Karl Hughes CTO The Graide Network

(Joel Beasley at 00:00:00) Today, we are talking to Carl Hughes, the CTO of The Graide Network, and we discuss the next generation of programming influencers, his upcoming talk this week at CodeMash, and his responsibilities as CTO. Carl was a fantastic interview packed with value. All of this right here, right now on the Modern CTO podcast. Here we go. This is the Modern CTO podcast.

(Joel Beasley at 00:00:36) So you just want to tell me real quick again about what Graide is?

(Carl Hughes at 00:00:41) Yeah. So The Graide Network, what we do is we hook K through 12 teachers up with college students studying to be teachers, and they'll grade their papers for them. And so typically, what that looks like is we go to the school. We basically sign them up for a package of grading hours. They distribute those hours out to all their English teachers.

(Carl Hughes at 00:00:58) And then throughout the year, as they have more assignments, essay assignments they want to send out to their students, they'll collect them and then have us grade those papers for them. So unlike fully automated grading solutions that only work for multiple choice or math kind of questions, we grade full essays. So short answers, essays, long written content that at some point computers might grade, but not anytime in the immediate future.

(Joel Beasley at 00:01:24) So no machine learning AI is doing this. You actually have people grading.

(Carl Hughes at 00:01:28) That's right. Yeah. I mean, obviously, long term, you know, the whole natural language processing mixed with some kind of AI is going to augment or help our graders. I don't know if it'll ever completely replace them for every essay, but maybe there's situations where we can do some of the easier work that way. Yeah.

(Carl Hughes at 00:01:45) It's like we're not just grading on grammar and spelling. We're grading for the actual content, like the themes of an essay that's written about a piece of literature. And that's so hard for computers to sort of break down into digestible bits at this point. Someday, once again, maybe we'll be there five, ten years. I don't know.

(Carl Hughes at 00:02:03) Yeah. But at this point, we kind of position ourselves as this manual solution because the automated ones won't be good enough for a while.

(Joel Beasley at 00:02:14) Yeah. And there's always that in between of the hybrid space where you can build technologies to assist your students.

(Carl Hughes at 00:02:20) I mean, even if automated grading does become a reality, everyone who does it is going to need a large dataset of real graded essay papers. And that's what we're going to have that upstarts in five years when this becomes more possible won't have. So I think we'll be in a good spot to do it even if we aren't today.

(Joel Beasley at 00:02:38) Do you think technology is reshaping education in the sense that the biggest brands, the MITs, the Dukes, and stuff, that they'll kind of die out? And then it's going to be an exchange of more real value education, like being able to go and work with a doctor. How do you think that's going to play out?

(Carl Hughes at 00:03:02) That's a good question, Joel. I mean, the student debt crisis that we're entering into is massive. I mean, this is billions and billions of dollars that the US owns in debt that the people who graduate with some degrees will never make enough money to pay off. I mean, if you have $100,000 in loans and you're going to make 35 to 50K a year, how do you ever get to a point where you can pay those off and buy a house or a car or whatever it is you want to, you know, save for your kids to go to college? It just is going to get unreal, and I don't know what the solution is going to be.

(Carl Hughes at 00:03:37) I mean, I do really appreciate—I went to college and, you know, it took me six and a half years, but I did get a four year degree. And I think it was beneficial for me. Do I think everyone needs it? No. I don't.

(Carl Hughes at 00:03:50) And do I think you need it, let's say a four year computer science degree to be a good software engineer? No. I don't. So I think there's a lot of room for people to—for this the—I don't know. Something's going to shake out there.

(Carl Hughes at 00:04:02) And I don't know if it's going to look like online courses or if it's going to look like some kind of augmented classrooms where, you know, fewer teachers could teach more people. We'll see. But it'll be, you know, this is a real—it's going to be a real problem in our lifetime, I think.

(Joel Beasley at 00:04:16) I think it's a—well, for me, I don't think it's a—well, the debts are definitely a problem. It's a problem when I'm at a restaurant and all the servers have higher education than I do. Right? Right. Yeah.

(Joel Beasley at 00:04:28) That's a problem. And yeah. So I think that's kind of when I started noticing it is when I started talking to the waiters and waitresses, and they all have four year degrees in business, and they all have college debt. And they're all, like you said, they're making that 30 to 50. And the answer to the question that you asked, and I know it was rhetorical, is, you know, what are they going to do?

(Joel Beasley at 00:04:47) How are they going to pay it off? It's math. They're not. They're going to die in debt.

(Carl Hughes at 00:04:52) Yeah. And in some states, that debt can even get passed to your kids, which is to me—

(Joel Beasley at 00:04:56) Are you kidding me?

(Carl Hughes at 00:04:56) Even more—That's

(Joel Beasley at 00:04:57) a thing?

(Carl Hughes at 00:04:58) No. Yeah. Yeah. Yeah. Yeah.

(Carl Hughes at 00:05:00) I'll send you some articles. You could put them in here. I have a huge list of articles I collect on the student debt crisis, because it really is the bubble of our time that no one's talking about, I don't think, or not enough people are talking about. Oh, wow. Um, but yeah.

(Carl Hughes at 00:05:13) It's huge. And, you know, even if you start to forgive it, that just means that's money the US government doesn't get and that we all pay in taxes. So at some point, you know, the rubber's got to meet the road.

(Joel Beasley at 00:05:25) Everyone's talking about the Bitcoin bubble. Let's talk about the education bubble. Right?

