lot of stuff up there about the use of this, etcetera. So sorry, Joel, it's just me and my limitations.

(Joel Beasley at 00:33:48) Oh, no worries at all. I'll tell you, I was definitely on the "DevOps didn't deliver the promise, everyone slaps the label on" train. And I think that GPT and Midjourney are the first two things, which they're kind of the same thing, but first two things that have delivered on the hype that I imagined would exist as a person who has built technology for, you know, twenty years. And I was actually very, very surprised when I started playing with GPT and realized that everyone's looking at it in the general public as this thing you can go try to prove is a bot and run the Turing test on and try to talk about tacos or whatever you can talk about to get it to, you know. And then I went to go use it as the business perspective and I said, wow, this thing has legitimate business value right now today. And then the same thing with Midjourney.

(Joel Beasley at 00:34:47) So my instant reaction was fear. Right? It's just like, okay. Typically, I feel fear. I'm like, okay, I feel fear when I don't understand something, so let's go understand it. Let's not be scared of it.

(Joel Beasley at 00:34:57) Let's go understand it. And so, fear of understanding how exactly everything would roll out in the economy and all of that. After my little journey of a couple weeks and inviting on some AI people and talking about it, I realized, to your point, that it will take time to get into the marketplace and time for all that to happen. That being said, your explanation of how that unfolds is the highest quality, most articulate, best explanation I've ever heard. So that's, Josh, we wanna make a comment.

(George at 00:35:29) Yeah, thank you. You know, but what's really interesting, Joel, is you brought up a comment, and it's also really interesting, you know, DevOps delivering versus ChatGPT delivering or any of the other AI pieces. A lot of the problems that happen with DevOps aren't because of the technology, they're because of people.

(George at 00:35:50) Alright, and I'm not faulting the people who are trying to live with what management just unloaded on them. But what management said would solve world peace and all this other stuff. The worst thing I could ever hear is "do DevOps." It's a good idea. Oh, don't do it. Don't do it. Back away at high speeds. You've gotta have a reason to do something. Alright, and if you have a reason, okay, then we can create a system around making that happen.

(George at 00:36:19) Channeling my inner Goldratt and everything. You know, a system is a series of interrelated parts that need to have a goal. No goal, no system. Okay, you have to have something that you're trying to accomplish. So a lot of people who didn't get anything from DevOps is because they didn't go in with any objective in mind.

(George at 00:36:37) Alright, they just did it because it was a good idea. Now you go into any of these AI tools that are out there and you're like, okay, I wanna do X. Alright, you're going straight into a tool and it's cranking out something. Right, you to machine. And it's giving you an output right there. It's a little bit of a different thing. I totally agree with you. I'm blown away by AI.

(George at 00:37:01) I do notice that some of the answers are formulaic. Alright, also, I asked them about who I was, you know, what do they call it, the thing where you always gotta pump yourself up so you put in your thing in the search tool or the whatever. And if they had some of it wrong, they had some of it right. But it's interesting because I'll just sum it up this way: DevOps has a huge people component to it, and that, of course, is the hardest part. ChatGPT and all these other generative AI tools have got these amazing technological capacities that they're bringing to bear that, we're all, like me as a lay person, I look at them like, alright, because I love graphics, I love music. I can't draw, I can't compose, but I love listening to these things. And looking at some of the stuff that these things crank out, it's just stunning. But you don't have that dependency on the human element as part of the deliverable to constrain you.

(George at 00:37:55) So it is interesting, but it's always also interesting to look at and say, now why is there a difference here? I don't know. Just some food for thought there.

(Joel Beasley at 00:38:05) No, I like your ideas. I'm hungry. Right? I like that food.

(Joel Beasley at 00:38:09) You have a book. I wanna give you a little shout out. Right? You got a book. You got an upcoming talk that you're giving with Gartner, if that's correct. Can you tell me what's the name of the book? What's it about? Where can I buy it?

(George at 00:38:21) Oh, the book is from a long time ago, The Phoenix Project. That was Gene Kim, Kevin Behr, myself. Are you kidding me?

(Joel Beasley at 00:38:29) I have the Visible Ops Handbook. You were part of the Phoenix Project?

(George at 00:38:33) I have Visible Ops Handbook. I have Phoenix Project. And see, what happened was, you know, I got hired by Gartner, and they're like, well, do you have anything you need to declare? You know, kinda like going across the border. Yeah, I got this thing right here called the Phoenix Project. It's a fictional book. Oh, it's fiction? Yeah. We're not worried about fiction.

(George at 00:38:48) Game on. Alright, so we did the book, but then when Gene and all the other folks went and did the other stuff, I had to kinda separate from them and focus on the Gartner side of the house, if you will. But yeah, no, that was me. That was one of the three there.

