Joel Hron - CTO at ThoughtTrace

(Joel Beasley at 00:00:00) Hello, my friends. Today we are talking to Joel, the CTO at ThoughtTrace, and we discuss the evolution of the CTO role as a company scales, the importance of practicing patience when recruiting, and the ethical decisions surrounding the advancement of artificial intelligence. All of this right here, right now on the Modern CTO Podcast. Here we go. This is the Modern CTO Podcast.

(Joel Derfner at 00:00:35) Hello. Hello. Hello, Joel.

(Joel Beasley at 00:00:37) Joel, this is so much fun. It's like two Joels, right? What's going to happen today?

(Joel Derfner at 00:00:43) This is a first or what?

(Joel Beasley at 00:00:45) Kind of. We had one. I told you it was the first in the prep meeting, but then I remembered we had one other Joel, and there was an audio issue, and so we didn't actually air that episode. But yeah, he was the CTO of UCF, the college in Florida.

(Joel Derfner at 00:01:03) Oh, yeah. Yeah.

(Joel Beasley at 00:01:04) Yeah. Really smart guy. So he did our name well. But yeah, sometimes the episodes just don't air because of the audio or lag or drift. But we've gotten much better at the technical aspect of it, so it should be smooth sailing from here on out.

(Joel Derfner at 00:01:21) Yeah. You learn every time a little bit, I guess.

(Joel Beasley at 00:01:23) We could do the whole episode on that, just that one seed of a thought, right?

(Joel Derfner at 00:01:30) Yeah. Yeah.

(Joel Beasley at 00:01:31) Oh, man. So I guess a good place to start—something I'm actually interested in because I saw a little bit about what you're doing now—but I was curious about where you got your start. How did you get involved with technology as a kid? Did you get involved as a kid, or was it in high school or college? When did you first fall in love with technology?

(Joel Derfner at 00:01:50) That's a really good question. I probably, I would say, had a nontraditional route, although I think everybody has somewhat of a nontraditional route. But I would say really college more than anything. And it was not so much technology at the time as it was engineering, which is my background. I was doing a lot of computational simulation work, mathematical modeling. Actually, I did some internship work with Lockheed through college, working on the F-35 program and things like that.

(Joel Derfner at 00:02:26) And then went to grad school and took a little bit of a different path but started working on sonar systems and doing a lot with signal processing and some more mathematical modeling stuff. And so that really, I think, more than anything, charted my path into the machine learning world. So software, strictly speaking, is not so much my background. It's more on, I would say, the analytics and machine learning side that I've got more of a background. And like I said, it was more of a stepping stones, starting with more traditional engineering sciences and then kind of migrating more into the computer science realm over time.

(Joel Derfner at 00:03:09) And then, obviously, when I joined ThoughtTrace, initially working on the AI machine learning part of what we do, but ultimately taking on the software engineering and product aspects of the company too. So it's been a consistent learning experience for me, which is awesome. It's what I like.

(Joel Beasley at 00:03:28) Yeah. How did you approach it? Were you like, I'm going to learn every single thing I possibly can about this software engineering, or I'm going to find some really bright people and empower them? How did you go about solving that part where you didn't have as much experience as the mechanical engineering?

(Joel Derfner at 00:03:48) Yeah. I would say more the latter in terms of bright people. I've got a great team of people, not all of whom I directly hired. Some of whom were here when I got here and some of whom we've hired since. But certainly building a good team of people around you and encouraging their growth and leaning on them for what they're good at and focusing on what you're good at, I would say it's been core to my strategy in terms of leading the team and growing the company. But certainly there's that personal growth aspect of it too, and trying to be a good listener and ask a lot of questions of my team when I don't fully understand things and just watching what they do and being maybe thoughtful about when I challenge and how I challenge certain things.

(Joel Derfner at 00:04:44) I think it's been helpful to me in terms of my own growth as well as the growth of our team as a whole.

(Joel Beasley at 00:04:51) Very important, right? Yeah. Learning how to listen is a tough thing.

(Joel Derfner at 00:04:57) It is. Yeah. Yeah. Especially when you come from a scientific background when you're bred to be correct. You know what I mean?

(Joel Derfner at 00:05:05) You just get value out of being right and knowing answers. And so you immediately think that's your job and everything that you do is to have an answer. You know? And so from a scientific perspective, it's very foreign to approach a situation and just not try to give an answer. You know?

(Joel Beasley at 00:05:27) Oh, man. For me, personal growth-wise, letting go of my need to be right was something I worked on majorly in my twenties because it can work when you're obsessed with work, which is how I was then. I was just eat, sleep, breathing engineering for the better part of two decades, right? And so you're working with computers and you have logs and you tell the computer what to do constantly. So you're issuing thousands of commands a day, right? Getting very explicit responses, and you can log and back-trace anything to understand exactly why it didn't work. And then you leave that environment and you go interact with your spouse or a girlfriend or another human. And it's like I could literally feel my brain shifting gears.

(Joel Beasley at 00:06:14) And it was so tough to play both of those roles and develop the human skills side of things.

(Joel Derfner at 00:06:22) It is. And when you don't do it right, it blows up in your face, and then it blows up in your face enough times, you kind of learn to adjust, and that's really where I come from.

(Joel Beasley at 00:06:33) You nailed it right there. It's when the pain is greater than the change, right?

(Joel Derfner at 00:06:38) That's exactly it.

(Joel Beasley at 00:06:40) I like you, my friend. We have—we're learning a lot of the same lessons, right?

(Joel Derfner at 00:06:45) Yeah. Yeah. I mean, it's definitely a learning experience, and that's what's been really fun for me is just to kind of not just learn from a technical standpoint or from a team building standpoint, but just, you know, my role itself evolves over time as well. Right? I come into the company—we were a startup. We had two customers, I think, and a handful of people working on the product. And so you're very hands-on, trying to tinker and figure out what's going to work and where you need to be going. And a lot of trial and error at those stages, but in a more hands-on way, and kind of evolved from that into very recruiting-oriented for a long time, right?

(Joel Derfner at 00:07:29) So trying to—okay, I've got something that works. I need to build a team around me and build up the team that's going to ultimately carry this forward and help us scale. To one that's maybe more customer-oriented and product-oriented, right? We're spending more time in front of prospects and customers and trying to define the longer-term strategy. And then one that's more organizational, right? I need to figure out process-wise and organizationally how to empower these teams to be independent but still stay aligned on what our long-term vision as a company is and what our product strategy and vision is. And so just sort of ebbing and flowing into those different needs as a contributor to the company has been fun for me because, you know, come to work one month and this is what I need to be doing and this is my top priority, and the next month it could be something completely different. And so it's been a really fun time adjusting to those different needs as we've grown.

(Joel Beasley at 00:08:33) I actually put a recurring event in my calendar that triggers every couple months that reminds me to go revisit that, just because yeah, it's like my step back. Obviously, okay, we're all perfect. We should be doing it every day or week or whatever it is. But for me, I just, every three months, I put this recurring event that says, "Are you doing what's effective? Are you doing what you love?" And just two or three basic questions. And that causes me to have to step out and evaluate where we're at. And it's not easy to do either.