(Carl Hughes at 00:05:28) Yeah. I mean, there's only a few—I don't know how many people invest in Bitcoin, but there's, say, say it's thousands of people, and they're mostly in the upper middle level brackets of income anyway. So they're not, like—it's not going to hurt them when they lose a few thousand bucks on Bitcoin. Right. This is going to be hundreds of thousands of dollars for, you know, millions of people who graduate from college, which is going to be huge.

(Joel Beasley at 00:05:50) Ah, let's run away from that conversation.

(Carl Hughes at 00:05:52) Yeah. Let's get this on the happy.

(Joel Beasley at 00:05:54) What do you—I know. Let's bring it back up, people. Maybe Eye of the Tiger, Eye of the Tiger or Eminem. Yeah. So I got to put the hoodie back on now.

(Joel Beasley at 00:06:01) What are your responsibilities as the—so everyone's always asking me, hey. What are the responsibilities of CTO? And I always respond to them, look. It completely varies based on where your business is at, what your business needs, what your type of product is, what your industry—everything varies.

(Joel Beasley at 00:06:15) There's obviously some core principles that flow. But for you, how do you see your responsibilities as CTO at The Graide Network?

(Carl Hughes at 00:06:22) Yeah. Good question. I wrote an article on this probably back in 2000—oh, I think it was maybe '15 or something. But it was kind of—I was at another startup and I was in a similar role where I was kind of only engineer slash just hiring my first couple engineers and moving from, you know, only doing all code to managing a team. And I worked with some offshore, you know, people in India as well.

(Carl Hughes at 00:06:45) And so I was doing just a wide variety of things. And I was very—I don't know. I think at first, I was very uncomfortable and unsure of myself. Like, is this normal? Is this what I'm supposed to be doing?

(Carl Hughes at 00:06:56) And as I started reading more and talking to more people in similar roles with startups and small companies, I realized that this is actually—there's a whole kind of industry of people who do this. I don't know if, you know, we identify as an industry or whatever, but the role is very varied. So everything from architecture and planning, even up to product planning is on my responsibility list. I've had, you know, contractors come in and help with design or some kind of individual projects and things. But typically, all that's set kind of in my responsibility area.

(Carl Hughes at 00:07:32) And then moving into actually writing code, actually figuring out what the technical architecture of our systems are going to be and then doing a lot of that is on my plate. And then making sure that every engineer on my team is also able to work with the specifications we drew up and all that and make sure they're staying busy but also have all their questions answered and all that kind of thing. And then finally getting stuff tested, getting pull requests merged, all that's kind of my responsibility as well. So for me, at The Graide Network for the last year and a half, I've actually been the only engineer, although I've had contractors on and off as well. But in my last role with Packback, I was managing a team.

(Carl Hughes at 00:08:12) And so the roles have varied a bit, but to be honest, it's like it's kind of everything. And then when you have a team that's filling in the gaps wherever the team lacks experience. So for example, we never had a dedicated DevOps person because we were just too small to justify it. And so my role ended up being, you know, do a little code, do a little this, do a little that, and then also you're going to do DevOps because no one else knows how to. So very, why—it varies a lot.

(Carl Hughes at 00:08:40) But at the end of the day, it all boils down to, and this is something you mentioned on your blog a lot, and I really appreciate this. It's like, what does the business need from the technology, and how do we sort of merge these two worlds together? So taking everything that the business needs for the customers that we have and that is essential to our growth and saying how can the technology we could build or have built help it and improve it and kind of tying it into key metrics like that. That's all sort of in the space of the CTO at this phase of company.

(Joel Beasley at 00:09:14) Yeah. Because that's, you know, the easiest—I always like to look at where I get lost because I can't help getting lost. But what I can do is pay attention to when it happens and minimize time between it happening and me getting back on track. Right? So I noticed that whenever my brain shifts from business and value to writing a feature or a spec or something like that, it's impossible for me to maintain both simultaneously.

(Joel Beasley at 00:09:38) It's either I'm—

(Carl Hughes at 00:09:39) Very hard.

(Joel Beasley at 00:09:40) Writing the code or I'm not. So the way I do it is I use the, you know, the in and out macro micro approach where I'll go, you know, write, and then I'll come back out and remind myself of the business stuff. I put—I'm a big fan of, if you look at my personal calendar, it looks is, like, 800 events a month because I—

(Carl Hughes at 00:10:04) I like that.

(Joel Beasley at 00:10:05) People ask me, oh, what's the advice? And I mean, like, you know, it's find the, you know, the eight things that really mean something to you, like, deep down, and then putting them in front of you all the time. Like, that's the hard part. The hard—it's the hard part is not the advice or knowing what to do. The hard part is taking that simplistic idea and putting it in front of you over and over and over and over and over.

(Carl Hughes at 00:10:30) Yep. And I have a very similar system, I think. Yeah. I use a mix of Trello boards with calendar reminders and emails to myself that all just kind of force me to remember these certain things at certain times and force me to reschedule time to do things like, you know, we talked about earlier, like, keeping up with my network. I have to schedule time to do that every month or so.

(Carl Hughes at 00:10:53) And so all that stuff, if you can automate it to a point where you don't have to keep it in the forefront of your brain, saves you a lot of trouble as your responsibilities change and everything else.

(Joel Beasley at 00:11:03) Absolutely. So right now, you're the only developer or the only engineer at Graide Network, but you have contractors come in and assist you with things from time to time. Is that right?

(Carl Hughes at 00:11:12) Yeah. And I'm working on actually making my first full time hire here in the next month. So we'll have somebody else coming on soon to onboard. But, yeah, that's the situation.

(Joel Beasley at 00:11:21) Excellent. So that'll give you some more time to do things like speaking and writing. I noticed you do a good amount of writing. You're doing some speaking now. That's exciting.