(Joel Beasley at 00:39:03) Oh, that's crazy. That is so cool. I wasn't expecting that. I had prepped for the interview. That didn't come up. And so, yeah, I have your book. It's on my bookshelf. It's notoriously, for those who are listening, if you don't know what The Phoenix Project is, you should go buy The Phoenix Project on Amazon. And it is by far the most notable fiction book about software development, ops, and all that, that's ever been written, I'm sure. Right? There's not a more popular book that's fiction on tech than for this category.

(George at 00:39:35) I can't say that.

(Joel Beasley at 00:39:36) You don't have to.

(George at 00:39:38) Yeah, well, you know, the thing was we realized that if we put out another dry textbook, nobody would read it. But if we created a story, you know, people would. And that's exactly what Dr. Goldratt did with The Goal, which is what our book is patterned on.

(George at 00:39:52) Goldratt came up with this idea and he'd had this body of knowledge that he'd been working on for a while called Theory of Constraints. And he said, well, okay, maybe if I put it in a manufacturing setting and make it kind of like a love story with a guy who's overworked and trying to get stuff done and all this stuff, maybe, you know, that will work in terms of getting the story out there. So he approached this professional writer who thought it was the dumbest idea he'd ever heard, made Dr. Goldratt pay him a flat fee up front to write it. Now that book is sold, you know, it's, I wouldn't surprise you at all if it's pushing 10 million copies. I got a copy of The Goal some years ago. It was already at 4 or 5 million. And let's see, it's years ago. How many years ago? A long time ago in a galaxy far, far away. I mean, it was a long time ago. But every morning, like, that writer wakes up and goes, should have done the percent, should've done the percent.

(George at 00:40:46) Yeah. Yeah. So there's that. And then the speaking portion of it. So I'm a Gartner analyst. I'm always at Gartner events. So I'll be at the Infrastructure, Operations, and Cloud Strategy Summit this fall. My friend Daniel Betts is gonna be the chair of the event. I'll be in both London and also in our Las Vegas one, and a whole bunch of my friends and colleagues, because you get those when you're at a company for thirteen years. And we'd all love to talk to folks.

(Joel Beasley at 00:41:14) Oh, nice. And how can people buy tickets for that?

(George at 00:41:17) Alright. Now this is where George has to say, George isn't 100% sure. We have these websites. So if you search, like, IOCS London or IOCS Las Vegas, Gartner IOCS Las Vegas or IOCS London, it'll come up with the events. And now if you're a Gartner receipt holder, you talk to your account team, they can help you figure out how to approach it and all that stuff. I was ready for everything, Joel, except, now how do we get a ticket? I don't...

(Joel Beasley at 00:41:42) How do we? Alright. So Visible Ops Handbook, that is, tell me a little bit about that book and why I'd wanna buy it.

(George at 00:41:53) Well, a long time ago in a galaxy far, far away, there was this issue of how do we stabilize production. Because when you look at it, a lot of IT work is firefighting. You know, for folks familiar with ITIL, we're talking incident management where you have a deviation from standard production. And the problem is that firefighting eats you alive. Alright, because people aren't multitasking, they're linear. We have to set up to work something, we're then executing, and then we get interrupted. Hey, George, it's a Sev 1, all hands on deck. So you gotta stop what you're working. That's a tear down cost. Now you gotta set up to work the incident.

(George at 00:42:32) People are yelling, they're screaming, all this stuff. Your train of thought just went right out the window. Now, you know, it's done. You tear down from the Sev 1. You come back over, except now you gotta get your thoughts back together. You gotta go, alright, what was I working on and why? And if you're like me, you're gonna declare all your variables wrong after that. You know?

(George at 00:42:52) But the point is, you know, you've lost your train of thought. Now, you know, incidents are just nasty. You know, when somebody says it was only five minutes, the negative impact was way more than five minutes. And for a lot of IT shops, you know, they don't have a lot of people. So if you want to get more productive, more productive work done, more productivity, which you can think of as movement towards the goal, stuff that really matters, not just silly things like lines of code.

(George at 00:43:20) If you really want to improve productivity, one of the things you need to do is to reduce the waste. Okay? Because it's eating up the time, the bandwidth of your precious resources, your people. So in the Phoenix Project, I mean, Phoenix Project, Visible Ops Handbook, what we're basically doing is saying, look, you know, 80% of incidents are caused by failed changes. Now that observation was originally from Donna Scott, who was also with Gartner years and years ago before I joined Gartner.

