Subscribe
Share
Share
Embed
CRAFTED. is a show about great products and the people who make them. Top technologists reveal how they build game-changing products — and how you can, too. Honored twice by The Webby Awards as a top tech podcast, CRAFTED. is hosted by Dan Blumberg, an entrepreneur, product leader, and former public radio host. Listen to CRAFTED. to find out what it really takes to build great products and companies.
[00:00:00] Kelsey Hightower: When you take the emotion away, you just don't care. Hey, fix this bug. You fix it. Like, uh, hope it works. Y'all let me know if it doesn't work. I literally don't care. And that software feels different. You can tell that there's no empathy in the product, but sometimes you use a piece of software, it's like, oh, someone cares about this thing.
[00:00:16] Dan Blumberg: That's Kelsey Hightower. This is CRAFTED., and Kelsey really cares about craftsmanship in everything he does. Kelsey's a legendary developer and storyteller. He regularly speaks in front of tens of thousands and likes to walk on a tightrope without a net. He improvises his live demos like he's playing jazz and speaks with incredible character and humanity.
[00:00:37] You'll turn to people
[00:00:38] Kelsey Hightower: you say, I know, right? You'll let them laugh. You'll pause, you'll tell the good part over again. That's
[00:00:43] Dan Blumberg: natural. Kelsey helped evangelize and build Kubernetes, and a year ago he retired at the age of 42, leaving behind his job as a distinguished engineer at Google.
[00:00:54] Kelsey Hightower: People are surprised that I'm still doing stuff.
[00:00:56] You still speak at conferences, you still advise startups. It was like, folks, I didn't die. Kelsey will also tell us why he almost stopped speaking at conferences. He thought they were a waste of time, but you always forget that there's someone looking through the camera on the other side of that saying, I'm gonna be like that guy.
[00:01:13] Dan Blumberg: Kelsey loves to level people up, and when I asked him what he wanted to focus on in this conversation, he said he hopes more people will see the power and innovation that happens after, sometimes long after going from zero
[00:01:25] Kelsey Hightower: to one. Those boring innovations. Those are my kind of premium ones because those are the ones that are harder to do.
[00:01:32] No one's asking for those things to change very much. I. And when they do, I appreciate it. 'cause that demonstrates very strong
[00:01:37] Dan Blumberg: engineering. Plus Kelsey shares what he's learned in his first year of retirement and what he hopes for both from new technology and from us.
[00:01:46] Kelsey Hightower: I just hope that if people are this excited about artificial intelligence, that they just get super hyped about real intelligence.
[00:01:54] That's the magic that I hope people wanna return back to.
[00:02:02] Dan Blumberg: Welcome to CRAFTED., a show about great products and the people who make them. I'm Dan Blumberg. I'm a product and growth leader, and on CRAFTED. I'm here to bring you stories of founders, makers, and innovators that reveal how they build game changing products and how you can too.
CRAFTED. is brought to you in partnership with Docker, which helps developers build, share, run, and verify applications anywhere without environment confirmation or management.
[00:02:27] More than 20 million developers worldwide use Docker's suite of development tools, services, and automations to accelerate the delivery of secure applications. Learn more at Docker.com.
And CRAFTED. is produced by Modern Product Minds where my team and I can help you take a new product from zero to one and beyond.
[00:02:46] We specialize in early stage product discovery, growth, and experimentation. Learn more and sign up for the CRAFTED. newsletter at modernproductminds.com.
[00:02:59] It is been almost exactly a year since you announced your retirement. So maybe let's just start there. How's it going? Good. You? I don't know how to do it. I've
[00:03:06] Kelsey Hightower: often had to tell people I'm retired, not tired. Mm-Hmm. People are surprised that I'm still doing stuff. You still speak at conferences, you still advise startups?
[00:03:15] It seems like you're still paying attention to stuff. It's like, folks, I didn't die.
[00:03:19] Dan Blumberg: Yeah,
[00:03:19] Kelsey Hightower: right. I'm just transitioning from the nine to five, the corporate America. But you don't throw away. All of this experience, right? You don't just ignore this thing you've been doing for 20 plus years. So that's it. So it's just a pivot.
[00:03:33] You know, hopefully I'll get to balance out things a bit more, spend more time with friends and family, but also, uh, kind of do things
[00:03:39] Dan Blumberg: on my terms. Yeah. In, in your tweet announcing your retirement, you said, I've spent the last 25 years learning how to work. I hope to spend the rest of my life learning how to live.
[00:03:48] Is there any one or two things you've learned in the past year? Oh, man.
[00:03:52] Kelsey Hightower: Just be patient. There's so many things that, you know, like learning things around the home. I bought a brand new house, but I started changing stuff. Right. It's like, Hey, I don't like that fireplace. I am not mounting my TV over the fireplace.
[00:04:05] It's not happening. I want, someone sent me a thread to this Reddit, post Uhhuh, or this Reddit thread, uh, TV. Too high. Yep. You could imagine when you look at this like, that can't be me. Well, what if I end up on this, on this message board? And so I just kind of refactored the fireplace. So we got the fireplace removed, um, reframed everything so it can support a properly mounted TV at the proper height.
[00:04:31] Yeah, redid it all with shiplap, reran, all the electrical study, the codes, and so just learning how all these other subsystems work and also just saying yes to everything. Hey, there's a friend's birthday party. You wanna go? I'll be there. And I don't have any good excuses anymore. Like, oh, I have work. I have something to do.
[00:04:48] No, now I can just say yes.
[00:04:49] Dan Blumberg: Yeah, right on. When, when we chatted to prep for this interview suggested, we focus on the question of how do the makers mature? Why is that question so top of mind for you right now?
[00:05:00] Kelsey Hightower: I think so much of the tech is all about new, new, new every year and half the people waiting on the new stuff don't even know how to use the old stuff.
[00:05:10] And so everyone's index on what they don't have versus mastering what they do have. And I think we all get caught up in that, especially if you are, you know, if you consider yourself a change, change it. You consider yourself on that side of innovation or making things. We tend to overly focus our energy.