(Joel Derfner at 00:09:08) No. It's not. It's—I kind of think of it as I'm trying to put myself out of a job at all times. You know what I mean? I think that's human nature. We build things to automate tasks and free us up to do other things that we wish we could do. That's why we invented things like tractors and travel and all this kind of stuff, right? And so I kind of approach my job from that same perspective. If I find myself spending a lot of time doing something, I feel like I should be figuring out a way to not spend so much time doing that.

(Joel Derfner at 00:09:43) And that might be building some process or building a team around it. And certainly there's somebody in the company who can probably do that thing better than I'm doing it by myself, right? And so I try to approach it from that perspective, and I think that's a helpful way to really have the company and have the team own those things more long-term rather than being single-threaded on them.

(Joel Beasley at 00:10:09) Can you give me some perspective as far as the size of the company and what the mission is?

(Joel Derfner at 00:10:14) Yeah. For sure. So we are just shy of 70 people today. We released a product in 2017. We are a document understanding platform. We actually have a new release coming out at the end of this month, really kind of recasting our platform into this document understanding framework. But from our perspective, document understanding really sits at the middle of what content management systems do as well as what newer-age contract analytics platforms do. And so contract analytics being the use of AI and machine learning to essentially read and interpret and extract relevant provisions and data from contracts to be able to ultimately take action on those things, right? And that was really where we got our start, was in the analytics space really with AI machine learning as it pertains to documents and contracts.

(Joel Derfner at 00:11:15) And, you know, as we kind of evolved in that space, we sort of saw that the management of those documents is inseparable from that. I can't analyze this thing and not manage it also, right? And so that led us to really create what we think is a really seamless integration between those two needs from an enterprise standpoint. And that's really where our focus is now.

(Joel Beasley at 00:11:42) That's exciting. I actually built some mock-up type beta version one applications for some law firms, you know, about seven or eight years ago, that was doing essentially contract analysis. It would extract from the—what was the word? The complaint or whatever the lawsuit was. It would extract information out of the lawsuit. Do you—it didn't progress. We built a version one prototype, and then we just didn't end up doing anything with it. I just did it as a contract software developer. But do lawyers use your product? Is that in the wheelhouse of what you guys are doing, or no?

(Joel Derfner at 00:12:23) Yeah. It is. I mean, we have a lot of lawyers who use our product. I would say more of the focus for us is direct sales into companies and businesses and enterprise rather than legal services firms and things like that. And so our users tend to be more business analysts and business users who are responsible for compliance with the contract or managing assets associated with those contracts.

(Joel Derfner at 00:12:53) Not so much legal work specifically, but in some cases that does lead its way into legal work. But, you know, things like diligence on acquisitions and divestitures are very common things for us where we get involved on the legal side as well. But really, the suite of problems are not just legal in nature. Legal, honestly, is an important but a pretty narrow scope of, I think, the whole scope of problems that exist around documents and kind of the information that's buried inside of them. You know?

(Joel Beasley at 00:13:28) And so what is one of the coolest use cases that you see happening a lot?

(Joel Derfner at 00:13:34) Yeah. That's a really good question. I mean, so we've been in the process of—we started with pretty, I would say, narrow industry focus, but over the last year or two have really grown into other markets and other use cases and industries as well. From our perspective, use cases around land and title and things like that where you literally have an individual going into a courthouse, pulling files out of file folders from ten, twenty, thirty, forty, fifty years ago, trying to create run sheets and things like that. I just saw, you know, a demo one of our subject matter experts gave this morning on our new product of literally doing three or four button clicks and creating a title run sheet in our application. You know? So there's some really, I would say, I want to call it simple, but some really archaic methods of doing simple tasks, like title, for instance, that are just awesome to see, really fun to see. And then there's really complex things where somebody's doing an acquisition of a billion-plus-dollar asset, and they identify $100 million in defects that they can go call back or fight back around. So there's the high end and there's the kind of big scale of how our application gets used, and then there's the real day-to-day improvement of process efficiency.

(Joel Derfner at 00:15:12) And so it's fun to see both ends of that spectrum.

(Joel Beasley at 00:15:15) So I don't know a ton about the space, but as you're describing it, it seems like across contracts there would be some similar things across all of your verticals, all of your different use cases, like participants or individuals mentioned in the document, right? That could be something that would be useful across all of them. Here are all the people that appear in this document, right? But then there's going to be, inevitably, very specific things that you're going to want to look for in certain verticals. So do you set up teams and prioritize how you build—I guess the parsing or the insights? You create insights from the documents. So you have to program the insights, and I'm assuming insights are vertical-specific. So how has that shaped the organization of your development team?

(Joel Derfner at 00:16:04) Yeah. That's a fantastic question, actually. That's one of, I think, the most unique things about our company and probably one of the most differentiable things, actually, is how we blend the knowledge of our subject matter experts with our data science team. And so some of that is tools. We've built internal tools for our subject matter experts who are typically lawyers in the industries that we work with, kind of ex-users of our software, if you would, who now work for us and support us in building those models, but also do a lot of work with our customers.

(Joel Derfner at 00:16:44) But they kind of have the insight and vision for what the needs of the user are from a contract interpretation standpoint, how that information will be used. And so they're critical in that development process. And having them work in a world of data science and machine learning is really foreign to them for obvious reasons. And so we've done a lot from an internal software standpoint. We've developed to help bridge that gap with them and give them more familiar means of identifying the things that are of interest to them and how they want to see them, how they want those displayed in the app.

(Joel Derfner at 00:17:22) But then on the back end, like you said, we're not dealing with a cat classifier, you know, where I can easily identify 2 million pictures of cats and I've got a really good image classifier. The types of models we're building are highly nuanced in that, you know, obviously language-based. But the language that we're dealing with in legal contracts, a lot of times, is intentionally ambiguous. It was written in such a way to ambiguate meaning, you know, for the purpose of interpretation. And so that interpretive element of labeling data becomes really, really challenging.

(Joel Derfner at 00:18:03) It's not something you can just go outsource very easily to an inexperienced person. And so our ability to take every labeled sample that an SME has provided and use it to its fullest, like you said, in terms of sharing across domains and leveraging similarities that might exist from this domain into this domain, is extremely important because every one of those labels and every one of those pieces of data that we use to drive our model is expensive. You know, there is no cheap data from our perspective. And so a big part of our strategy is how we use that data to its most effect in our platform.

(Joel Beasley at 00:18:52) Yeah. I saw that you guys—I think you called it community-driven AI—that you leveraged that. Is this what you're talking about? You have these subject matter experts that are also users, and they can help influence and train the system?

(Joel Derfner at 00:19:03) Yeah. You bet. So that's definitely a part of it. The other part of it, I would say, is really around our security model of essentially federating the training of those models.

(Joel Derfner at 00:19:17) And so, obviously, you know, dealing with legal contracts, there's a high degree of sensitivity and security around the information itself. You know, it's not this social platform where all the information is available to you, then just use it however you want. The contracts themselves are very critical from a business standpoint in terms of the information that is contained within them and the confidentiality of it. And so our security model in terms of how we leverage customer information to train our models in a global sense allows us to build more generalized models than one company could do on its own. So, you know, if I am ABC Operating Company, I probably have a form for some agreement. And that's the same agreement that I use for this, or at least 90% of it is the same agreement that I use time and time again. But when I go do an acquisition of a company, let's say, I may get a completely different form that I've never seen before. Right? So I've trained this model on this form, and I've really trained to the test.