(Carl Hughes at 00:11:30) Yeah. Definitely. That was one of my projects this year, sort of, was to get into speaking and learning what it took to get to conferences and get, you know, basically get your way paid to go to conferences, but also get to learn from other people, get feedback on my ideas. And so most of the talks I've been doing are pretty technical and kind of specific to certain things, but I've been, you know, sort of experimenting with a lot. So I applied to probably close to 30 or 40 conferences in the last six months.

(Carl Hughes at 00:11:59) And I got accepted to, I want to say, five. And I—scheduling didn't work out with one, so I'm doing four of them. And so, yeah, it's exciting. It'll be fun to see how that goes the rest of the year.

(Joel Beasley at 00:12:12) How far in advance do you have to apply for a conference?

(Carl Hughes at 00:12:15) About six months in most cases. It can vary. Some of the smaller conferences, maybe it's as early as three or four months away when you apply. But a lot of the bigger ones are six months to a year even.

(Joel Beasley at 00:12:26) Do they set it up to where they say, okay, you apply for a talk. And if we choose you as a talk, we pay for A, B, and C? Or do they set it up to where you apply and then you have to negotiate it?

(Carl Hughes at 00:12:37) Yeah. Usually, conferences are pretty willing to—they have sponsors, so they're pretty willing to cover air travel, your cost of the conference. Usually, they'll cover a hotel as well. There's slight variations, and I've been trying to be more selective. Just pick the ones that'll cover everything so I don't have to pay out of pocket.

(Carl Hughes at 00:12:56) We're such a small company at The Graide Network that we don't have a budget for conferences yet. So, but, yeah, most of the time they'll pay for almost everything, at least the better ones. And, you know, that part is great. But it also, you know, for me, I have friends all over the country. So for example, in November, I got accepted to speak at a conference.

(Carl Hughes at 00:13:17) And so when I didn't take their hotel stipend, I instead just went and stayed with my friend because I hadn't caught up with him since college. So nice. There's all sorts of other pros out there, you know. You get to go travel and visit other people too. Yeah.

(Joel Beasley at 00:13:30) So when are you speaking next?

(Carl Hughes at 00:13:32) Next up, next week, I'm speaking at CodeMash, which is in Sandusky, Ohio. And that's a pretty eclectic conference for web developers and software engineers. They cover all sorts of languages and topics. I think there's—it seems like there's at least five or six tracks of learning at a time going. So it's very, you know, a lot going on, which will be cool too because there'll be a lot of stuff to see.

(Carl Hughes at 00:13:56) It's also challenging because I have to pick a favorite track to go to, you know, or a few—you know, pick one talk per hour to go see, and that's always tough. But yeah.

(Joel Beasley at 00:14:05) How many people are going to be there?

(Carl Hughes at 00:14:07) Oh, I don't know, actually. Yeah. That's a good question. I imagine it's in the—it's got to be in the hundreds or low thousands because it's a big one, but I don't know for sure. You'd have to look on their website.

(Joel Beasley at 00:14:19) And there will be, like, three people in your room?

(Carl Hughes at 00:14:22) Yeah. Right. I—that's what—well, you know, you laugh, Joel, but they put me at the last time slot on a Friday. And so I have a feeling that a lot of people will be checked out by then.

(Carl Hughes at 00:14:36) I mean, I'm totally very early on in my process of doing this, so I'm not offended by any means. Getting accepted is awesome to me.

(Joel Beasley at 00:14:44) Dude, that's the win. The win is going through the process. You know, when you're the keynote, you're done.

(Carl Hughes at 00:14:50) Yeah.

(Joel Beasley at 00:14:51) It's like you've done it. You grind. You went through the grind. You had the struggle. You had the hustle.

(Joel Beasley at 00:14:56) You went through everything, and you made it to the top. And now it's just like, alright, I'm at the top now. So this is the fun part, you know, taking the crap spot and buying the bus ticket because you can't afford airfare.

(Carl Hughes at 00:15:08) Yeah. Yep. Yep. Yeah. But it'll be fun.

(Carl Hughes at 00:15:12) And, you know, even last time when I gave this talk in November, I got a lot of great feedback from people in the audience and learned things from other speakers as well about, you know, specific to the topic I was giving a talk on. And so it's great to get feedback on your ideas and kind of hear other people validate them as well.

(Joel Beasley at 00:15:28) Tell me about it. That's what I've been so pumped about the blog is because, you know, I'm sitting here on this collection of writings, and I'm talking to my editor and stuff. And, like, you know, this was a couple months ago before we started putting content out. And I'm like, I know that this is valuable. Like, I know that it is, but I'm still nervous.

(Joel Beasley at 00:15:47) You know?

(Carl Hughes at 00:15:48) Yeah. Yeah. Yeah. I think you'll get over that pretty quickly. It seems like you already have a bit, but I think I've been writing on my blog probably for five years or so.

(Carl Hughes at 00:15:59) And at first, it was a little intimidating. It was a little like, what am I doing? I have no qualifications for this. And even now, I don't know that I do, but I've stuck with it. I feel more comfortable with it even if I'm not really the smartest person in the world.

(Joel Beasley at 00:16:14) But you know what? That's interesting because I messaged out to every CTO that had registered for the book. Right? And I said, hey, you know, I looked them up. There was like 2,000 of them.

(Joel Beasley at 00:16:27) So it took me about 100 people a day for twenty days. It was like all I did, like mad science style. So I looked them up, read about them, said, alright, cool. This person looks awesome.