[00:05:27] On bringing something new that people have never seen before, when really, if most companies were at the same technology excellence that Google was or Facebook was 10 years ago, just imagine that your insurance company with the same type of technology stack that Google had 10 years ago. Mm-hmm. That would be a phenomenal insurance company, but that's not what we have.
[00:05:48] We have insurance companies that are still, it's not just that they're using old technology, their mindset. Is also stuck in the same era in how they do things. And that's the thing that I think if you're gonna mature as a change agent or someone who makes things, your job is to try to even take those companies.
[00:06:05] I. And move them slowly towards the trend, and that takes a lot more work. You have to understand how the old thing works in order to do the surgery necessary to get that thing integrated. And most people are just not up for the task because that's actually hard. I.
[00:06:21] Dan Blumberg: It's definitely hard. I, I've consulted to some 200 year old banks and, uh, and so I've definitely seen what you're talking about.
[00:06:27] And then sometimes it's even like, oh, well, well, competitor X has this feature. We should have this feature. It's like, yeah, they do, they have it today. The process of shipping new features that, that, that bank might take, you know, a year or more. So like, let's where they going, you know? It's definitely not easy.
[00:06:41] Are, are, are, are there examples that come to mind of companies that you think have done this well? I mean, Apple's
[00:06:46] Kelsey Hightower: my, the top of the line. So for example, apple gets a lot of slack. People will say things like, oh, Android did it first. Android allows you to do this thing where you can move your icons anywhere you want.
[00:06:58] Apple's just copying. It's like, I don't think people understand how this works when you have like 2 billion. People who use your devices all the time, every day. These are their kind of go-to daily drivers. They're not looking to switch between Android and and Apple every year. What they're doing is they're locking onto your technology train.
[00:07:18] And so what Apple tends to do is they have the iPhone, the watch, iPad, the various things that they have, but every year, you know it's gonna get incrementally better every year. And so the thing that's hard to do is come up with something that was like a game changer. More than a decade ago, and it still be worth using a decade later where people are patiently waiting for those announcements.
[00:07:40] So you might see an announcement, it's like, oh, it's still just Siri. Right? You ask your questions, but look what it allows you to do now. Yeah. So to me, I admire that because they get a lot of pressure, right? They're judged by all of the industry peers. And then I also give credit to a lot of the people who do.
[00:07:55] Infrastructure type work. You know, think about if you're make tires, like a run flat tire is a really cool thing to have right before you drive down the road and your tire blows out and then you're upside down. Now you have these tires that they're flat-ish, they can drive and make it to where you gotta go.
[00:08:14] And we don't think about it. It just works. It integrates with the cars we already have. It works with the road we already have, but it's an improvement. Mm-Hmm. Over everything else. So those boring innovations, those are my kind of premium ones because those are the ones that are harder to do. No one's asking for those things to change very much.
[00:08:32] And when they do, you're not gonna get a lot of fanfare. Uh, but I appreciate it because that demonstrates very strong engineering.
[00:08:39] Dan Blumberg: Yeah, and, and also just understanding what the problem there is, right? In the case of the tires to not, you know, be, be stranded. What is it that you see as people progress in their careers?
[00:08:48] How, how can they embrace this, this harder type of innovation that you're talking about? Also, let's just think about
[00:08:55] Kelsey Hightower: people at different levels of their career and we'll just jump on something like Docker, right? Docker is this container technology. It comes out a little over two, 12 years ago, and this thing comes onto the scene, and if you're a junior person in your career.
[00:09:08] You have no reference point. All you know is that there is a lot of people really excited about something. Let's say you're still in college, you're six months away from graduating, and you're looking at all the news postings about what's going on, what industry are you about to walk into? And so for you, you're like, well, maybe I'll just learn this docker thing.
[00:09:26] So everything you know about deploying applications to servers. Docker, everything you know about the cloud is Docker. And so that's your starting point. And so for you, you might say things like, uh, I wanna be a Docker system administrator. I wanna be a Docker developer. I might wanna work at Docker, or I just wanna use Docker.
[00:09:46] And so that's your world and nothing's wrong with that. And so if you were to join a company and they ask, what would you bring to the table? You would say something like, I know Docker. And it's like, Ooh. Doctor just came out. That's the new hot and shiny, and you think that will separate you from the rest of the pack.
[00:10:02] Right. So that would be your way of doing it. Mm-Hmm. Now, if you've been in the industry for about five years, that means you probably already have an approach to solving some of the problems that Docker solves. So you may have experience with other tools, puppet Chef Ansible, DevOps, configuration management tools.
[00:10:16] So you look at this Docker thing, even like, Hmm, how's it different than what I'm currently doing? And so you'll be able to at least evaluate whether you need it or not. Now, if you're not a kind of forward thinker or you're kind of married to what you already have, I. You'll dismiss Docker for a decade.
[00:10:34] You'll say, ah, it's not production ready. Look at all the features that it's missing. I'll just wait. And I think the majority of our industry is in the, I'll just wait category. We'll keep brute forcing this thing. We'll keep this homegrown solution. Yeah, we'll just keep hiring more people. And so that's kind of like your common group of folks, but then you have this kind of person that's been around.
[00:10:56] The loop a few times when Docker comes out, they'll ask the good questions like, what problems is it trying to solve? And I was like, well, you know, today when you package your apps, it's all over the place. You're using all of these tools. Everyone's building their own root Goldberg machine. And so you look at Docker because you already have a mental roadmap.
[00:11:14] If I had enough time, I would've built something that takes this pain away. Right? That's, that's your mental roadmap. Maybe the team has put that as a low priority. Maybe you don't have the skill or the budget to pull it off, so when it shows up, you can see it. You're like, Ooh, that's the thing I've been thinking about, and it's finally here.
[00:11:35] I have an idea on how to integrate it because this thing was on our roadmap and there's some work that it needs, but it's a good jumping off point. This is gonna save us at least a year. Maybe two. And so then those people, they don't go around screaming Docker from the rooftops. Those people are just saying, we're gonna keep doing what we're doing, but better.
[00:11:54] So those things, we've promised the business that when the code is ready, we're gonna ship it to our customers. Maybe the best you can do is like a day. Some of people are still at a week, and then Docker helps you get that down to an hour. What do you walk in the room and say, do you give all the credit to Docker?