(Joel Derfner at 00:20:31) And when I go throw something new at it, I can't generalize outside of what I've trained from my silo. You know? And so we are able to, with our approach, kind of break down those silos of forms and sort of formats of different language and deliver a lot more accurate and more generalized solution around these different contract types.

(Joel Beasley at 00:20:57) Yeah. Yeah. 100%. And when you were growing, it looks like you've grown pretty quickly because you said you started in 2017. That means you added, you know, 50-plus people in the past two or three years. What did you learn from that when it comes to recruiting and finding the right people?

(Joel Derfner at 00:21:14) Yeah. It's a good question. I would say patience, maybe first and foremost. You know, I don't even know the statistic, but I would venture to say single-digit percentage in terms of our hire rate of interviews. You know? So we interview a lot of people. We involve a lot of our team in those interview processes as well and try to get a variety of inputs and perspectives on it. But really, I would say patience and really a specific view of what need you're trying to fill. You know, it's hard to go out and hire—let's say I want to just go hire a data scientist. Right? Well, my data scientists come in a million different colors these days. Right? I have people who are experts in BI platforms who call themselves data scientists. I have practitioners of machine learning. I have academic machine learning individuals who are super focused on research and development of model architectures. Right? And we are not developing new model architectures. We're very much operationalizing machine learning and doing it in an applied way. Right? And so you need to really be focused on the right individual who's going to excel in that kind of environment.

(Joel Beasley at 00:22:34) And where are you spending the most time today?

(Joel Derfner at 00:22:38) Yeah. So I would say today is really on our product. You know? And the last six months has been on our product strategy, particularly as we come up to this new release in terms of us really taking what were previously a few disparate products and unifying them under a common document understanding platform going forward. And so I really say the execution of that strategy and that vision has been the dominant portion of my time here over the last, you know, call it six to twelve months.

(Joel Derfner at 00:23:22) I think, you know, if I look forward just a little bit as to where we're going, like I said before, we are now growing into several different markets simultaneously, which kind of creates a new and different challenge for our company. And so I think looking forward, you know, I'll be back on the recruiting train again, I imagine. And really, you know, figuring out organizationally how we can be more efficient in terms of our structure and process, in terms of how we are leveraging our employees, particularly as we are now fully remote and will likely be for the foreseeable future, just to ensure that the culture is maintained, you know, that that sentiment within engineers and that they love what they do when they come to work every day doesn't fade away, and that we maintain some efficiency and effectiveness in what we're doing and what we're trying to accomplish.

(Joel Beasley at 00:24:25) So before the COVID, you guys were all coming into this main office?

(Joel Derfner at 00:24:29) Yeah. So that was an interesting recruiting learning, actually. So we were all coming into a main office. We're based in Texas. And we struggled recruiting, honestly, into Texas. It's not like we're in the Bay Area, and you've got engineers jumping ship left and right, ready to come work at the next new thing. So recruiting was a little bit of a challenge for us, you know, in the Texas area at the time. And even—but up until that point, I would say well into 2018 even, we were completely based in Texas. And then probably, you know, early 2019, maybe late 2018, we started to do more of a hybrid remote culture. And we had, you know, maybe 10 to 20% of our development staff that were based remote and the rest were based in Houston.

(Joel Derfner at 00:25:25) And so we had grown that percentage of remote folks over time. And maybe at pre-COVID, we were maybe at 30%. And so it was something we were growing into when COVID hit. And obviously, when that did happen, we went fully remote. And it's really worked out, I think, extremely well for us. A lot of our employees were able to move back to hometowns and things like that. And I would say really are achieving a better balance of work and life and being able to spend time around their families and things like that and living where they want to be because of that. So it's—I would say it's been a little bit of a blessing in disguise for us from a work-life standpoint. I hope that continues. You know, our HR team and, really, our whole management team is really diligent about doing things to maintain that culture and make sure that that doesn't kind of fade away behind the webcam and the microphone.

(Joel Beasley at 00:26:26) Of course. That's very important. And you said that you're based in Houston?

(Joel Derfner at 00:26:32) Yeah. Yeah. So I'm actually a COVID transplant. I was based in Houston. And then about, I guess, about a month and a half ago, I relocated back to New Orleans. So I'm from there. My wife's from there. And, you know, we lived in Houston for about ten years, and we would spend two weekends a month in New Orleans. So we were back and forth all the time. And so with COVID, you know, it was kind of unforeseen as to when we'd get back into the office full-time. And so we decided, you know what? We're making the leap as a company. We might as well make it too. And we made the move to New Orleans and enjoying the time with the family. And certainly, it's a good thing for my kids and my wife and everything like that.

(Joel Derfner at 00:27:23) And from a work standpoint, I make it a point—I'm in the office this week. We're having a kind of on-site where a lot of management team will get together and just be on-site at least, you know, a couple days out of the month, just to maintain some continuity and face time. And so we're putting together a lot of things like that just to make sure that we're staying aligned under this new norm. But all in all, I think it's been a good thing for a lot of my team and certainly myself too.

(Joel Beasley at 00:27:57) Yeah. It's important to find the positive in anything that's happening. Right? And what you mentioned about family a couple times, I have found at first it was very difficult. Right? But then after a little bit, it just became, like, this is amazing. I'm really getting to know my kids better. I'm really getting to see who they are, and they're right at that age with three and one and a half. So they're small. The girl is three. The boy is one and a half. And so, you know, it was harder on my relationship at first, and it was harder on the family at first. And then something flipped and happened, and then we are closer than we've ever been before. And that was—you know, I've never really talked about or thought a whole lot about a home life in relation to professional development, right, as an executive. But, man, that foundation at home just bleeds through your entire life, and you don't really realize it until you're on an extreme, like an extreme of it working very poorly or an extreme of it working very well. And I was kind of floating around the middle there for a while, just keeping it going.

(Joel Derfner at 00:29:13) I know. And, I mean, you know, I think there are some drawbacks of the work-from-home thing too in that it's harder to turn it off. You're always plugged in. There are some things to definitely be conscious of as we do more and more of this. But for me, you know, to your point, the small things—there are some things I see now my two-year-old do and I just walk out for five minutes to grab a cup of coffee in the kitchen and then I get to just have that five minutes. Those kinds of things with my older seven-year-old—I don't remember ever seeing that side of her, you know? So it's like, I think it's even just those small moments that are helpful to kind of maintain some alignment and connection with the family throughout the day. That's been really cool. And everybody now is so open. You know, my two-year-old will come barge into the office and sit on my lap and just watch a video chat or something like that. And it's completely normal, you know. And so I think that that level of acceptance of that obviously wasn't there pre-COVID. And so it's really nice for that to be more accepted, not just within our company, but across the board.

(Joel Beasley at 00:30:27) And how—now that the environment changed, how are you thinking about burnout or unwinding, you know, hobbies? How do you get the family from going stir-crazy?

(Joel Derfner at 00:30:42) Yeah. Well, me personally, I like to do a lot of stuff outdoors. So that's actually, for me personally, one of the motivating reasons to move back to Louisiana. I love the outdoors, and that's where I grew up, you know, down in South Louisiana. So it's good to just go out to the lake, you know, on Saturday and sit on the levee or take the boat out on Sunday and just spend time with the kids like that. It's a good way to not be at the mall and with a crowded group of people, but still just be kind of clear in your head and spend time with the family. So we try to do quite a bit of that. You know, every evening, I'm out there playing volleyball with my seven-year-old because she loves that. So, you know, those kinds of small things where we're able just to get outside and enjoy nature a little bit is kind of what we thrive on.

