091: ABC’s with CTO of Cast AI – Leon Kuperman | Artificial Intelligence | Brazilian Jiu-Jitsu | Cloud Computing

[00:00:13] Zach White: So we had a really fun little chat before I hit record today and I asked about something on your LinkedIn profile that I think is super interesting.

[00:00:22] I wanna explore it again. Brazilian Jiu-Jitsu. So here’s a, a CTO doing some incredible technology, and we’re gonna talk about AI and about careers and digging into some really important and time sensitive topics today. But before we get there, I’ve gotta get into like, when did you start Brazilian Jiu-Jitsu?

[00:00:40] Where does that fit into your life? . Yeah, Zach, so I 

[00:00:43] Leon Kuperman: have to take you back all the way to like junior high and high school. Okay. No, no surprise. I was a scrawny, nerdy kid. Like I started coding when I was like nine or 10 years old. Apple 

[00:00:56] Zach White: two oh . 

[00:00:56] Leon Kuperman: So I was picked on all the time. there’s this classic story that my friends still remind me of. This one bully kept me in a headlock for the whole lunch period, 1, 1 1 day, right? And it was like a challenge for him and his buddies to see if he could hold me in a headlock for the whole lunch period.

[00:01:11] I was picked on constantly. Anyway, so if you fast forward, oh man, in grade nine or 10, I joined the wrestling team and I A I I didn’t want to join the wrestling team. I asked the coach if he would help. Teach me how to defend myself. Okay? So I could stand up for myself and he said, you know what, Leon? I will show you, but you have to join the team and stick it out till the end of the year.

[00:01:34] So I did, and I wasn’t great at it at the beginning, but I got better and better over time and ended up, competing more seriously. And I really fell in love with it. But, wrestling is kind of a thing that ends in college and high school, right? Like not a lot of professional capability afterwards.

[00:01:50] So, . When I was living in Los Angeles, I discovered this grappling art right around the time when U F C came out called Brazilian Ji Jisu. Yes. And honestly, by then I was in my early thirties. I was kind of outta shape, didn’t really, I hated running, didn’t really have any other exercise that I wanted to do.

[00:02:07] And it saved from a physical perspective, it saved my life. I really benefit so much from being able to train and I’m almost 50. and then I still train. The kids are still much younger, but now, the, from a skills perspective, the playing field’s a bit more 

[00:02:22] Zach White: level.

[00:02:22] Wow. So, I mean, we could talk about jujitsu all day. There’s so many questions I have, but I’m actually curious, having gone into that training and the comment you just made, you know, it saved, it saved my life. how does jujitsu and engineer. relate for you. Where have you found overlap or how does it inform the way you approach your career?

[00:02:46] I mean, obviously extremely successful in your career. We’ll talk more about that, but where’s the parallels for 

[00:02:51] Leon Kuperman: you? Jiu-Jitsu teaches you it’s very much like a chess game or I would say it’s more complex than chess. It’s more like go, because the in the patterns are much more infinite that you can kind of follow and you have physical attributes that you have to kind of, uh, understand about your opponent and understand about yourself.

[00:03:08] Are you flexible? Are you strong? What skillset do you have? Okay, so I actually wrote a really interesting article about this years ago about like what cybersecurity and Brazilian jujitsu have in common. Like when you are facing a cyber, adversary, , how do you think of that in terms of a physical altercation?

[00:03:25] But anyway, so the sport has taught me, to be patient and to be analytical far more than one or two moves in advance. Yeah. To really understand surrounding, and patience is such a big part of it. Zach, you know, because often when you try to rush something, it doesn’t work out because the circumstances around you, aren.

[00:03:47] But if you learn to time and anticipate the reactions of the environment and of your opponent in this case, you really get a much better result. So it’s taught me to think critically in a lot of aspects that I now use every single day. Yes, 

[00:04:01] Zach White: yes. So, Is it for you something that that connection happens naturally.

[00:04:08] So you’re, you’re in a board meeting or you’re working on a piece of technology and you just subconsciously find yourself applying a principle like timing and patience or looking at my adversary in a certain way. It just happens or it. Okay. Yeah, see, I think that’s, I’ve had the same experience with things that I’ve done.

[00:04:29] You know, for me, people who listen to the show know I, I have a ballroom dancing background and all of the things that I’ve done in dance create these subconscious experiences that show up in eras. Outside of dance, including business and engineering and other things. And that’s what I love about this kind of multidisciplinary mental model way of living.