[00:12:11] Or do you say that your team has finally done it? We've gone from a day to five hours down to 30 seconds. Now that progression is like, oh yeah, that's amazing. So that's what you're supposed to do as an engineer. You tuck this new thing that's new and shiny, but you have the proper perspective to put it right in its place on the roadmap, and then it just gets adopted.
[00:12:32] And then you show results. Not talking about the thing that's like the flavor
[00:12:36] Dan Blumberg: of the week. Mm-Hmm. I have a, a couple different quotes from you. You've said a career in tech is less about learning a specific tool as it is the willingness to learn a different tool when the time comes along. I'm wondering if you could expand a little bit more on what you mean by that.
[00:12:49] Yeah, I mean like when we
[00:12:50] Kelsey Hightower: start our careers, you usually apply for a job like Linux system administrator. I know Linux perfect match. I'll come and I'll do Linux stuff and then some people end up with 20 years of one year experience. Any other skills? No. It was like, well, we kind of use Linux for everything, so we're gonna use the Swiss Army knife on every problem.
[00:13:07] I see. And so the truth is like, look, if you're really good at Linux, you have 80% of the skills you need to manage any other operating system. And so then those people that are willing to, they're willing to step back for almost like, we could totally do this with Linux. Here's how much it's gonna cost.
[00:13:23] Here's how efficient that approach is. But I've never used Mac before. Is it better than this? So you got this reference point. Those people that are willing or able to kind of switch when the time came because now your company's doing like Mac development and they can actually use a Mac to do these tests.
[00:13:39] Uh, you step back and say, Hey, instead of reinventing the wheel or trying to come up with an emulator. Let's just use Mac, and then before you know it, you're the Mac person and it turns out the head start you got from learning Unix on Linux, you can apply those skills to Mac and it's a game changer. Yeah, but some people never go that far.
[00:13:58] They're just, ah, I don't know Mac, we need to go hire a Mac system administrator because that's not me.
[00:14:03] Dan Blumberg: Were there particular moments in your own career where you were ready to make that that shift? No.
[00:14:08] Kelsey Hightower: I used to argue about nonsense. I wish I used Puppet for this. Puppet is the greatest thing. Ah, I don't know.
[00:14:12] We don't use chef. That one was written in Ruby. No one puppet for everything. Puppet can do that. Well, if it can't do it, I can make it do it. And so you're just trying to push everything in this box. And those type of people, me included at the time, they're really hard to work with because the team may find a solution that works and here you are trying to shoehorn it.
[00:14:32] And to the solution you have, and that's a challenge. And then when the maturity kicked in, it's like, look, our goal is to build software and get it out on time. How do y'all want to do it? So I haven't always been like that, but I noticed that once I was able to make that mental shift, life got much easier.
[00:14:53] Dan Blumberg: Tell me more about the challenge that you see when A, as you put it to me, people struggle from going from hello world to, to, hello, revenue, I, these are your words, but I'd love to dig in more to this, this challenge that it's not sexy to work on stuff that's not new. And people often struggle in their career and otherwise to, to really innovate at that sort of, I think as Scott Belsky calls it, like the messy middle.
[00:15:14] So hello
[00:15:15] Kelsey Hightower: World comes from the idea that when you're learning software development, usually start with a simple program, hello World. And if you can get that, then you kind of have the foundation to learn more advanced tasks. So you can go from hello world to building a web server, making a website. So it's a good foundation and that's how most people start.
[00:15:33] Hello? Revenue is different. A lot of people in the software industry think our jobs are the end all, be all. Hey, we are the software developers. We're the ones that perform this alchemy on the computers to make them do things that no one else can do. And in some ways, you start to think that you're the most you important piece of the puzzle.
[00:15:53] But when you talk about Hello revenue, it takes a lot to sell software. There's marketing, there's vision, there's regulatory. Things you could have to adhere to their sales, their support. I mean, there's so many other functions that I think those people don't understand what it means to have software as a business.
[00:16:12] And this really shows up in like these popular open source projects. Like if you go to GitHub, you'll see all these open source projects and they're very popular projects. I mean, you have some of these have millions of users. And then when people go on to start a company around these projects. They find it hard to actually get a dollar from any of those companies, and so then those companies go away.
[00:16:32] And then engineers start scratching their head. It's like, Hey, I mean the software is great. Like we're like, number one, everyone uses this stuff, why won't they pay? It's like that's a different game. And that's where you start to do the research. You start to ask yourself, what type of things would people be willing to pay for?
[00:16:48] Then you build Mm-Hmm, right? Projects. You just throw something out there and if they like it, they like it, but this, this is different and a lot of, a lot of people can't make that pivot. So if you can make that pivot to Hello Revenue, that means you can talk to customers. You can do support, you understand the full lifecycle of our products or after it ships what needs to happen.
[00:17:09] That's the set of skills that I think take you to the next level.
[00:17:12] Dan Blumberg: Yeah. A another great quote of yours that I found is, is you've said creating good software is very emotional, and I'm interested to hear more what you mean by that. Yeah. Because I
[00:17:21] Kelsey Hightower: know like when you're a developer and you write code, when you ship it, you're hoping that it works.
[00:17:28] Like, you know, a bug is around the corner. Some user's gonna do something you haven't anticipated. 'cause there's no way you could test every mutation of that app. Mm-hmm. And so you're, you're, you're a bit nervous, right? If you're on call, you might have a little bit of anxiety on this thing, right? No one wants to come in, work in the morning and be like, yo, the whole thing caught fire.
[00:17:48] No one got any sleep. So if you really care about the things you're building. You're gonna have this kind of set of emotions on that first pass. And then a lot of times these things are a reflection of our identities in some ways, right? My craftsmanship, my skill level is reflected in that final product, right?
[00:18:07] If you ever build something and you step back from me, it's like, yeah, I did that. And I think that's what they mean by it. Has your name written on it? Mm-Hmm. Because when you take the emotion away, you just don't care. Hey, fix this bug. You fix it. Like, uh, hope it works. Y'all let me know if it doesn't work.