(Joel Beasley at 00:16:37) Let's ask them if they have any experience that they want to share, like, come on the show. Right? So I started messaging out, messaging out, and I got two types of responses back. The first one is, yeah, I'd love to.

(Joel Beasley at 00:16:47) And I was like, oh, that's awesome. And then the second one was, I don't know if I'd have anything valuable to share. My experience is going from developer to CTO. It's probably, you know, rare. Right?

(Carl Hughes at 00:16:59) Yeah.

(Joel Beasley at 00:16:59) So I started, you know, I collected all these responses, and I'm looking at them, and my mind was blown away. And there's not really any studies or anything like that because it's just such a new exploding field. Right? But from the numbers I've gathered and the people I've talked to and all this great stuff, you know, we're looking at 60 to 80% of all CTOs are developer to CTOs.

(Joel Beasley at 00:17:21) Mind. Yeah. I was thinking they were all, like, you know, medicine company executives and it's just on the 60th floor.

(Joel Beasley at 00:17:30) You know? Right?

(Carl Hughes at 00:17:32) Right. Yeah. Well, there's—I mean, there are by numbers a lot more small businesses than there are large ones. But I find that true as well, Joel. I've been a couple different CTO sort of meetup groups here in Chicago, and at both of them, most of us are pretty new to the whole management thing and the whole, you know, executive level thing.

(Carl Hughes at 00:17:52) And we're all just figuring out how it is to make that transition away from being just a hands-on developer. I say just a hands-on developer, but, you know, it just feels like a totally different world.

(Joel Beasley at 00:18:03) That's honestly one of the biggest themes in my book because—and it wasn't even intentional. It's just what happened, because, you know, I meet with these CTOs. I work with some venture capital companies, and they go in and inject capital. And usually, it's a developer and a business person or a marketing person, and they're going to, you know, build a business from their MVP or prototype. Right?

(Joel Beasley at 00:18:23) And the hardest thing for the developer to do is to become the CTO, because they'll wear the label of CTO. And then when you start probing them about CTO things, they're just like, I don't know. And so one of the pivot points or the crossroads that I've, you know, seen recurring over and over is directly tied to their ability to requantify their value.

(Carl Hughes at 00:18:50) Yeah. No. I was going to say that you have to measure yourself in completely different metrics than you did as a developer. So lines of code are not important. The, you know, quality of the code coming out, I mean, that does matter, but it's also not your number one metric.

(Carl Hughes at 00:19:04) Like, what matters is keeping the people that work for you happy and keeping productivity high and, you know, keeping in line with what the business needs. And so it becomes a totally different mentality. And when I was managing a team and less in this role where I'm the only developer, what I would do is block out some time in the day to code and then other times of the day to do basically just people and management type stuff. And I had to have a separation because those two things you kind of mentioned are very different in your head, and keeping them both on, you know, on tap at the same time is almost impossible.

(Joel Beasley at 00:19:40) Yeah. It's a big culture shift, not just in our world, but in the global economy where, you know, in the fifties, we imagined that you would go get a job and work the same company for sixty years and retire. You had one job, you did one thing, and that was it. Right? Now we're an economy where people are programming in the afternoon, Ubering at night, barista-ing three days a week.

(Joel Beasley at 00:20:01) Like, you have these multiple ideas and these multiple skills that you use, and that's just a totally different economy. Totally. And one of ours, the one that we're running into is that transition from developer to CTO. Yeah. What are you talking about at—first of all, I know it's CodeMash now.

(Joel Beasley at 00:20:19) But this whole week, I've been calling it CodeSmash. I don't know why I've been so pumped up. Like, every day, I'm like, yo, we're going to talk to Carl from CodeSmash. And then this morning, Jake's like, dude, you got to stop saying that.

(Joel Beasley at 00:20:30) It's CodeMash. And I'm like, oh, man. It needs to be—we need to—and then the icon needs to be a fist. So I'm going to talk to the branding people over there. It's going to be called CodeSmash.

(Carl Hughes at 00:20:38) Yeah. Send them an email. I don't have much sway yet. Maybe next time.

(Carl Hughes at 00:20:44) Yeah. So what I'm talking about—so I have a few talks kind of that I've been submitting. One of them—the one that's gotten the most acceptances and has kind of, I guess, so I guess it, you know, ends up being the most popular from that perspective. But it's about testing distributed systems. So for the last few years, when I've been working with startups, we've built stuff on microservices.

(Carl Hughes at 00:21:06) And so the technical challenge there is not only do you have to test each microservice internally, but you also then have to test them in conjunction with one another. And those testing across those boundaries is to me the most interesting and challenging part of architecting tests for that sort of setup. So if you're a developer and you've been writing unit tests, you got no problem with, you know, mocking stuff out and writing up a unit test and making your assertions, and it all is good. But once you've got, like, five or six microservices feeding into a single page front end application and maybe some proxy layers in there, how do you test all that? And it does get to be pretty tricky.

(Carl Hughes at 00:21:45) So in the talk, I kind of go over some of the, like, less on the implementation of actual tools that do it, more on the mentality and the philosophy of how you break that system up in the layers and think about it in terms of this kind of test versus that kind of test.

(Joel Beasley at 00:22:01) So you kind of, you know, you give the value through the heuristic. Think of it like this.

(Carl Hughes at 00:22:06) Yep. Exactly. Yeah. I mean, because at these conferences, you know, everybody's writing code in a different language and uses a different tool set. So I'm not trying to tell them, you know, this is exactly how you do it in Java and PHP and JavaScript. I mean, those are all just so wildly different that I couldn't cover that in one hour. But to give you, like, an idea of, like, here's at a high level what this system looks like, and here's how the testing layers fit into it. That can be done.