(Joel Beasley at 00:31:39) Yeah. We were just in Texas last week in the Dallas-Fort Worth area, exploring that as a potential place to relocate because we're in Florida. And we've both—my wife and I are natives to the town that we currently live in. And so we've been here our whole lives, and it's only been getting hotter and hotter and hotter. I remember when it got cool, you know, and it's just, like, 90. We were looking the other night to go for an evening walk, and the sun's already down and it's 92 degrees. And then, you know, on Christmas last year, it was ridiculously hot. And so we were looking online in different cities and their weather patterns and, you know, looking to go explore. And when we landed in the DFW area last week and we stepped out of the plane and there was half the humidity that we have, and then the temperature was 30 degrees cooler, we were just so excited. And so we're still exploring different areas because Texas is very big. The U.S. is huge. And I was looking up how big Texas is. Texas has a bigger economy than Russia. Did you know that?

(Joel Derfner at 00:32:54) Yeah. I, uh, so I spent four years in Fort Worth, actually. I did my undergrad at TCU. So I'm very familiar. I love Fort Worth, honestly. And I spent two years in Austin in grad school at UT. Yeah. So I've been a lot of places in Texas. It's an awesome state. There's a lot of people moving there from all over the country because it's really cool spot. I heard—I have to check this stat—I heard on the Interstate 10, it is farther from Dallas to El Paso than it is from El Paso to San Diego. It's a huge state. There's a lot to do. Like, West Texas, beautiful terrain and, you know, ranches in the central part, like the hill country through Austin and everything like that. So when we were living in Texas, we would do some camping trips through there and things like that. Dallas and Fort Worth is kind of like old cattle town, ranch-style, you know, homes and things like that. Then Houston's the oil and gas, you know, more than anything within the state. So just a ton of diversity across Texas. It's a cool spot.

(Joel Beasley at 00:34:02) You got to do some work in the oil and gas industry. Right?

(Joel Derfner at 00:34:05) Yeah, that was kind of where I really started from an engineering perspective, doing more mechanically project-oriented things. One of the first things I worked on when I joined the industry, there's this platform called Independence Hub in the deepwater Gulf of Mexico.

(Joel Derfner at 00:34:31) The thing, it's just impressive, the scale of the things that you worked on. Right? This thing was a facility that floated in 9,000 feet of water. It produced, at the time, I think it was 2% of the natural gas consumed by the United States. And if you were to put it on an aerial map on top of Houston, the wells that fed this thing spanned 120 miles.

(Joel Derfner at 00:35:04) Yeah, and it all flowed back underneath the sea—9,000 feet of water in the Gulf of Mexico, rather—all back to the central facility. So just the scale of those projects and, from an engineering perspective, the complexity of things that go into building and installing those systems was just a cool, fun thing to see. And as my career evolved, it got more into the analytical domains of that, understanding productivity and failures of machinery, things like that that might have been more machine learning-oriented.

(Joel Derfner at 00:35:45) And so that's ultimately how I meandered to where I am today. But yeah, that's how I got started.

(Joel Beasley at 00:35:53) In that industry, was that the first time you went from individual contributor to team lead?

(Joel Derfner at 00:35:59) Yeah, that was. At some point—I forget exactly when this was, but a handful of years ago—this was, I would say, the front end of when machine learning was becoming a big deal. And so the oil and gas industry was like, "We need to do this machine learning stuff. Let's do it." And so the company created this innovation initiative and analytics initiative around really leveraging AI and machine learning within the business more fully, all the way from geophysical interpretation of the formations underneath the surface of the earth, through well productivity and performance, even into some aspects of safety and environmental health and things like that. And so I helped lead the engineering team oriented around those initiatives as kind of one of my last leadership roles with them. And it was really fun. There are so many smart people that work in that industry.

(Joel Derfner at 00:37:07) And, obviously, they have been smart at a plethora of things, from more traditional engineering and construction-oriented things, like I was saying with those megaprojects, all the way through well performance and things like that. And now into machine learning as well. There's a lot of really innovative things that have happened in the last five years or so in the industry around the use of machine learning. So it's been a cool thing to see there.

(Joel Beasley at 00:37:39) And at what point did you meet ThoughtTrace? What's the story of you joining ThoughtTrace?

(Joel Derfner at 00:37:46) Yeah. So I think I was boiling crawfish, actually. That's what I was doing. I think my wife and our CEO's wife met each other and had kind of become friends. And Nick, our CEO, came over for a crawfish boil at our house one day. We're eating crawfish, and we just got talking about machine learning and some of the things that he was doing, some of the things I was doing, and really hit it off. And from that point, I think the thing that I really was drawn to about Nick and what he and the team were working on and doing at the time was really bridging the gap between machine learning and AI as an idea versus as a solution. In working in the role that I had in oil and gas previously, we would, every week, probably have three different companies come in and pitch us a machine learning solution that was not a solution at all.

(Joel Derfner at 00:38:54) It was a way to sell services so that they could build us a machine learning model that would do some task. It was really difficult for people to integrate machine learning into software in a way that just worked out of the box on day one. That sounds very obvious and simple, but it's a really hard thing to do, even, I would say, today. Operationalizing machine learning is not an easy thing to do at scale. There's a lot of ways that machine learning is used today to generate a report or generate some insight and then take an action.

(Joel Derfner at 00:39:31) But day to day, every day, how it integrates into a business process is not an easy thing to do. And I think they initially had done that, and I think we have done that as a company really well in our software, in terms of really kind of making machine learning almost fade into the background of the software that you're using. You have type-ahead on a text message—you don't care whether they've got a monkey somewhere telling it what to type or whether it's a machine learning model.

(Joel Derfner at 00:40:05) You just care that it gave you the right suggestion, and you use it. And so really making AI as a background component of business process has been something I think we've done quite well.

(Joel Beasley at 00:40:20) That's a funny visual. That'd be a good—I was almost going to say that's a good startup until I realized PETA would just rip us apart.

(Joel Derfner at 00:40:30) I like that. They got monkeys that can do some pretty advanced things these days.

(Joel Beasley at 00:40:38) Look at us, man. We're crazy. Yeah, we're beaming light around the world right now, talking in real time. That's chaos. It's so cool. I constantly find myself just amazed about how much we've done. I know I say it a lot, but a hundred and ten years ago, we got electricity.

(Joel Derfner at 00:41:00) I know. Things have happened so fast. It really is unbelievable.

(Joel Beasley at 00:41:06) And then I'm just excited. Right now, I'm pretty excited about the advancements with the Teslas. Right? These types of cars. I love running outside a lot. I'm a big outdoors person, and so I'm always excited when Teslas drive by because I don't choke.

(Joel Derfner at 00:41:23) Yeah. Right? Hey, if you move to Dallas–Fort Worth, you may see fewer Teslas, but I'm with you on that. I think it's super cool. And the way they, obviously, consumerized not just the vehicle itself, but the whole buying experience of it and the ownership experience of it—it's really cool. I'm hopeful that that sort of same thing catches on. I think competition will ultimately prevail there. And I think to see more and more folks do what they've done—it's really cool how they kind of reinvented that industry, though.