[00:18:21] I literally don't care. And that software feels different. You can tell that there's no empathy in the product, but sometimes you use a piece of software, it's like, oh, someone's used this before. Someone cares about this thing. And I think this is where those concepts around dog footing come from. Where, you know, when you meet a developer that doesn't use the stuff that they work on.
[00:18:42] It's super interesting 'cause they just don't have any empathy for us. Like, I don't know, I don't, I'm not an accountant. I work on QuickBooks, but I'm not an accountant, so I don't know if it's good or bad. So I just shift stuff as sketched out by the UX team and hopefully the customers like it. I will never know.
[00:18:57] That's like the equivalent to a race car driver with the, you know, like a, a mechanic with no driver's license. Like you work on cars. Yes. Do you know how to drive? Nope. Don't know how to drive. But I work on cars. It's like, man, I don't know. I, I kinda like. The situation where someone can actually feel their product, use their product, and approach it with empathy, uh, versus just something you do from nine to five.
[00:19:16] Dan Blumberg: Yeah. It's funny, the, the analogy that comes to my mind actually is I worked, before I got into tech, I was in radio. One of the hosts that I used to work with, she was the host of Morning Edition in Chicago Public Radio. She'd be in a studio all alone at 6:00 AM and she'd literally had taped in front of her on the wall.
[00:19:31] A smiley face, like an emoji smiley, just to remind herself she was talking to someone and just a smile and like, you know, and that, and what I think you're saying is like that emotion of like, it gets into the keyboard, it gets into the code, it gets into the actual user experience. If, if the developer, him or herself is, is emotionally invested, I mean, who are you building
[00:19:47] Kelsey Hightower: this for?
[00:19:47] I mean, the first time I learned how to cook and I got to cook for like Thanksgiving where you make one of the dishes that everyone's like, Hey, you got the macaroni and cheese this year. Lots of people are looking forward to the macaroni and cheese. Some people, that's all they eat. And so if you messed that up, but when you do a good job, everyone's smiling at you.
[00:20:09] Who made the macaroni? You're just like, ha. Or me? Yeah. Yeah. Or good job on this. So I think it's just one of these things where you know that you're, who are you cooking for? Right? We make food with love and people can tell when you did that. Now, some people just might say, look, I'm just gonna microwave some macaroni and cheese in one of those cups.
[00:20:26] You know, you just add water, you mix it up. Maybe you make 50 cups for Thanksgiving. Yeah, but what is this? I dunno. This is not good. Uh, you know, someone took a shortcut here. You didn't do this with empathy.
[00:20:40] Dan Blumberg: You are very famous for your live demos, and I would love to understand what is it about the demo?
[00:20:45] I imagine it's some of this emotion we're talking about that's so powerful, and also what are some things that people can do to improve their own demo skills?
[00:20:53] Kelsey Hightower: So we work in the tech industry. I mean, there's products coming from everywhere. Right. There's new phones, there's new software to help you do your taxes.
[00:21:01] There's microphones, there's editing software. There's just so much. It's impossible for anyone to discover half of this stuff, let alone use it. Especially hard when you're making developer tools, developer tools for those that are not in a software game. It's like going to Home Depot. There's like six different hammers.
[00:21:19] They all appear to do the same thing.
[00:21:21] Dan Blumberg: Yeah,
[00:21:21] Kelsey Hightower: it's a hammer. Can't be that interesting. But some of them just pick the right color scheme. But imagine going to Home Depot and you're walking by to pick up a hammer. Mm-Hmm. And one of the vendors has a video playing. Our hammer always pulls out the nails You did wrong.
[00:21:38] Look at the diamond laser cut nail extractor tool we have on the back. He said, Ooh, nail extractor. I don't think these others have nail extraction technology and like the other people make cameras, like there's no such thing as nail extraction technology. It's a wedge. You put the thing we, we have that too.
[00:21:56] It's like, no, no, no. I'm looking for nail extraction technology. And so when you see it in action, you're watching the little demo. They always have the hard rock music video in the background. It's like, man, look how I was getting those nails. I wanna look like that. When I pull nails out in front of everybody else on the construction site, so I gotta show up with the right thing, and you show up and everyone's got the same hammer hanging from the belt, you know you bought the right one.
[00:22:23] So when it comes to the live demo, unfortunately, when a lot of people do public speaking, they public speak like a robot. They take all the human elements out of it. The way we talk to our friends and family, the eye contact, the body language, reading the room, they throw all that away. What they do is say, I'm gonna pre-calculate every emotion from the audience into some speaker notes and bullet points.
[00:22:46] And so what they end up doing is essentially copying the website. Here is our product. As you can see on the slide, you just like, oh my God, if you, if you don't stop reading the slide to me, so everyone just zones out. The thing is when you use a new tool for the first time, like in the software space, like let's say you're working on a text editor, writing software is already hard, but the first time you use a feature like auto complete where you just open a file, it's like.
[00:23:16] OS dot, and then it's like open file and it guesses the file name based on the directly in. And you see that for the first time. It's like, oh my God, this is amazing. Yeah. You know how much typing I can save if I can just hit dot and tab and it just auto completes it. It's like magic. So the first time you experience this, you feel a certain set of emotions.
[00:23:39] And so what I tend to do is as I'm building out a demo. And I feel this way because first thing you gotta do is learn how to use it yourself. And then maybe you've been using it for a while, but then you tend to forget on how valuable it was. And you can get too far away from those original emotions.
[00:23:54] But for me, I go back and remind myself, and when I get that emotion, I just record it and say, Hmm, I need to get people there. And what you'll notice is that you have to back up. You can't just get to the thing you do because it doesn't make any sense. There's no context. You need the bookends. And so what I'll do is I just build it up.
[00:24:12] I was like, Hey. You ever try to manage a server and you run the automation tool and the server gets deleted and then you gotta start over. And people are like, yeah, but what happens when you have a thousand servers? You gotta do all these things. It's like, yeah, yeah, Kelsey, I know that experience and I hate it a lot.
[00:24:29] It's like, great. So now we have shared context and then I give them the moment I had when I got the demo working and I'm using a new tool like Kubernetes where I give it a hundred machines and I just start deleting them and things just start coming back automatically. I. I'm like, oh my God, this Kubernetes thing is dope.