[00:04:48] And just encourage everybody, get, get, get a hobby and go dig in. Especially we use our minds for a living. When you do something with your body for recreation, I think that’s a really powerful way to balance your life. Yeah. But. 

[00:05:03] Leon Kuperman: You know. Exactly. You know, Zach, it’s really interesting you say that cuz like usually you would think of an engineering team to be more sedate and more kind of focused on you know, the book smart, so to speak.

[00:05:12] In the current team that I have on the field, I’m really amazed by them. Almost every single one of them is an athlete. yeah, triathlon. One of them was talking to me about a biking tour they did, and Spain this, this week, I love the fact that the, the engineering discipline is moving towards also having a physical counterpart outside of work.

[00:05:32] Because you’re gonna have a bad back, you’re gonna have bad posture. All of these things are gonna affect you. Yeah. And if you want a long, and thriving career, you need to take care of your body. I agree. 

[00:05:43] Zach White: I agree. So Leon, I want to unpack. Something that’s pressing in this time of history with you at the point we’re recording this, it’s early 2023.

[00:05:54] Most of the, at least engineering world is spending their nights and weekends playing on chat G P T right now. We just chatted about that before the call and everybody’s asking new questions about what is AI going to do in our industries for our jobs, and you have a really powerful perspective and expertise in this and we want to get there.

[00:06:15] but rather than start at the end, I wanna begin with where did you become that expert and, and how you now are in the space that you’re in. So, back us up maybe all the way, if you’re willing. And tell us about from childhood to getting into this world, what are those key milestones and moments that pulled you into technology and engineering and how we got to today?

[00:06:40] So you can go back as far as you want. . 

[00:06:43] So I come from a family of engineers, so it’s kind of unfair Eastern European background. So I was born in a city called Odessa, and my mom was an electrical engineer. My dad was hands-on, I wouldn’t call it a mechanical engineer, although he, he was basically the guy in the shop putting the machines together from all the drawings.

[00:07:02] and then my brother became a mechanical engineer and they have a very successful family, like robotics, business in, food processing. And my mom, , so I was doing sales for them in college and, I’ll just kind of skip forward a little bit. I had this offer to go and work, in an internship at IBM here in Toronto.

[00:07:20] Leon Kuperman: And my mom said, you know what? This family business is super hard on work-life balance. Like, you know, basically family has no life outside of work. And she wanted more for, for me and for my family. And she said, go do the IBM. because you’re gonna at least have a chance to discover what that work-life balance.

[00:07:37] I don’t want you to be an entrepreneur. It was essentially interesting because it’s too hard, it’s called the immigrants dilemma. I don’t know if you’ve ever heard the term before, but like immigrants come to the country, they want to do well so that their kids can become doctors and lawyers and engineers so that they don’t have.

[00:07:54] As hard as, you know, previous generations. Anyway, so I went to go work for my b m and I got very fortunate that I didn’t get stuck in a boring project. I, I started working on this e-commerce thing back in 1996, and it was just a rag tab team of like four or five devs. that put together a product that wasn’t even a product really, that became the launchpad for IBM’s e-commerce platform.

[00:08:20] Essentially. It’s called, uh, WebSphere Commerce Suite. Okay. And it launched like LL Bean was our first customer and then Sony style and a whole bunch of others, right? So all of a sudden this little team of six got all this budget and notoriety and We kind of quickly outgrew our founding team and it became a team of hundreds and hundreds.

[00:08:40] I had this interesting inspiration and I’m like, we shipped this product and it takes us a year to ship it and shrink wrap. Right. It was still actually shipped on CDs back then. Oh. Oh, wow. This is just the wrong way to deliver software. Why doesn’t someone come to an online experience, sign up, get provisioned?

[00:08:58] Basically, it was, software as a. before Salesforce existed, before all these things existed. Wow. Yes. And that was kind of one of my foray into entrepreneurs. So that was my first and almost last big company job. And then I’ve been an entrepreneur ever since. Right. have had great successes and failures and spectacular failures in starting companies.

[00:09:18] And you need to have both in order to be ultimate. You know, it’s like, uh, Elon Musk always says it’s the rate of failure, right? It’s not about optimizing for individual successes, but how quickly can you get to failure on mm-hmm. , those attempts that don’t work out. And, um, that’s what kind of brought me all the way through startup.

[00:09:37] and I would say, 

[00:09:39] Zach White: so real quick, let me back you up because mom, mom gave you this fair warning. Don’t go down the path of, burnout and entrepreneurial life and the things that we’ve experienced. You know, go experience some balance. Be a W two employee. You did that, had great success there. The team scaled, and then you leaped out into entrepreneurship.