(Joel Beasley at 00:42:02) Yeah. And then when the COVID thing happened, I had Ken on the show, who's the CIO at Ford. Yeah. And in our prep meeting, I was trying to figure out how to ask him a question about why are the electric vehicles so ugly from every other car company? Why can't they do that? Or why can't they just take the model as it looks? Just take that car, your Mustang or whatever your Ford truck, and just make it electric. It doesn't need to be ugly futuristic things. And as I was trying to figure that out with the prep team, I said, "Wait, let's go Google and make sure that they're not."

(Joel Beasley at 00:42:39) And we found a new Mustang that looks amazing. It looks like a Mustang, and it's electric. And inside, it's got a very Tesla-esque feel. It's very simple. It's got the screen and everything. And then I did more research and found out all the car manufacturers picked up on this, and their models coming out over the next couple years are going to be electric and look normal, and that's the big push now. And so we scratched that question from the show. But yeah, it's exciting for me. Yes.

(Joel Beasley at 00:43:17) I love the Teslas, but it's exciting for me to see everyone else follow suit. And they're just better cars. If you've ever driven in a Tesla, it's just a better car. It's hard to call it a car. It's an entirely different experience. And the words don't mean anything until you actually go drive it around.

(Joel Derfner at 00:43:35) Yeah. And from an engineering perspective, the performance of the vehicle—it just makes sense that an electric motor would have a faster zero to 60 time than a combustion engine. You know what I mean? So you see these tests and videos. Actually, good buddies with one of the guys who has a podcast called Tesla Geeks—huge Tesla fans, super fans of Tesla. And they post a lot about new cars and things like that.

(Joel Derfner at 00:44:09) But yeah, I mean, the performance videos and things like that that those guys have posted, seeing them—it's a really impressive vehicle. And so I think I've been most impressed with them about the range that they've brought to the cars. That was always my biggest concern. Particularly for somebody like me, I'm driving every couple weeks back to New Orleans. It's 400 miles, 350 miles one way. I'm like, I can't stop and charge a car for two hours in the middle of that drive. You know what I mean? But now what they're doing with the range and fast charging and things like that, you know, they really moved on that curve, I would say, just in a matter of a few years, which has been super impressive.

(Joel Beasley at 00:44:58) Yeah. And it was a smart move what they did with the battery patents and then becoming the battery energy company. Right? They're essentially—Tesla's going to become an energy company. Right?

(Joel Derfner at 00:45:08) Yeah. Yeah. I mean, I fully believe it. I love their ambition. And one of the things for us at ThoughtTrace is we've really grown quite a bit over the last couple years into the renewable space. And so that's really, personally, opened my eyes a lot to that industry and the complexity of it and the pace at which it's growing as well. So Tesla's certainly there, but other companies too. And I'm really impressed by what they're doing there and how that's coming along.

(Joel Beasley at 00:45:39) So some of the renewables industry—the people in the renewables industry use ThoughtTrace for their document analytics?

(Joel Derfner at 00:45:46) Yeah. So a renewables asset is just as complex as any other asset. Right? Anywhere where you have asset-intensive industries—in a renewables sector, I've got tons of solar farms, I've got tons of wind farms that I'm managing, distributed over many acres or many states or whatever it may be. Or maybe I've got rooftop solar installations, which I've got multiple times more leases associated with those assets. So I'm leasing every roof that I've put a solar asset on. Every one of them's got a power purchase agreement associated with them or an agreement to interconnect the power into the electric grid. So these situations—that's a renewables example—but in real estate, obviously, you've got buildings and office space and warehouses. Telecom would be things like cell towers or with 5G. You've got a bunch of smaller towers now that will be going up.

(Joel Derfner at 00:46:48) So anywhere where you've got a lot of assets, generally, you've got a contract around each one of those assets. And those are particular industries where ThoughtTrace is really proficient because we're really good at kind of understanding those contracts and bringing information to bear on them and allowing users to manage and take action on them more efficiently. So renewable certainly fits that bill. It's been a really cool area for us to grow into. It's been fun.

(Joel Beasley at 00:47:21) What is the thing you are most excited about at ThoughtTrace?

(Joel Derfner at 00:47:25) The thing I'm most excited about at ThoughtTrace right now? I'll take the geeky answer for you. So in the natural language space, I would say deep learning is a few years behind where computer vision is in terms of advancement. And so deep learning has been a thing within language for a while, obviously. But in the last couple years, there's been a lot of new advancements on that in terms of the use of new model architectures, particularly transformer architectures.

(Joel Derfner at 00:48:02) And, you know, we—like I said, we don't have a big team. I think we're just north of 60 as a company, and only a fraction of that is developers, and only a fraction of that is data science and ML engineers. And so we've really, I would say, taken with a small team of data scientists and operationalized that scale. We've processed, I would say, tens of millions of documents at this point, in the many billions of words, through our system.

(Joel Derfner at 00:48:37) And to be able to operationalize a deep learning system like that is, I would say, nontrivial. You've got the likes of the Googles and Facebooks and Ubers of the world who do it maybe without blinking an eye with teams of thousands of engineers. But I've just been really excited that we've been able to build a system—not just from a data science standpoint, but really, across the board from our platform team to our app teams, our data scientists—that has handled the scale. And certainly, it seems like every year you build something with an expectation of this scale, and then a year from now, you're 10x that, and all of a sudden you have to scrap everything you did and rebuild it from scratch.

(Joel Derfner at 00:49:20) So, you know, the problems get harder for sure, but it's been fun and really exciting for me to continue to kind of reach and achieve new barriers of scale on our platform. So that, and in particular the context of deep learning there, I'm really excited about where that's going from a natural language standpoint.

(Joel Beasley at 00:49:46) That's exciting. Right? Shipping product, operationalizing it, overcoming these huge obstacles in order to do so with a small, lean team. That seems like a trend.

(Joel Derfner at 00:49:59) And, I mean, you go look on a Medium article, somebody will post, "Hey, this is how you do natural language generation with BERT." And it's, you know, they've got 17 lines of some PyTorch code to show you how to do that. But it's like, you know, that only scratches the surface of really what is required to put that thing to life. Right? And there are so many specifics around how that implementation needs to function for your organization and for your product. And so, you know, the line that says "dot fit" is really, honestly, the simplest line. It's all the stuff that comes before and after that that makes it really hard.

(Joel Beasley at 00:50:40) Yeah. Including the sales side of things. Right? So not only do you have to build it, it has to work, and you have to document it and support it and then sell it, and you have to do all of these things. It's like a symphony playing together. Right?

(Joel Derfner at 00:50:53) Yeah. Yeah. Absolutely. And making sure that we're building the right products and things like that. You know?

(Joel Derfner at 00:51:00) I can tell you, we I think have had a strategy, and we've really been executing on that strategy now pretty consistently for the last eighteen months. But early on in particular, there was a lot of different directions we could have gone with the business, right, and with the business model. And the idea is only one component of that. The business model and how you execute on that idea can make or break that idea.

(Joel Derfner at 00:51:28) And so really learning that business model and how we can be most successful at implementing that and being successful in growing the company with it has been a learning experience and still is, honestly. We try new things and learn new things every day on that front still.