[00:24:47] And I say, oh, that's the demo. And so then I put the story about what we're all currently doing. So there's this buildup. Where's he going? And I say, Hey, look at these same a hundred machines. I'm not gonna do anything to them. I'm just gonna run a bunch of apps. You know that Kubernetes is picking where they run and people are like, Hmm, how's it doing that?
[00:25:06] What happens if he deletes one of them? It's gonna look exactly like my current problem, and I delete one of them. And they're like, did it just come back by itself? While they're sitting there? They get the same aha moment that I had, and that's where my live demos come from. I'm just trying to transfer the emotions that I got all by myself, and in some cases try to give it to an audience of like 20,000 people.
[00:25:29] And when they all get it, you can just see the collective light bulbs go off. Right. And you know that those people will never be the same again after seeing that.
[00:25:37] Dan Blumberg: Yeah. No matter what you're doing, it sounds like you're focused on storytelling one way or another. And I'm curious, where did you learn this love of telling stories in this, this way to do it so well?
[00:25:47] I mean, that's all what we're born to do.
[00:25:48] Kelsey Hightower: You're naturally born this way. It's just that corporate America extracts it from you. It just tells you, that's not professional. Use a slightly different vocabulary. Walk this way, dress this way, talk this way. And so everyone starts doing it the same way. Then it becomes incredibly boring and stale.
[00:26:06] But in real life, no one does that in real life. We all do what's natural. You'll use the whole room, use your body language. You'll turn to people you say, I know, right? You'll let them laugh, you'll pause, you'll tell the good part over again. That's natural. And so what I learned to do is just do that. And I dropped the speaker notes.
[00:26:28] I get rid of the slides most of the time, unless the imaging is gonna really help me. And I played the demo like live jazz. Some people try to script a demo perfectly. You've seen those kind of demos where they run every command in sequence, and then when something goes wrong, fish outta water. Oh, the demo wasn't supposed to do that, and the whole audience is like, okay, keep going.
[00:26:51] It's like, I can't, I've only did this where everything works. Now that things aren't working, I guess I'm done. I do mine more like jazz. I'm playing. If something goes wrong, then you slide over and you do something different. You improvise. Yeah. And then you get back on track or you just know it's gonna fail.
[00:27:07] So that's kind of been my approach to it. Just be as natural as possible. Do what you do in front of friends and family, and if you can find the courage, because there's a lot of vulnerability in that. You're gonna have to be you. And a lot of people don't want people to see who they are, so they hide behind.
[00:27:23] At the slide decks and the speaker notes and the way you're supposed to speak versus the way they actually are.
[00:27:29] Dan Blumberg: Yeah. I wanted to ask you about vulnerability. How did you get to be so confident on stage? Or did this come naturally or did this come over time? When I was
[00:27:37] Kelsey Hightower: first speaking, I think I was doing everyone else's doing.
[00:27:39] Get your slides, have your speaking points, have everything tight. Even do a rehearsal and dry run. But I remember there was a couple times where I was like, you know what? I don't have time to produce slides, so if I can't do slides, then what I gotta do? Well, well, you're gonna have to show 'em some. Well, how about you show them the thing you're actually talking about?
[00:27:58] So, oh yeah, I could do a live demo, but that takes a lot of work to prepare. It's like, yeah, but don't make something that's too pie in the sky. Give them something real. And so you go up there and you walk on stage and you just look at people for a moment. It's like, have you ever written software that just didn't work?
[00:28:15] And everyone was looking at it like, yeah, I was like, me too. And I just did that recently. I wrote some software that didn't work, so I'm gonna show it to you really quickly. So here's the little software I was working on and I'm gonna run it and it didn't work. And everyone's like, this sounds surprisingly natural.
[00:28:33] And you're looking at them and they're all like putting their phones down, like, where's this going? And he's like, now I had to ship this before I came here. So I had to fix it and I didn't have a lot of time. And look, I've been ignoring debuggers like my whole career. My debugging has been putting print statements, going to the shell, looking at everything.
[00:28:54] But look, it's been taking too long. I decided to use the debugger. Did you know these IDs have a debugger? And look, some of you, all this is just normal. But a lot of, you're probably just like me, so here's how I use a debugger, or at least here's how I use a debugger. I'm not saying I'm the expert at debuggers, but here's how someone went from print statements to using a debugger.
[00:29:16] Here's what I like. Here's what I don't like. In the future, I can see myself doing a bit of both, and then people are like, yo, I really enjoy that talk. I use it to bark all the time, but I never thought about why. Or someone in the audience was like, man, I've been using Prince Davis for 20 years and after seeing this, I think I'm gonna try that debugger thing out for a spin.
[00:29:38] And so once I'm getting this kind of feedback where people are like, my favorite part was when you did this thing. Mm-Hmm. And I also noticed that there was room for jokes because we're all doing this work and it's like when something's funny, I just laughed on stage. I was like, why? Do we check in the print statements?
[00:29:57] Have you ever checked in one of the print statements and people are like, oh yeah, I remember doing that, where you put all these print statements for local debugging and you're supposed to take them out before you check in your code, but you don't. And so when you get a code review, it's like, Hey Kelsey, um, what's this print statement?
[00:30:11] It's like, oh yeah, don't worry about that. And so that's natural. And so then once I got that feedback loop, I was like, you know what? This is what I'm gonna do by default. And I'd rather people hate the talk. Than to just not remember it. Mm-Hmm. And so I just decided that's just the way I'm gonna do it going forward.
[00:30:29] Dan Blumberg: Yeah. That's a similar thing with products, right? You'd rather, people either love or hate your product, but God forbid they just, you know, it's nothing. It's meh. Is there a favorite moment on stage or joke on stage that that comes to mind?
[00:30:41] Kelsey Hightower: Oh, oh yes. A hundred percent. So when I was at GCP and you know, the big cloud providers, actually, all the big companies, when they do their keynotes, I.