(Joel Beasley at 00:22:30) Yeah. That's why, you know, I was a big fan of Martin Fowler, Ben Orenstein. Yeah. And Metz, like, those types of people because when I'd read their content while, you know, Ben and Sandy were Ruby, you have Martin over in, like, Java. And but the point was, yeah, the code samples are there. You could virtually—like, they're—you don't need to know necessarily about the code.

(Joel Beasley at 00:22:52) It's more about the structure and the heuristics of putting it together.

(Carl Hughes at 00:22:55) Right. And if you're out—as an advice for anybody who's looking to speak at conferences, if you're looking to speak at conferences, it helps a lot if your talks are language and framework agnostic. So the reason is there's just a ton of conferences that take general talks. There's only a few that take, say, PHP or even more specific one framework in PHP or, like, a content management system like WordPress. You know, there's just only so many places you can go with that.

(Carl Hughes at 00:23:21) But if you go into something a little more general, it's a lot easier to get accepted.

(Joel Beasley at 00:23:24) So I shouldn't go publish my own micro library on GitHub and give a talk just on that.

(Carl Hughes at 00:23:32) Yeah. I mean, it's funny because I actually—some of the talks at conferences are basically that. Usually, people can only pull that off if they work at a place like Google and they have enough, like, swing behind them that people just want to see them talk about anything or if they're a conference sponsor and then they can kind of talk about whatever they want. But for those of us who are unknowns, it helps to have a more generally applicable talk.

(Joel Beasley at 00:23:52) But I have three stars on my project and only two of the accounts are me.

(Carl Hughes at 00:23:56) Right. Yeah. Yeah.

(Joel Beasley at 00:23:58) I was looking at your—I love that when you're speaking at this conference, you then went and wrote content and produced content about the topics you're most excited to see at the conference.

(Carl Hughes at 00:24:09) Yeah. Definitely. I mean, I'm a big fan of, like, if a conference is going to do me the favor of allowing me to come speak, then I should do at least a little bit to help market, you know, what they're doing. And plus, it's just an easy idea for content. I think one of the hardest things when you have a blog for a number of years is keeping good ideas flowing, you know, and stuff that you're passionate about writing a little bit about.

(Carl Hughes at 00:24:34) And so, you know, some months I'll come up with a ton of topics. Some months I just barely have anything. And so this was like—I've always done this at conferences that I'm going to or speaking at. Just write a little pre-conference blog post about the five or six talks that I'm most interested in seeing. And usually, I'll mention the people giving those talks on Twitter just to say, hey.

(Carl Hughes at 00:24:53) You know, here's a shout out. Like, thanks for coming to the conference and speaking. I'm looking forward to it.

(Joel Beasley at 00:24:58) Yeah. I use the, uh, Gary Vee style document don't create. Have you heard of that?

(Carl Hughes at 00:25:03) No. I haven't.

(Joel Beasley at 00:25:04) Alright. He's this business media guy, tech, and he's early investor in a lot of the tech companies, but he's mainly like a business guy who's really good about selling and communicating and super, super smart. So he always talks about document don't create, and that really helped me get over my hump with the book as well because what I had been doing was sitting down trying to come up with good ideas to share. And instead now, I just—I look at whatever's happening to me, whatever, you know, because every—all the issues recur.

(Joel Beasley at 00:25:36) Right? There's nothing new.

(Carl Hughes at 00:25:37) Right.

(Joel Beasley at 00:25:38) I'm not—that was the main premise of the book is, like, I'm not running into new stuff. It's everybody making the same mistakes, and I made the same mistakes too. And then I figured out how to deal with them. And I don't make those mistakes. I'm definitely not perfect, but, like, you know, I want to communicate this.

(Joel Beasley at 00:25:52) So now I'm just kind of writing about—I let whatever's happening around me inspire me, and then I don't care if I cover the same topic in a different context with a different story. So in your post, you listed the five talks that you're most excited to see at CodeSmash, uh, 2018. It appears it's like a monster truck jam. We got Hulk hands, and I'm just—I'm building it up in my mind. I apologize.

(Joel Beasley at 00:26:21) If anyone goes to this, they're going to, like, be so disappointed. They're so upset. They're like...

(Carl Hughes at 00:26:27) I don't know. Maybe they won't. It should be fun anyway.

(Joel Beasley at 00:26:30) And then they'll hold out, like, hashtag wait for it, and then they'll see Carl's talk at 5 PM on the day everybody—after all the flights that have happened. And then they're going to be like, it was worth it. Carl told us about distributed systems.

(Carl Hughes at 00:26:42) All worth it.

(Joel Beasley at 00:26:43) You know?

(Carl Hughes at 00:26:45) Right. Yeah. There's nothing better than listening to a talk about testing at 4 PM on a Friday.

(Joel Beasley at 00:26:51) Right? So sexy. Like...

(Carl Hughes at 00:26:54) I know I'm not getting any sexy, you know, points for this, but that's alright. It'll be good anyway.

(Joel Beasley at 00:27:01) Here's the talk I want to see. You ready for it? I want to see neural networks that generate Game of Thrones scripts. Right.

(Carl Hughes at 00:27:09) Yeah. So have you done any work on machine learning or some just basic natural language processing stuff by chance, Joel?

(Joel Beasley at 00:27:18) Well, me, it's very little. I booted up a couple projects, played with it, taught it some, like, Robert Frost poetry, and was like, okay. Yeah.