(Joel Beasley at 00:51:48) Yep. I'm always fascinated with communication. It seems to be one of the things that affects every area of your life.

(Joel Derfner at 00:51:56) Yeah. Yeah. Absolutely. And the different styles of communication that you need in the different aspects of your life. You know what I mean?

(Joel Derfner at 00:52:06) That's the hardest part. You need to be a chameleon a little bit. You can't just bring one phase to every situation. You need to kind of change tack depending on who you're talking to or what the scenario or situation is. So that's what makes it really hard.

(Joel Beasley at 00:52:23) Oh, yeah. And then as you start talking to a lot of people, whether you have a company and you have a lot of people or, like I do, a lot of interviews. When you're around a bunch of bright people, right, which you are at ThoughtTrace, when you're around a bunch of bright people, you learn there's so much to learn that, for me, it's become interesting to watch how I forget.

(Joel Beasley at 00:52:46) Because I get all this great advice from all of these different people on all of these topics, and it's amazing how my mind will just delete certain things but keep other things. It's been fascinating.

(Joel Derfner at 00:52:58) It must be incredibly challenging to show prep. You know? The mental gymnastics that you go through every week to talk to somebody about cybersecurity all the way to structuring teams. Those are big deviations in terms of what your mind's doing.

(Joel Derfner at 00:53:17) So how do you prep?

(Joel Beasley at 00:53:18) Yeah. So prep is pretty fun. We've gotten better at it over time, but we have a producer and associate producer. We have prep meetings. We get ideas and topics of what people want to talk about.

(Joel Beasley at 00:53:32) And then we sit down and we research the person. So we've got, like, in front of me right now, I have this document that has your face, all the links to all your social profiles, summaries from our producers, and bios on your company. Yeah. It says you like fly fishing and have two daughters. Like, we research the heck out of you.

(Joel Beasley at 00:53:52) Right? And then we sit down in these meetings and it takes a couple hours to prep every episode, but we ask ourselves, what is the area that's going to make this person shine? What type of information do they have? How many people are at their company? What are they experiencing right now?

(Joel Beasley at 00:54:08) Where can we pull useful knowledge from them so that the audience will enjoy the podcast?

(Joel Derfner at 00:54:16) That's awesome. That's really hard to do because, obviously, everybody brings a unique set of experiences and perspective to every conversation. So I don't envy you. That's a tough task for sure. I don't think I have a machine learning system to do that.

(Joel Derfner at 00:54:32) I'll say that your job is protected.

(Joel Beasley at 00:54:38) I'm just such a geek though because, I mean, I did software development. I wrote code for like seventeen years and built engineering teams, and so I'm genuinely interested. And it's like my job is to get to understand these technologies and explore them and do deep dives. I did a big deep dive the past couple months ago into quantum computing. Right?

(Joel Derfner at 00:54:59) I was in an episode or two where you're talking about quantum. Yeah.

(Joel Beasley at 00:55:04) Yeah. Did you get the Honeywell episode?

(Joel Derfner at 00:55:07) No. I did not. You went deep quantum on Honeywell?

(Joel Beasley at 00:55:10) Yeah. So I was trying to figure out if it's business ready. Like, what's the state of quantum computing? Because like you said earlier with the Medium articles, you can find an article on here's how you program a quantum formula and an API. And it's just this cookie cutter example.

(Joel Beasley at 00:55:25) And I wasn't really sure what the purpose of it was because you could do that with traditional—

(Joel Derfner at 00:55:31) Depending on what your search history is, you can find a Google search to reaffirm any bias you have.

(Joel Beasley at 00:55:37) I know. That's the rule, isn't it? And so I couldn't wrap my mind around something. And that is a huge driver for me because I'm a very curious person. So because I couldn't wrap my mind around it, I was like, alright.

(Joel Beasley at 00:55:48) I'm going to go a mile deep into this. I started taking the courses on algebra and, apparently, linear algebra happens to be a good starting point for quantum computing. And then I started inviting all the guests on. So at the time that week I was doing that, I saw a press release for Honeywell releasing the world's fastest quantum computer, and it reminded me of the Intel races where every three months there's a new company. They're all going after each other.

(Joel Beasley at 00:56:18) And what my takeaway from the entire multi-week experience and really drilling down and asking tough questions to all of these people involved with quantum computing is that it's currently like—if you could imagine the analogy—it would be like classical computing when the computers were very, very, very large, like the size of a room. Now I'm not saying that quantum computers are very large, although some of them are. But what I'm saying is, like, in that sense, the big computer room, it was mostly mathematicians and nerds doing nerd things that hadn't gotten to the point where there were developers building for business logic type deal. So my—and, you know, correct me if I'm wrong—but my takeaway of the current state today is that it's useful for people running quantum algorithms and doing quantum things which are usually like quantum physicists. Right?

(Joel Beasley at 00:57:12) Because they need to process quantum computations, and they were doing it, you know, maybe by hand or something, or maybe weren't able to do it. I don't know. I'm not a physicist, but it's useful for them right now. So there's APIs and there's a business model behind it. And in the future, it might be useful for some other things.

(Joel Beasley at 00:57:28) And then the last takeaway was even people like—I talked to the creator of Ripple, you know, the cryptocurrency, he's not even worried about it for the next ten years because that's a big hot headline if you Google, oh, if you can go build a quantum computer, you can crack and have all the Bitcoin in the world, and there's going to be all these problems and everything. Well, no. There's actually quantum safe algorithms for hashing. They're just a little bit slower today. There's a ten year—he's like, I worry about things that are less than ten years out.

(Joel Beasley at 00:57:57) So they're not worried about it in the industry right now. And I talked to multiple people. No one's freaking out about this.

(Joel Derfner at 00:58:02) Yeah. Yeah. Yeah. It's certainly not something—not that it would be rude. Probably just leverage it if it was available and useful, but, you know, it's not something that strikes me as something that will affect our business in any way in, like you said, the next five to ten years. There's a huge kind of—it's not just the technology itself, but almost the bigger leap is the democratization of that technology. Right? That's often even a harder thing to accomplish. And we're still not there with machine learning today.

(Joel Derfner at 00:58:37) Right? The platforms that are being developed by a lot of companies are trying to democratize access to machine learning more so to kind of non-development-centric people. But even that, it's not really fully democratized just yet.

(Joel Beasley at 00:58:55) I have faith, though, that it'll happen as it needs to happen. Right? We're moving so fast, and I love the fact that, you know, back in the eighties, some new technology came out. The whole world would know about it because there was one source of—you know, handful of sources of information that would stream out. But now there are so many advancements happening in so many different, you know, areas, like nooks and crannies of the world and so many industries that you literally cannot catch up.

(Joel Beasley at 00:59:28) You cannot hold it all in your mind. It's too much information. Right. So for me—yeah. Go ahead.

(Joel Derfner at 00:59:35) No. I was going to say, I think what you said there around it'll happen when it needs to is spot on. It's like the need creates the solution at the end of the day. And if there's a business need or social need or something like that, we'll figure out a way to make it work. You know?

(Joel Beasley at 00:59:51) Yeah. That the—what? The necessity, mother of invention phrase? I love cliches because cliches are things that are so true they make the generation mad because their parents said it.