[00:30:52] This is serious business. These things cost tens of millions of dollars to do these events. They got the cameras, the lights, the set design. This ain't no meetup, and so typically they want everything perfect. Pretty much all the stuff that I said that I don't like to do, that's what they want to do. They even have speech coaches, they have people that write stuff for you.
[00:31:13] There's even table reads where everyone gets in, like, do your part and you really have to sit there for the whole time and there's a chance that your whole section may get cut if the whole thing's too long. And so they were like, Hey, Kelsey, you wanna give, um, you know, a keynote during GCP? I'm like, yes, but, um, I don't want no input from nobody.
[00:31:31] I don't want no table reads, nothing. And they're like, yeah, but you know, you gotta, it's a team effort. It's like it is a team effort and I promised to do my part, but y'all gotta gimme a little leeway. 'cause I just want it to be what I've grown to be able to do. And luckily at the time the CEO of Google Cloud was Diane Green.
[00:31:47] I. One of the co-founders of VMware legendary technologists, and she was like, you know what Kelsey, you can kind of have that free reign. And I remember going to one of the rehearsals and everyone's got their slide deck perfect. They got their talking points, the teleprompters in there, and they were like, Kelsey, no slides, like no slides, no script, nothing to put in a teleprompter.
[00:32:09] And so they just put ad-lib. And so I get to the driver and they're like, you know, can we kind of get a preview? What are you gonna do? I was like, no, I just know that there's a demo that I, that works. Uh, I think I know the arc I want people to take and, uh, yeah, see you tomorrow. And you could just see the executive team at Google, like, oh my God, what is he doing?
[00:32:31] 13 minutes of live demo. Oh my God. Is it recorded? It's not. It's not even recorded. Oh my God. If the internet doesn't work, this thing is over. Like Yeah, it's over. Yeah. And so I'm doing this live demo. I'm gonna build an app from scratch. We're gonna start with the database, let's get the database schema database is working.
[00:32:48] So now I'm just trying to show people the big picture of the cloud and how far we've come. And so I'm just deploying pieces and deploying pieces and I had this joke. And I added it in this morning. I saw a buddy of mine. I was like, Hey man, I need you to help me with something. I wanna talk about the reality that a lot of developers, we pull stuff from Stack Overflow, we don't have all this stuff in our head.
[00:33:08] He's like, what do you need me to do? I said, look, just post this answer to Stack Overflow and I'll use it in my talk. And so I was gonna write some code for this piece of the talk, and I decide to say, you know what? I'm gonna be honest with you all. I don't have time to write all code from scratch, so I'm just gonna go get it from where all the great developers go, and I start typing it in my browser, stack overflow, and the audience starts laughing.
[00:33:28] First pro tip. Now you gotta write some code. Now this is what we do. You do this. If you haven't heard of Stack Overflow, like this is high quality code that you can find on this thing, man. Trust me. I was like, this is what we all do. Stop pretending. And so I'm trying to copy the code from Stack Overflow, but it's a brand new Chromebook that they had set up as a demo laptop.
[00:33:52] So my key bindings aren't right. Things aren't the way I want it to be. And then there's all these popups. 'cause it's a new browser asking me if I wanna turn on over. I'm like, I'm on stage dying. That's like, Kelsey, you are like, not even halfway. Grab this Uhhuh, let's see what we're gonna do here. So let's grab this code.
[00:34:08] Come on. I don't care about cookies. Where's my copy and paste? If you stop here, it's gonna be embarrassing as hell. And they told you not to do this. All of them told you not to do this. And I'm sitting there and it feels like an eternity on that stage. Like, why are they trying to sell me stuff? I'm trying to get my job done.
[00:34:27] Oh, come on. Like the hardest part of this. All right, look at this ads. 'cause you need to stop on time. You can't go too far over. Yeah, and I remember eventually I got it, my cursor right in the box where I can start to highlight and my goal was to get the cover stack overflow and put it in my ID and it just came to me.
[00:34:48] Good developers. Copy. I went to my IDE and I said, great developers. Paste and it pasted. All right, copy. So one thing is good programmers. Copy, great programmers, paste.
[00:35:07] And I and I ran it and I was like, who? What a safe, all right, so we got our code. Now the next thing we need to do is set up a database. Now I am the world. And then the rest of the demo went by suite and then I kept building it. And then I showed off dialogue flow where you can like do the whole okay, Google thing.
[00:35:24] And talked to the app we just built. And at the end. As the Google assistant was talking back to me, I was like, all right, thank you. And it said, I gotta admit that was extra dope. And it came through the house speakers and it was just like this big applause. Thank you. Okay. I gotta admit that was pretty dope.
[00:35:45] And Diane Green. CEO, Google Cloud is watching this play out, like masterful. And as I'm getting my microphone ticking off backstage, she goes to the back, she's like, oh my God, Kelsey, that was extra dope. And I was sitting there in my mindset like we were about four minutes from it not being mm-Hmm. But that was like one of my greatest saves live on stage because there was no fallback plan.
[00:36:07] Dan Blumberg: Yeah. Yeah. On the high wire without a net. Um, another one of the, the quotes that I, I love that I, I read from you, and I think the story is an example of it. A lot of people don't realize how much power they have until way later in life. Mm-Hmm. I'd love if you could explain what you mean by that.
[00:36:23] Kelsey Hightower: I mean, it's the simple things too.
[00:36:25] When you're going to the airport, you're kind of rushing. There's a dude that's probably been at work for seven hours and 30 minutes. He's ready to go and nobody sees him. Right. He's just a guy there, you know, probably has to sweep the sidewalk or something, and so you just walk by like everyone else. And this person.
[00:36:46] Might feel invisible. And so one day you stop and you just smile, Hey, how you doing? And they look at you like you talking to me like, yeah, how? How you doing? And you might even have a conversation like, yeah, I'm doing pretty good. I said, man, I love the way they were designing the airport so much better.
[00:37:02] It's like, yeah, I've been working here for seven years and the one thing we used to complain about was this thing and they're fixing it now. I was like, oh, they're fixing that. So now you can actually cross between the two corridors. Oh yeah. Oh, they also got a new Chick-fil-A over there. Oh, they got Chick-fil-A in there.