[00:09:58] Anyway, I’m curiously on if you were gonna go. and give yourself advice the way Mom did, would you give the same advice or how do you feel about that decision now? 

[00:10:11] Leon Kuperman: Yeah, she wasn’t right, like she wanted to protect me, but I wouldn’t, I can’t imagine and you know, this is not for everyone.

[00:10:18] I’ll tell you, entrepreneurship’s super hard. Startup life is super hard. Most people fail. but I can’t imagine a life where I am a cog in the. , of a much bigger organization. It just doesn’t fit my personality that well. Yeah. but other folks are perfectly content in that world and they find other things other than entrepreneurship for that feeling of success, and that’s cool.

[00:10:40] But yeah, my mom wasn’t right about that one and I Interesting. I think she understands that now that it wouldn’t have fit my personality type. 

[00:10:47] For the people who it’s. Man, it’s great and it is so rewarding and for the people who it’s not for being an entrepreneur will be the worst experience of your life.

[00:10:57] Zach White: And it really is important to answer that question clearly for yourself and at least Leon, my perspective in the last five years, 10 years, it seems like it’s becoming really fashionable to want to start your own business for a lot of good reasons, but a lot of people I see going doing. Are not actually asking the question, am I an entrepreneur?

[00:11:19] They just want the money or the notoriety or the freedom and this idea that I can work on whatever I want, which sounds great on paper, but it’s really hard. would you share your story about. Some of the challenges you faced with regard to balance and energy in your life for anybody who might be asking like, oh, was it all, paychecks and roses for Leon after he left IBM

[00:11:41] Yeah, 

[00:11:41] Leon Kuperman: it wasn’t, certainly not. there are so many, adversarial stories. It, it’s probably easier to tell the happy ones, but let me, let me try to kind of give you some color. . Firstly, for folks who think that they can do whatever they want. Once you raise a dollar of money from the outside world, you are no longer in control of your destiny, especially as you start going from seed round to series A to series B, your control of the company diminishes significantly, right?

[00:12:08] And I kind of equate it to being on a treadmill in some ways. Once you start that growth path, you’re on that treadmill and you’re. and sometimes it, it feels like it’s very difficult to get off, like if ever, right. Like you’re gonna basically either run to the end or you’re gonna fly off in a dismal failure.

[00:12:25] Mm-hmm. . So I would say there’s an illusion of control that doesn’t exist. Right. Yeah. And I will tell you, there have been many times where we have been on the end of the rope and hey, we’re two weeks away from not making payroll. not recently, thank goodness. But you have a hundred people on the hook that, you know, their livelihoods depend on you, and you’re two weeks away from not being able to pay those people, nevermind not taking a salary for a couple months, because that’s just what you have to do. Yeah. Yeah. So don’t do it for the money, that’s for sure.

[00:12:55] Because, there’s a nine and 10 chance that, it’s not gonna work out financially, right? Mm-hmm. . even now, like even starting this latest business cast, it’s debatable, even if this is a, a, a great success and I maybe see that financial success in a few years, it’s debatable whether this would be more financially lucrative than, staying at a large company like a Google or Right, right.

[00:13:17] Or not 

[00:13:18] Zach White: Microsoft. Right. Yeah. I appreciate your candor on that and personally being on that treadmill. What are. , the risks that you’ve experienced, the consequences of, if you’re not intentional on that treadmill, if you get stuck, what happens? 

[00:13:34] I would say, and then probably this happens to a lot of folks, what I’ll call in that kind of invincible quadrant, right?

[00:13:41] Leon Kuperman: Where you, I maybe out of your late twenties, you have enough experience to really kind of feel like you can make an impact on the industry. You’re maybe in your early thirties. , you feel like you still have the stamina of a, of a team. So what do you do? You burn the candle on both sides. my only restrictive factor in my mind back then was time.

[00:14:00] Like I just need more time to do things. Mm-hmm. and I will ultimately be able to outpace others around me. So you start sleeping less. you know, when you’re sleeping less, you know, you are not paying as much attention to your exercise and your diet regime and your body starts to kind of break down.

[00:14:17] This is why I was saying exercise is so important to keep your body in shape. Yeah. Yeah. And like I think people break in one of two ways, Zach, so people either break emotionally and that that’s what you traditionally see as a. Somebody who’s just been disengaged or just, you know, takes a leave of absence because they, they, they can’t emotionally get back into their work.