(Carl Hughes at 00:27:27) Yeah. So there's this idea—apparently, I've read about before—Markov chains, which allow you to sort of—basically, you have the computer read, quote unquote, a bunch of text, and then the Markov chain will predict, sort of the next word, given a specific word. And so it uses the text that it knows about to predict what the next word might be. So you could do this.

(Carl Hughes at 00:27:47) I've seen it done for, like, Harry Potter...

(Joel Beasley at 00:27:49) Uh...

(Carl Hughes at 00:27:50) Style books. We got, like, ingested all the Harry Potter books and then wrote a Harry Potter book from its computer, which is cool. This is apparently the next level of that, which uses neural networks, which are, you know, that kind of unsupervised machine learning technique. And that's about the edge of my understanding. So I'm excited to see the rest of it, how it goes.

(Joel Beasley at 00:28:10) Right. I just hope that, like, on the slide that he uses that he uses the picture of the blonde dragon girl from Game of Thrones. Her storyline was the most interesting because I like seeing the dragons. Like, I like the cool dragons, like, hung out with her and stuff. And I'm like, I want, you know, less incest and more dragons.

(Joel Beasley at 00:28:28) Like, can we—let's do a little bit of that. Right?

(Carl Hughes at 00:28:31) Right. Well, you know, Game of Thrones, everybody's got what they come for, so it's got to keep all the fans happy, I'm sure.

(Joel Beasley at 00:28:37) Pun intended, I'm sure. Alright. So Markov chains. Oh, by the way, speaking of blondes, we actually have coming on the show in the next two weeks, we have Coding Blonde, and she's—that's her Instagram handle.

(Carl Hughes at 00:28:52) Okay.

(Joel Beasley at 00:28:52) Coding Blonde. And she is a blonde programmer who was working at Google, very smart. And I was looking through her posts, and I was like, oh, sweet. This is good. And then I started looking through other posts of influencers, I guess, people with a lot of followers and what they're doing on Instagram.

(Joel Beasley at 00:29:12) I wanted to feel what the culture was like for programmers. Because I'm 30. How old are you?

(Carl Hughes at 00:29:19) 29. Right there with you.

(Joel Beasley at 00:29:21) So we're right there. I had just started my family. I just had my baby girl like three months ago, first child and all that stuff.

(Carl Hughes at 00:29:27) Congrats.

(Joel Beasley at 00:29:29) Yeah. Thank you. And so if a high schooler looked at me, they'd say he's old. Right? Which is crazy for me because I still feel the same as I felt when I was like 13 in my mind. Right?

(Carl Hughes at 00:29:39) Totally.

(Joel Beasley at 00:29:40) And so it's blowing my mind how life is working. It's unbelievable. It's very fascinating. So I'm thinking about what the next generation of these people are doing. I was hardcore programming 13 through 18 all the time. Right? And so I'm wondering, like, what are these people doing now? What would Joel be doing then? Because we did not have any of this stuff. We did not have Facebook. We did not have Instagram.

(Joel Beasley at 00:30:05) We did not have the iPhone. We did not have any of this. Right? We had like the Nokia. You know? Like, come on now.

(Carl Hughes at 00:30:12) Yeah. You could play Snake on that phone. That was the best part.

(Joel Beasley at 00:30:14) That was it. Right? So I'm wondering, alright, well, now that everyone can hang out online, let's go and do whatever wherever their communities are and see what they're doing. And so I looked them up on Instagram, and that's a popular community for that age. Well, I mean, I'm on Instagram personally anyways, but I wasn't ever going into the culture of programming on Instagram. Right? So I started to go peek into that corner. Here's what I found. Okay. There's about ten, fifteen people in the programming space with those type of hashtags that have over 20,000 followers. Right? Only like 10 or 15 people, at least in my country. Right? At least in the United States.

(Joel Beasley at 00:30:57) Right. And eight or nine of them are girls. Overwhelmingly female.

(Carl Hughes at 00:31:04) Okay.

(Joel Beasley at 00:31:04) Overwhelmingly female. And I was like, that's awesome. Like, that's so—yeah. Like, who's poking around looking at that? Like, I haven't read that article yet.

(Carl Hughes at 00:31:12) Yeah. Yeah. It's interesting you mentioned that because one of the things I've noticed from speaking at these boot camps and meeting people who graduate from boot camps is the diversity level tends to be a lot higher than what I saw going to engineering school a few years back. And that's both women, it's people of color, it's people from poor backgrounds. It's just all levels of diversity is better coming out of boot camps.

(Carl Hughes at 00:31:37) And I wonder if—I don't know if that has anything to do with Instagram at all. But I do think it's a really exciting trend because it's like I finally have job applicants who are not just exactly like me and look exactly like me and think exactly like me like it's been for the last few years. So I applaud it. More power to them.

(Joel Beasley at 00:31:57) Yeah. I super don't care about—you know, it's quite interesting for me because I care so little about it. But at the same time, I'm like, oh, I kind of have to care about it. And I'll give you some context here. I care about the quality of someone's work and their attributes as a motivated, accomplished person. Like, I don't care about your height, color, size, weight, sex.

(Joel Beasley at 00:32:22) Like, it doesn't matter to me. What matters to me is that you can produce, that you can produce value and grow what I'm trying to do and help me on my way to helping others. Right? Because that's what I'm doing. And so I just try to surround myself with people that produce value.

(Joel Beasley at 00:32:37) And so I started saying, like, ah, it doesn't matter. Just find the people with value. But then I realized I actually came across it. Like, you know, because I am—anthropomorphism, and you'll like this. Right? If you don't know what it means already, because you're super smart. It's when you associate that everyone else around you is similar to you. Right? So we do it with our dogs. We imagine they feel certain ways.