[00:37:16] Oh, I'm gonna make sure I ate it. It's like, I appreciate that the smile they have. From being seen and they got a family and they got a kid, and maybe the kid has been feeling like dad ain't see them too. And then dad's like, man, someone saw me today. They may go knock on the kid's room and be like, Hey, how you doing?
[00:37:36] And they do it with patience and really explore the answer and repeat what you just did. Who knows what that spawns into. You don't have to be Superman, you don't need a cape. You don't have to win an election. You can literally choose to be that way everywhere you go. I'm not saying unrealistically, you're gonna do this all the time, but I'm saying that you hold that kind of power.
[00:37:58] So that's what I mean by just this ultimate power we have. Now. You can scale it up because. I didn't realize how much power my voice had doing those keynotes, but I was, I was at a conference in London and this guy was like, Hey Kelsey, and he was whispering. I said, why are you whispering, man? Wait at a conference.
[00:38:14] You kidding? You kind talk as loud as you wanna have to whisper. He was like, no, no, no. I was in jail watching you do this Kubernetes thing in jail, and when you're in jail you have lots of time. So I'm studying Linux and all of these things, and I just want to be able to do what you're doing. Because I had all his time on my hands.
[00:38:32] So I come outta jail and in tech, a lot of these companies aren't doing deep background checks. If you have the skills, you're gonna get a shot. He's working for the starters. I say, Hey, show me what you built. And he opens the laptop and he gives me like a little mini demo and I'm blown wha? I'm like, I'm not li.
[00:38:49] You're actually really good. This is really good. And I saw his C-E-O-C-E-O walk by. I was like, this guy is good. This product. It is good. He's like, I know. We're so lucky. We found him. Ever since we added him to the team. It's been a big boost to the bottom line, the product. That's where we're at Coup Con today, and I'm thinking to myself, I was like, man, there was a point in time where I thought the key notes and the talks were a waste of time.
[00:39:16] But you always forget that there's someone looking through the camera on the other side of that saying, I'm gonna be like that guy. And so that to me is extremely powerful to be able to have that impact on people.
[00:39:27] Dan Blumberg: Yeah. I wanted to ask you about what it means to be a black man in tech and be on stage and how that representation matters.
[00:39:34] I mean, that question that that's what
[00:39:35] Kelsey Hightower: came up in London. 'cause my background is hands-on technologist. You can see my source code on GitHub. I always wanted to be that first. I didn't want anyone to be giving me a, a, you know, a prize just for being my skin tone. I didn't want anyone to kind of guilt themselves into an opportunity.
[00:39:56] I wanted to earn it just like everybody else. And so you're on stage and you work real hard to have that be your persona. That's the thing. So we're coming off those stages. You find yourself in those kind of call 'em the hallway track in those circles, and you're just kind of talking about. You know, what's the future of Kubernetes?
[00:40:13] Should you do it this way or that way? And this one gentleman comes in and he asks that exact question, what is it like being the only black person in the room? How do you navigate this? Because it can feel weird if no matter what your color is, if you find yourself in an uncomfortable environment, it can feel weird.
[00:40:32] Like if you don't know how to play basketball and you go to the basketball court for the first time. And you are looking at everyone else look like they have superior skills and you're like, man, what am I gonna do if someone allows me to play? I don't wanna make a fool of myself. It can feel uncomfortable.
[00:40:47] And so you gotta learn how to navigate that. And it's hard to do it when you feel like you're by yourself when there's like three new people there. There's like, Hey, we're gonna stick together, right? Maybe pass me the ball if you get a chance. But if that's not the case, and I think for him, like many others myself, at certain points in my career, you do find yourself in these uncomfortable scenarios.
[00:41:05] So he asked that question, and then all my industry peers are also in this circle, and they're like, yo, we ain't, we don't have an answer for that. And so I'm sitting there and I'm, I'm kind of stumped a little bit, and I have to just dial it in and just let 'em know. It was like, you know what? That's still a thing I.
[00:41:21] Luckily for me, people know what my name is, so if I walk in the room, they say, oh, that's Kelsey. No one makes me prove myself time and time again. But I do remember what it's like before they knew my name. And so I kind of explained that to him. And honestly, I probably didn't even give a great answer. But then he said something that was dope.
[00:41:37] He said, Hey, I never thought I needed a mentor until I saw you doing what you do. And when you knew you were gonna come to London. I brought my son with me and he had his hands on his son's shoulder the whole time. And we were looking at him in his eyes and I didn't notice his son. And so I said, I just wanted my son to see that not only can we be here, we can lead the way.
[00:42:00] Wow. And that was kind of like that moment of like, woo, representation does matter. And it, it could be anything. And so I think life is just like that when we see ourselves do something. Uh, it just gives us a lot of courage and, and gives us the ability
[00:42:14] Dan Blumberg: to at least try. Yeah. Talk about the power of, of open source and how open source opens doors, like the kind you're talking about.
[00:42:21] I've never seen anything
[00:42:22] Kelsey Hightower: like this in human history. I'm 43 years old. I'm not saying I've seen everything, but most things in life we're kind of taught, especially in least western societies. You know, I grew up in the United States. If you create value. You tend to wanna hoard it for yourself, package that thing up and sell it for as long as you can, and you may even sue someone who tries to copy your idea depending on how it comes to market.
[00:42:49] The thing that's really interesting is that when you show up to some of these communities, the very same people and other people in those communities take the time to teach you all of the things. They show you their code, they explain things to you, they write docs to you, and they give you something more important than just free software.
[00:43:07] They give you the knowledge to create your own. There's not a lot of industries where there's that opportunity for profit. And then we have all these people that just said they just wanna balance the scales. This is fringe community that's like we want to be able to build the software too. And software is this humongous multiplier effect where one person can truly produce software.
[00:43:30] That can be used by millions. It's almost like musicians, right? A musician can write a song once and a world can listen to it billions of times, and some of those songs can go on to become the national anthem. And so like when you think about like a true musician, yes, they probably love the money that comes with their hit record.
[00:43:47] They probably love the fact that they can do these concerts and make all the money on the concert tickets, but most of them just like music. They like to listen to music. They like to make music, and in many ways they are showing you like their creativity is expressive of that. For a lot of us people in the tech world, especially the software side, we also have an outlet.