(Joel Derfner at 01:00:02) Yeah. They're true because they just work. You know? They make sense.

(Joel Beasley at 01:00:07) Yeah. I've always been fascinated with the humans, as if I'm not one. Right? I've always been fascinated with humans because we find these truths and then we kind of hate them. And it's like, that's kind of strange.

(Joel Beasley at 01:00:23) Right? So we found things that we know are universal truths and our natural reaction is to sort of hate them.

(Joel Derfner at 01:00:29) Because they put us in a box. Right? It's like, well, this truth can't be true about me. I'm unique. I don't fit that mold. I can't.

(Joel Beasley at 01:00:38) Joel, you're a snowflake.

(Joel Derfner at 01:00:42) I'm just—yeah. I'll leave that be.

(Joel Beasley at 01:00:46) I like humanity because we do believe that we're special and we're individuals, but we also do things as collectives. And I'll tell you what, if you kind of take a step back and get weird for a second, and we'll wrap. Do you have another five minutes?

(Joel Derfner at 01:01:03) Yeah. I'm good.

(Joel Beasley at 01:01:04) Okay. So if you take a step back and get weird for a minute and you start—if you imagine the universe that we are in almost being like a container, right, and then it's going to exist for a very long time, like hundreds of millions of billions of years, wouldn't this kind of be a fun thing to do? You know, like, go live a life on Earth? Like, there's a hundred years.

(Joel Beasley at 01:01:27) Right? And then it would be kind of interesting because if we were a civilization that existed outside that could exist forever. Right? That could exist for, let's say, a billion years, not even the whole universe. If we could exist for a billion years, we would get pretty bored. Right? Once you can zip around and go everywhere, it's like, okay. And then what?

(Joel Beasley at 01:01:56) And so maybe going through, you know, booting up a planet and going through life would be an interesting activity you could do.

(Joel Derfner at 01:02:05) That's right. That's the next frontier.

(Joel Beasley at 01:02:07) Right?

(Joel Derfner at 01:02:08) I think so. One of my—I told—I mentioned those books I go back to regularly and read. My favorite book of all time is Thinking Fast and Slow by Daniel Kahneman. In the academic world, you're familiar with academic references. And you write a paper, and your goal in writing that paper is to be referenced by other people. Right?

(Joel Derfner at 01:02:35) And if you are referenced by many other academics, that means you've contributed to the academic knowledge in a big way. Other people took your work and built off of it in a substantial way. So you were an important building block to knowledge. That's, I don't know, in a nutshell, how I think about academic references. Well, this book, you could listen to almost any business book, leadership book, you know, self-help book, you name it.

(Joel Derfner at 01:03:06) This book is referenced. You know? So in the realm of academics, I feel like it's a really important book for a lot of people to read because it's so foundational to so many things that we do. But it's a behavioral economics, cognitive science book. And the premise of that book is even though everybody thinks they're so different and unique, they're really not.

(Joel Derfner at 01:03:29) You know? And he gave a really good example of—he and a team of—he's from Israel. Him and a team of other scientists were writing some textbooks as part of, I think, the armed forces over there. And they all sat around and they said, how long do you think it's going to take us to write this textbook? And they were all like, oh, we're really smart people.

(Joel Derfner at 01:03:53) We're cutting edge of our fields. We'll do it in six months. We'll do it in nine months, maybe a year at the most. And then they went and looked at the data, and they said, how long does it generally take teams to write textbooks? And they're like, you know, seven years, no one does it less than five.

(Joel Derfner at 01:04:09) And they were still all like, no. We're going to do it in a year. And seven years later, they finished the textbook. You know? So it's like, we get this human bias that says, you know, no.

(Joel Derfner at 01:04:19) I'm going to solve that problem because I know I'm smart, and I'm going to work my tail off. But, really, you can't separate yourself from being a statistic sometimes. It's kind of the lesson of that book as grim as that may sound.

(Joel Beasley at 01:04:34) Yeah. But it's also—it's almost like that's required. That bright eyed, bushy tailed optimism, lack of experience because, you know, that's most entrepreneurs. Right? You have to have that, like, oh, I'm going to go out there and crush it mentality to get yourself into gear. And then what tends to separate, you know, the winners from the losers is persistence.

(Joel Beasley at 01:05:04) Right? When you have that moment where you realize I was wrong, like, when they got to six months, right, they could have stopped. Right? They could have been like, nope. We didn't do it or we're not there, but they didn't.

(Joel Beasley at 01:05:16) It's like that initial bright eyed and bushy tailed mentality got them started, and then they continued, and then they finished. And the next time they go through, they have a realistic expectation of it and understand the process. And then it's like, I've already done this once before. We can do it again. Let's just do it again.

(Joel Derfner at 01:05:34) Yeah. It's a phenomenal characteristic of people. You know? But you're absolutely right. That is what causes us to be so innovative as a race and as a society.

(Joel Derfner at 01:05:47) You know? It's really impressive.

(Joel Beasley at 01:05:50) That's why I tend to be optimistic with everything from social dilemma algorithms to, you know, election stuff. All the parts of life, right, where fear will set in pretty quickly, I tend to be optimistic because I just run into a stream of amazing people constantly that their heart's in the right place, and we're all just trying to improve and grow. And almost never do I run into the evil villain.

(Joel Derfner at 01:06:18) I bet. Yeah. What'd you think of that social dilemma thing? You mentioned that. I'm assuming you watched that.

(Joel Beasley at 01:06:26) Yeah. I thought it was super interesting. What did you think about it?

(Joel Derfner at 01:06:30) Yeah, I thought it was super well done. You know, I think everybody hears about the robots are coming around AI. I thought it was just a well-done documentary because I think it brought the threat of AI in a different light, right? It's not this robot with a laser that's gonna come kill us all, you know, but it's humans creating the incorrect incentives around the use of AI. It's not like the machine learning models are really good at what they do. And honestly, so much has been created by the technology that those companies developed.

(Joel Derfner at 01:07:15) BERT is a model that is available to everybody as a starting point because Google has done so much work around language and understands that data so well and has been able to leverage it in such a way to push forward the rest of the community and the rest of society on that front. So there's so much good that comes from it. It's just that one situation where it seems misaligned, the incentives of how it gets used that has these ripple effects that we obviously couldn't have foreseen and didn't foresee. You know, so it was really enlightening to me in terms of seeing how easily that can happen in terms of taking something that is, I think, on its own, in isolation, a good thing and then putting it into the wrong situation and how that can really snowball very quickly on you.

(Joel Beasley at 01:08:20) Yeah. I also have faith in people. So I think one of the things that works against us right now, which is why I really liked The Social Dilemma, and back to it'll happen when it needs to happen, it's becoming so important right now. It's really impacting our lives, and so people are talking about it. They're making movies about it. They're writing books about it. Discourse, conversations going to happen. Thought leaders will emerge from that. Standards and common ways of thinking about this stuff will emerge from that, and then moral compass will be able to connect with that, right?

(Joel Beasley at 01:08:56) You will be able to calibrate against it. In that way, we're decentralized as humans, right? We're all individuals, but we're also a group. It's like we can play both. And so what will happen is these engineers, once it's ironed out, once it's talked about enough, and once there's these thought leaders and there's these common ways of thinking about it and kind of these things we all agree on as humans, at least for right now, then people will start raising their hand or speaking out if those things are being violated. If the no-no's are happening, you get enough, you get a hundred people together, some of them are going to say this is against what we should be doing. This is not the thought in the industry, and they will tell on the bigger organizations, right?