(Joel Beasley at 00:33:00) Oh, I feel so bad for that. We do it with inanimate objects. We do it with everything, anthropomorphism. So I thought everyone's kind of like me. Then I realized, dude, there are actually sexist, racist people.

(Joel Beasley at 00:33:13) Yeah. And I came across—like, I was in a meeting one day. Yeah. This is seven years ago, right, when I had this realization. I'm in a meeting one day, and they're talking about hiring someone.

(Joel Beasley at 00:33:24) And like, they actually didn't want this person because of one of their attributes. And I'm not gonna say which one it was because it's such a variety, and it could be anything. But I, like, no joke. Jaw hit the floor. I went out of that meeting, and I texted this one friend of mine who's very into advocating for that group.

(Joel Beasley at 00:33:44) And they just responded back to me. I was like, yep. Because I've been talking to them seven years ago, and I said, ah, it's not a big deal. Like, I was like, this is the modern age.

(Joel Beasley at 00:33:53) No one in my group, in my generation is like that. We're modern and coming up. And then I saw someone was like, oh, man. I was like, alright, so now I'm not gonna say it doesn't matter. Now I'm not gonna say—instead, I'm just gonna constantly set the example for validate people off of the quality of the value that they bring you and what they do, and double down on that.

(Carl Hughes at 00:34:16) Totally. Yeah. And it's not only at the point of hiring, you also have to worry about it from like, does the whole work environment we're creating, is it allowing us to get and keep the best people? You know? So like, I don't wanna be in a work environment where I have to worry if any woman who joins the engineering team might wanna leave in six months because it's hostile to her.

(Carl Hughes at 00:34:39) Like, that would be a shame, especially if she's a great engineer.

(Joel Beasley at 00:34:42) Right.

(Carl Hughes at 00:34:42) But even if she's just an okay engineer, it's expensive to hire people. It's expensive to onboard people. Turnover. I don't wanna lose them in six months. Yeah.

(Carl Hughes at 00:34:51) So there's a lot of selfish reasons to be pro, you know, having a diverse and friendly to all workplace. And so I'm a big advocate for it. And anyway—

(Joel Beasley at 00:35:03) So the rainforest—like, some of the world's largest rainforests in science, take any collection of organisms, and what makes them stable is their diversity.

(Carl Hughes at 00:35:13) Right? Yep.

(Joel Beasley at 00:35:14) Right. So I love that aspect of it, like the rainforest being filled with so many different living things is what keeps it alive.

(Carl Hughes at 00:35:21) Right? Yeah.

(Joel Beasley at 00:35:22) So I love it. So when you were writing a little bit about and talking a little bit about marketing products. Yeah. Right? So how do you feel—do you have your hand in marketing over at your company or no?

(Carl Hughes at 00:35:39) Not too much. You know, we don't do a ton of marketing, true marketing. We do a lot more direct outreach and sales kind of—that's how we get our customers. So I'm less involved now. But because I've worked with consumer-facing companies quite a bit in the last few years, I've gotten to see a lot and understand a lot about marketing.

(Carl Hughes at 00:36:02) And so I always think it's an interesting space. It was also—you know, when I very first started this journey of figuring out I wanted to be a programmer and all that, I kind of thought I might wanna do more of a marketing automation kind of role than a true programming CTO role. And so I explored it a lot more a few years back. And then, you know, now I kind of regularly keep up with what's going on in marketing, especially for startups and side projects and things like that.

(Joel Beasley at 00:36:29) You're gonna love Gary V then.

(Carl Hughes at 00:36:32) Yeah. And it's—Gary is it Vaynerchuk or something like that? Yeah. Yeah. Yeah.

(Carl Hughes at 00:36:38) So I have heard his name. I don't think I follow him directly right now, but I definitely have heard of him.

(Joel Beasley at 00:36:43) Yeah. So we actually took—I've seen him for a couple months, and we took his information, and we're executing it with the podcast. And for no reason other than to just share value with people. And I always call myself like a life scientist, because all I do is go around testing things and seeing what works and then doing more of them and then figuring out what doesn't work and then moving forward. You know? So I said, oh, well, you know what?

(Joel Beasley at 00:37:10) Let's take this guy who is clearly growing his company because it's huge. Right? And I've watched him grow it. And let's take the—and he sits up there, and he goes, this is the advice. This is how you do it.

(Joel Beasley at 00:37:21) Right? This is exactly how you do it. And I can tell you all of this because none of you will do it.

(Carl Hughes at 00:37:27) Right. The execution is the hard part.

(Joel Beasley at 00:37:28) Yeah. Oh, you know what? I was talking about that, with the execution of the very simple things over and over and over. I learned how to execute. It was difficult, and I'm always working on it and growing on it. It's always a thing that I'm working on. But I said, you know what? Let's take this guy's advice and execute it. Boom.

(Joel Beasley at 00:37:44) Eight weeks, like, tremendous results.

(Carl Hughes at 00:37:47) Okay. Cool. Well, yeah, it's like you don't hire people because they've read the most books or had the most theoretical understanding of things. You hire them because they have real hands-on experience doing stuff, and that's the hard part of any job.

(Joel Beasley at 00:38:02) Oh, yeah. We're always battling over here thinking, like, whenever we come up with an idea, what we'll see is, like, alright. Let's do this. And then instantly, the next words out of your mouth is a queue that builds up of the things that have to be done before you do it. Right?

(Joel Beasley at 00:38:17) Yep. We're like, oh, let's livestream. Oh, we need this. Oh, and now we need this. Oh, and now we need this. Now we need—I'm like, it got out of control. You know? And—

(Carl Hughes at 00:38:27) Yeah.