[00:44:06] We can create applications and software and it could just work only for us, and then we could choose to share it with the rest of the world. And if you like it, you can use it too. And no strings attached. If you want to change it, knock yourself out. So when we say community. I think that's what people mean.
[00:44:26] It's just hard to do when all the other things get in the way. And so I've been really appreciative of open source community. 'cause as someone that didn't go to college, someone who's wanted to be competitive in this industry, meaning be the best that I could be. You can't do that without all of these people willing to share their tips and tricks and show you how they did things so you can get better over time.
[00:44:47] Dan Blumberg: Yeah. Back to where we started with from, from Hello World, the Hello Revenue. I'm, I'm interested in how, what we just talked about with the power of open source and you know, sharing your knowledge, sharing everything, how that jives with the first conversation we had my last year at Google
[00:45:04] Kelsey Hightower: was very interesting.
[00:45:06] I was there for almost eight years and learning to go from Hello World to Hello. Revenue really helped because when you become a principal engineer, I ultimately retired. A distinguished engineer. At that point, you're an executive, so you do executive briefings and you know, you have to think about revenue as in terms of your impact.
[00:45:23] Your impact is no longer expressed in, uh, I wrote this code, or I helped this feature get shipped At some point it's like, here's the impact to the business and then revenue plays a big part of that. So once I started getting good at this hello revenue thing, I also started to really deeply examine my own career.
[00:45:41] I was like, what is my actual impact on society? Why do I actually work? Look, a lot of us, we work because we need to pay for things. It's really simple equation. But one day I said if I was able to buy my time back, would I do it right? So you gotta have enough money from here until the end of your time and all the people you want to be able to support and, and as for that process, would I buy it back?
[00:46:03] And then I was faced with this situation where I can go from principal engineer to distinguished engineer and they have this like mentoring program where each business unit, you can think YouTube or cloud. A select group of people are given an opportunity to do this thing where they kind of get you on the right path.
[00:46:18] This could be mentorship, it could be a leadership program, and these things take time. And I thought about it, I said, um, I was so honored to get this because a lot of people don't get this. And I remember sending back a letter to HR and saying, Hey, I'm done. I'm done with the latter. The thing I need. Next goal.
[00:46:37] Can't give. If you're gonna be an employee with responsibility, ultimately these are gonna be success factors you're gonna be measured against, and it makes sense. It's a deal. That's what you're signing up for. But where I need to go next, revenue cannot be the dominating predicate to my decision making.
[00:46:54] And so being able to get to a place where, what you'll be remembered by, I promise you and I yet to even think about all the great legendary people that we study in school. I never really heard anyone really over index on how much they were worth, how much revenue they generated. We just remember the impact they had on like your lives or the lives of others.
[00:47:15] And some of those people died broke. Some of them died rich and famous. But either way, we all still just remember their impact. So I kind of figured I wanna make sure that I have those type of stories to over index on. 'cause no one's gonna really care about all the revenue I generated. That's just secondary.
[00:47:33] Yeah. Right on.
[00:47:39] Dan Blumberg: Is there anything that today sounds like science fiction that you think is gonna be totally commonplace in the near future? You know
[00:47:46] Kelsey Hightower: what, so I think of you can't ignore ai. And the interesting thing about AI is like when people look at AI and action, they're just floored by it. Look at this thing. You can talk to it.
[00:47:57] It doesn't always give straight answers, but look at it. It feels magical. It seems so smart. I asked it to draw a picture. To draw a picture. This is the future and we're gonna be in another galaxy because of this moment. And then you're like, damn, did you know humans could do this too? Five year olds can do this with like, I don't know, 20 watts of power, a couple bananas, 800 calories.
[00:48:21] They can do this too. And we also have mental models of the world. You are, you have these lived experiences where you can look and see a shadow and know that there's a cat playing with yarn on the other side, and a lot of these ML models will struggle with that task. But you can do it, but you don't look at other humans in Marvel.
[00:48:42] At the future, you don't look at this other person and say, wow, this other person's studying cancer cures. It's the future of cancer. But if I show you this computer with the blinking lights mm-hmm. Crunching data, you're inspired to the fifth degree, but looking at your fellow citizen, you're not. So, yes, it feels like magic that somehow this thing that is still not capable, does it have a better memory than the average human?
[00:49:07] Uh, of course. Does it have more computing power than the average human? Collectively, maybe, of course, but there's no match for all the things that a human can do at all. It's not even close to that, right? This is where they would like to be, but in many ways you already have it. But the moment someone can take this pizza box and make it do something similar to what we do, all of a sudden everyone's inspired again.
[00:49:33] I just hope that if people are this excited about artificial. Intelligence that just gets super hyped about real intelligence. Yeah. And we start to treat everyone else. Like what happens if you start fine tuning your own model? What happens if you go out and get new experiences so that you can be a little bit better at understanding the world so you too can make better predictions or try new things and show off your creativity.
[00:49:57] 'cause you can also generate stuff too. Yeah. That's the magic that I hope people wanna return back to. Yeah. Kelsey, thank you so much.
[00:50:06] Dan Blumberg: Thanks for having me. That's Kelsey Hightower. I'm Dan Blumberg, and this is CRAFTED. CRAFTED. is brought to you in partnership with Docker, which helps developers build, share, run, and verify applications anywhere without environment confirmation or management.
[00:50:24] More than 20 million developers worldwide use Docker's suite of development tools, services, and automations. To accelerate the delivery of secure applications. Learn more at Docker.com.
Special thanks to Artium where I launched. Artium is a next generation software development consultancy that combines elite human craftsmanship and artificial intelligence.
[00:50:47] See how Artium can help you build your future at Artium.ai.
And CRAFTED. is produced by Modern Product Minds where my team and I can help you take a new product from zero to one and beyond. We specialize in early stage product discovery, growth, and experimentation. Learn more and sign up for the Crafted newsletter modernproductminds.com.
[00:51:11] Please share. Crafted with a friend. We have some incredible episodes coming up and some shiny new features too.
Ooh, nail extractor!