(Joel Derfner at 01:09:43) Yeah, yeah, yeah. I agree. And I think from a society standpoint, I bet if you went around and took a survey about any of these platforms, like, is this good for your life? I don't know that the majority of people would say, yes, this is good for my life. I think everybody sees some negativity associated with it, whether it's addiction and time that they spend using it and things like that. And so I'm with you. I think the more that conversation just happens amongst individuals and amongst society, I think the more action you'll see being taken around some of those things. And I think that documentary just, in a very objective, kind of nonpartisan way, did a great job of just framing the issue in my mind.

(Joel Beasley at 01:10:36) Yeah. And also, we're in America, so personal responsibility is the foundation of Americans, right? And that's your phone, right? There it is. You know, you can put it down. You can delete the apps. You can, for me, eight, nine months ago, I turned off all notifications. I don't have notifications anymore. I noticed myself picking up my phone a hundred, two hundred times a day reflexively. It just bothered me so much. I just said, I remember a time when I was going through high school, we didn't have cell phones. And for the first, you know, eighteen years of my life, I wasn't picking up a cell phone, right? And then it's this thing I do, and you get the screen time reports. And so to take control, I turned off all notifications. And when the new iOS came out, I removed all the screens and the icons. So it's basically just a picture of my kid when I open it up, and then I have to go search for any specific app I want to use. I'm moving them all into the app library. So, you know, I think it's really important to, when it's this easy to have junk food, right, you have to come up with ways to get the junk food out of your house.

(Joel Derfner at 01:11:47) Yeah. And that's, I think, where we reach the end of machine learning's benefit to it. It is too good. I've never been an Apple person. I think Apple does a great job with security. I've been on Android from day one, Google, everything I do. But I use DuckDuckGo now for a while. And I'll be the first to tell you, it is not as good. It is not as good. I don't find what I am trying to find as easily. But that is a cost that I think is worth it. You know what I mean? Because in other situations, I don't want to find what is, quote unquote, the right thing for me to be looking for. It should take some effort in some situations. I had to go write a book report in high school. I was going to the library, and I had to sit down. I had to read the three books on the shelf that were about that to understand all the perspectives and write an opinion. It should be, I think, a little bit more of a lift to find the information you need to formulate your own opinion. So I think it'll take some getting used to as well from people that maybe this system or this service or this solution does not need to be so good. Maybe it can be a little worse, and that's okay.

(Joel Beasley at 01:13:13) Yeah. DuckDuckGo, right when I started using it, I noticed that exact same thing. But I like the fact that you could research things and not it not be a part of your profile and you'd not be put into that box.

(Joel Derfner at 01:13:28) Right. That you have to go, you're starting from scratch every time you go in as a person, which means you get all the information, the things you agree with, things you don't agree with, and you have to make heads and tails of it. And I think that mental, those mental gymnastics are actually good.

(Joel Beasley at 01:13:46) And I fully agree. That's why, as parents, I think to kind of wrap this whole privacy conversation up and the future and the fear and the happiness and all of the good things that will come with it and difficult conversations and all of that, one thing I found out is that as parents raising the next generations, we can do a better job raising them to think and how to think and how to deal with these things. Because yes, it's a problem, and we both saw on the documentary that the teenage girl suicides went up. But where was the parenting in that? You know, it's not all on the algorithms. Yes, they're addictive. Yes, they have scientists doing that. But where was the parenting on that?

(Joel Derfner at 01:14:36) Yeah, yeah. I mean, it highlights the need to be engaged in that stuff. You know, I mean, just like, it's no different than, like, active bullying at school. You know, it's just now obviously a digital one, and it's harder to see that those things are happening, right? So it just takes an added level of vigilance from a parent's perspective to stay on top of those things. It's a harder job in a lot of ways because of that.

(Joel Beasley at 01:15:05) We just connected all the humans on the planet. It's like we're going through puberty right now. We're trying to figure out how to deal with this, right? We've got the hormones going. It's crazy. We're all immature. Where's dad, type deal, right? There's no central agency to control this stuff. This is not necessarily a government. There's nothing you can point at. And so we're all just kind of these angsty teenagers maturing. And over the next decade or two, things will figure it out. We'll calm down a little bit. And then in the next hundred years, we'll be wise thirty-year-olds. And then in two hundred years, we'll be forty, and three hundred.

(Joel Derfner at 01:15:44) I like your optimism. I'm looking forward to those days.

(Joel Beasley at 01:15:47) We'll get there because that's the beautiful thing. As difficult as all these problems are, historically, we solve these problems and move forward. So the storm will be here, but I'll tell you what, Joel, learning how to stay centered in a storm is similar to communication. It feels like it's something you're always working on.

(Joel Derfner at 01:16:08) Yep. Yep. Gotta constantly keep yourself in check and kind of go back to your priorities as a person, you know, and make sure you stay standing on those.

(Joel Beasley at 01:16:19) Yeah. And surround yourself with great people. You know, Joel, this is awesome. Dude, we made a podcast.

(Joel Derfner at 01:16:26) Yeah. Joel, it was really good getting to know you a little bit. That's fun.

(Joel Beasley at 01:16:30) Is there anything that we didn't get out that we want to get out here at the end?

(Joel Derfner at 01:16:34) I don't know. I think we touched on everything, honestly.

(Joel Beasley at 01:16:38) Do you have a demo on your website, ThoughtTrace? Do you have a demo, or do they contact you? How does that work if somebody wants to see it in action?

(Joel Derfner at 01:16:45) Yeah, awesome. My marketing director, Britney, who's phenomenal, would have beat me up if I didn't say something like that. So, you know, yeah, we've got demos on our website. We're actually going to market, like I mentioned, with our new release on the twenty-seventh. So a ton of new content hitting the website on the twenty-seventh around that. Done some really cool things with really specific experience videos that are really targeted to the industries and the users and the types of problems that they're solving with our new product too. So all that will be on the twenty-seventh, and really looking forward to going live there and continuing to improve the product over the course of the next handful of years. It's been a fun ride, and it'll be, I think, an even more fun ride once we make this new launch.

(Joel Beasley at 01:17:40) Oh, that's exciting. And we'll put links in the show notes and everything so people can access that really easily.

(Joel Derfner at 01:17:46) Cool. Cool. Thank you for that.

(Joel Beasley at 01:17:48) Dude, this is great. Joel, do you ever have questions or you hear someone, you're listening to the podcast and you hear somebody talking about something and you want to talk to them, you just reach out to me and just say, hey, can you connect me with this person? I want to talk about this topic. And I'll introduce you. And everyone's, I get letters from people all the time that this is something that happens. So I figured I would help facilitate it as well. So if you ever need anything like that, you just reach out. You know, Jake, our producer, can help connect you. I can help connect you. We'll make it happen. But whatever you need, because we want ThoughtTrace to be successful, I want you to be successful. So however we can help, just let us know.

(Joel Derfner at 01:18:25) That's awesome, man. I really appreciate that. Thanks, Joel. It's fun talking to you.

(Joel Beasley at 01:18:30) Thanks. Have a great day. Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.