(Joel Beasley at 00:38:27) We boiled it down, and I said, look. You know, and we did it together because I do everything with my team. I said, let's just say, like, how do we get the image on the screen? Like, let's only do what's required. Let's MVP this thing and only do what's required. Zero.

(Carl Hughes at 00:38:43) Yeah. Yeah.

(Joel Beasley at 00:38:44) And then so we did it. So do you have any—any other than the speaking and the writing, which—

(Carl Hughes at 00:38:50) Or—

(Joel Beasley at 00:38:51) Or you can have video of that. Right? I wanna see your talk.

(Carl Hughes at 00:38:55) Yeah. So, occasionally, they record video. I don't—I haven't asked the people at Code Mesh, but if nothing else, I'll record a kind of a screen and audio capture if that's all I have available. So I'll have some recording of that. There's a couple of recordings I have on my blog. There's a tab at the top for speaking, I think, that may have some videos too. So, getting some recordings there. Yeah.

(Joel Beasley at 00:39:17) So there's Amazon Basics, you know, super cheap. They have a $10—like, $12 tripod with a $10 iPhone attachment thing.

(Carl Hughes at 00:39:28) Yeah. I was looking into that, actually. That was my first thought. And then I was like, well, I don't know if I wanna have one more thing to deal with and make sure it works. Yeah.

(Carl Hughes at 00:39:38) And carry it. So I could do that too. I'll think about it for next time.

(Joel Beasley at 00:39:43) Oh, I just wanna see you. Like, I wanna see the talk. Yeah. You know, like, if you put your phone in Do Not Disturb, then you can use it pretty effectively as long as you make sure your Do Not Disturb settings are really strict because your wife will call you three times and the recording will stop.

(Carl Hughes at 00:40:03) Yeah. Totally.

(Joel Beasley at 00:40:04) Amazon Alexa, you playing with that?

(Carl Hughes at 00:40:07) You know, I have not. I play a little bit with Google's Assistant because I have an Android phone, the Google Pixel, and they're pretty big on having their assistant do things. It does not work for me yet. It is not robust enough to fit into any workflows that I use at this point. I think there's a lot of potential there, but like, when I say something like, you know, add a to-do list item, Google wants to add it to its to-do list thing that I don't use.

(Carl Hughes at 00:40:37) Like, can I—I would like to be able to customize that. And I hear you can do more of that with Alexa. Is that right?

(Joel Beasley at 00:40:42) Yeah. Alexa's pretty amazing. We have both at the office. Okay. And we're primarily using them for hype music.

(Carl Hughes at 00:40:52) That's worth one.

(Joel Beasley at 00:40:54) Yeah. Maybe we'll have some Eminem on one and then some Eye of the Tiger on the other. Oh, and then, of course, and then we'll queue up with like a third one, Thunderstruck, so we can pan, you know, switch to it after it gets upbeat enough. Yeah. Alright.

(Joel Beasley at 00:41:08) Anything else that you want to communicate to our group, like, to these—we got modern CTOs, you know, crazy amount of us, and most of us went from the developer to CTO. Do you have any—maybe any advice or any other thing you wanna communicate?

(Carl Hughes at 00:41:26) I, you know, I think we've covered a lot of what I always give people as advice. It's just, you know, that networking is a huge part of this and meeting people and knowing people, staying in touch with people, but also, you know, reading and learning and staying on top of what's out there, but not getting too caught up in it. It's a daily struggle. But, yeah, if anybody wants to continue the conversation, I'm always happy to chat whether that's just a little Twitter interaction or it's a longer email or somebody wants to hop on a phone call, all that's good. And you can put whatever contact info you have for me in the show notes, but Karl L. Hughes on Twitter and GitHub and most other places.

(Joel Beasley at 00:42:04) And please do not send dev-related questions or CTO-related questions to the actor, Carl Hughes.

(Carl Hughes at 00:42:10) Right. Yeah. Carl Hughes with no L. He's probably not me. He's the guy in England who's an actor.

(Carl Hughes at 00:42:15) If you're interested in an actor, though, he is the one to go to. I am not. So there's a Carl for everything.

(Joel Beasley at 00:42:24) That's awesome. How great would it be if Carl Hughes, the actor, started looking into programming because everyone kept reaching out to him?

(Carl Hughes at 00:42:34) That son of a gun. As long as he doesn't change his branding—I mean, he's got an IMDB page. I can't compete with him.

(Joel Beasley at 00:42:40) Show up at your talk, and he's gonna start arguing with you about testing.

(Carl Hughes at 00:42:45) Right. That would be a sad day for me.

(Joel Beasley at 00:42:47) If you had your camera rolling, though, you're gonna be YouTube famous. Right?

(Carl Hughes at 00:42:52) Right. Well, that's true. Yeah. Yeah.

(Joel Beasley at 00:42:54) Alright. Thank you so much for coming and sharing.

(Carl Hughes at 00:42:57) Thank you, Joel.

(Joel Beasley at 00:42:58) It's been—like, dude, I love calls like this. Carl is awesome.

(Carl Hughes at 00:43:02) Oh, I had a great time too, Joel. And I look forward to listening to this episode, but other episodes of the podcast as well as the book. And does it have a due date?

(Joel Beasley at 00:43:12) Yeah. So we just did editing, and we're thinking that the safe date is March, but we're imagining that's gonna start shipping in February. Thank you so much for listening to the Modern CTO podcast. Share this. Get the word out.

(Joel Beasley at 00:43:32) Thank you guys so much. I couldn't do it without you. I appreciate it. You guys are the absolute best.