(George at 00:43:49) And what we saw was true. A lot of the incidents are self-inflicted. You know, somebody made a change and then it blew everything up, usually right before they go on vacation or right before the weekend or whatever. Right? And so then we all have to all go into forensic firefighter mode trying to figure out what happened, and there's gotta be a better way. Right? So Visible Ops was about saying, alright, look, what could you do to reduce this firefighting? Well, a lot of the stuff that you see in there eventually, you know, became some of the things that we talk about in DevOps, some of the things that we talk about in site reliability engineering. And I'm not saying that they built on Visible Ops.

(George at 00:44:28) I'm just saying the concepts parallel one another. Okay? Because there's some tremendous bodies of knowledge out there. But the idea is, okay, number one, why don't we have people talk to one another so that you build stuff that's gonna operate the way that you need it to operate.

(George at 00:44:43) At the start of an incident, why don't we ask the question, what changed? If we know that the majority of incidents are caused by failed changes, then we should be asking right up front, what changed? Which means we need to have the change-related information right in front of us. Now for you and I, you know, like our discussion earlier, okay, I wanna be using machine learning to be running herd on all the stuff I'm doing with monitoring and telemetry. Alright, and I wanna look for something that's going on that's anomalous behavior. Then what I should be doing is saying, okay, I've got 5,000 alerts, all of them are using this load balancer. I'm then diving into the load balancer.

(George at 00:45:24) I'm grabbing the change-related information on the load balancer, whatever, and then I'm presenting it in front of somebody instead of then Joel having me say, oh, look, 5,000 alerts, oh my God. Instead, you know, you're able to dive right in and your mean time to restore improves dramatically. So Visible Ops is really about improving stability, improving reliability. And even then when we were finishing it, Gene, Kevin, and I are like, you know, all this is necessary, but it's not sufficient.

(George at 00:45:53) And then after many, many, you know, "Hey, let's start, let's stop, let's start, let's stop," we eventually did the Phoenix Project. So this BLOB is still very timely for folks who are wanting to stabilize their production environment. It's just a little dated in terms of stuff. Like we say, "Oh, bare metal builds," and now people are like, "What's he talking about?" "Oh, can you build an environment from scratch in the cloud, you know, programmatically?" Yeah, you know, they just take it as a given. But at the time, you know, it's like, "Wow, okay, you know, bare metal build from tape? Oh boy, that'd be cool." You know, it's just, so there's a few little comments in there like that where it's showing its age, but it's good for folks who are looking to stabilize. But I would very quickly take Visible Ops and treat that as food for thought and then also look at concepts from Google Site Reliability Engineering, which I think is so cool.

(Joel Beasley at 00:46:45) Yeah, we actually had some conversations near that with a guy named Tony. I think it was Tony from Broadcom, Josh?

(Josh at 00:46:53) Tony Davis from Broadcom.

(Joel Beasley at 00:46:54) Tony Davis, Broadcom. Yeah, and he was talking about meaningful observability, which was the extent of all of that industry in those years. And I thought that was pretty interesting. But what's your next book going to be? Is there a book coming out?

(George at 00:47:09) Well, a lot of our stuff now happens within Gartner. So I will do these little white papers with colleagues. So last year, myself, Roger Williams, Daniel Betts, Hassan Ennaciri—I feel like I'm forgetting somebody—how do we introduce agile concepts into infrastructure operations? So we did a little research note that linked to a PDF that talked about that, you know, which a Gartner client could get ahold of. What we're working on right now is taking the same people and setting that happened in that first story and introducing some of the concepts of platform engineering into that. So we're just kind of extending there. So right now, you know, as long as I'm in the Gartner family, I have no plan on making any changes. My stuff's happening within Gartner.

(Joel Beasley at 00:48:00) Do you and your colleagues—I know you're doing these white papers—do you guys do a podcast? Can we give that a shout out? Or have you not gotten there yet?

(George at 00:48:09) Well, personally, I haven't gotten there yet. My friend Daniel, he does stuff on blogging and whatnot, and we've talked about it, but we don't have anything yet, Joel, as much as I'd love to have you do a shout out.

(Joel Beasley at 00:48:19) I think you would really enjoy it, but you're already getting to do it. You already get to wake up every day and talk to other technology experts and everything like that. You're just doing it within your Gartner job requirements, right?

(George at 00:48:28) Yeah, yeah, yeah, yeah. And like I said, I count myself lucky, man. Thirteen years I'm still here, which is another way of saying they've tolerated me for thirteen years, Joel. I'm stunned.

(Joel Beasley at 00:48:37) They like it. You're much needed, right? There's all different types of humans needed for all different types of reasons, and everybody needs a George at their company.

(George at 00:48:45) Oh, you don't know what you just unleashed!

(Joel Beasley at 00:48:50) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.