[00:14:39] Some people break physically. So that’s me. I, I’m on that kind of, I won’t disengage mentally, but I’ll break physically. And in my case it was, I had a heart attack when I, I think I was, I think I was 42 when, when that happened. Right. So about eight years ago. Right. Wow. When, and then when I saw the doctor and after they put the stent in, it was one artery that was 98%.

[00:14:59] and it was from lack of sleep. Like I think I had slept three hours a, a night on average that week. Right. when the doctor said to me, he goes, look, I see you’re not overweight. You are, you don’t smoke, you don’t drink. I’ve never seen this before where someone came in with a completely closed artery at your age without any of, if you had told me this was a case study before, I wouldn’t have believed it.

[00:15:22] it just goes to. , how important is your stress level and your ability to maintain sleep and diet and exercise relative to your ability to continue the, the work that you want to do? 

[00:15:35] Zach White: Yeah. What was the biggest change you had to make to recover from that heart attack in that season of life? 

[00:15:44] was, on the emotional side, it was just like figuring out what’s important and, 

[00:15:47] Leon Kuperman: how do you stick around for your kids? Cause I, I have three of them, right. all the way from 25 to nine right now. and then there is how do you maintain a pace that you want to maintain without a recurrence? Right? So it started with diet, it started with putting exercise higher in the priority list.

[00:16:07] Yeah. And then I started doing a few things like tracking my. . So I know every night I get seven and a half hours of sleep. and I started intermittent fasting and that’s after trying all vegetarian diet, that’s after trying like statins, you know, to get my cholesterol level down. And then I finally hit on this combination that works for me, which is intermittent fasting.

[00:16:29] So I don’t eat for 16 hours a day and that shot my cholesterol almost to. , you know, beyond perfect levels, but almost to like when you were 18 or 19, that’s what your kind of cholesterol readings would be. 

[00:16:41] Zach White: Wow. Well, glad you’re doing well. And what I love about that story, and I’m the same way. And, you know, my clients who sign up for our coaching programs, one of the things I encourage them in is like, we love engineering.

[00:16:54] Let’s treat these things, you know, like an experiment and something you want to continuously improve and learn on just like we do at work. And so what’s working for you may not work for someone else, but rather than throw our hands up or make excuses about it, let’s, keep trying new things, keep testing, keep collecting that evidence and data to learn.

[00:17:12] so important. So Leon, I, the coach in me wants to talk about this all day, but the engineer in me wants to get, get onto the, the juicy bits about your zone of genius. So would you just describe from a technology perspective, your sweet spot, your zone of genius and expertise with what you do now and through the, the decades of work you’ve done, and kind of give us the box that you.

[00:17:35] Sure. So 

[00:17:35] Leon Kuperman: look, so I, let me just take you back in just a couple years, so, I’ve been doing cybersecurity as my primary kind of s point of focus for many, many years, right? let’s call it 15 plus years or so. in the last startup that we did, this is a company called Xen Edge. we decided to deploy the whole product into the cloud, and we said, look, as customers come on, We’ll just order more infrastructure.

[00:18:00] It was in AWS at the time, and we’ll just, our costs will keep up with our revenue. And what started happening was the, the finance team would send my CEO the invoice every month. Yeah. And it would blow out a control because we were growing pretty quiet pretty quickly.

[00:18:16] And the truth is, I didn’t know what to tell him. I didn’t. I didn’t have an answer for why the build was growing exponentially for cloud computing and services For cloud computing specifically, and we went through all of these layers of let’s buy reserved instances, all these kind of discounts and saving schemes that we would try to do that the cloud was recommending that AWS was recommending.

[00:18:36] And while the product was really, it was a good product, I think customers really loved it, analysts loved it. We ended up being acquired by Oracle and it was integrated nicely into the Oracle Cloud suite. We never really solved the cloud cost problem because we just moved them into Oracle data centers.

[00:18:52] the cost of good sold line net and became zero. But did we ever really solve the upside down equation cost? Right. And then during. The beginning of Covid, right? We got this clear message saying, we as a world are short on chips. Everyone needs to conserve their computers. The supply chain is gonna be disrupted for a while.

[00:19:13] So I noticed these efficiency programs with customers and within Oracle themselves, like, Hey guys, we gotta cut down the amount of compute. And it was all a manually, all a manual effort for customers, for us, for everyone. It was a manual effort. And I’m like, I can’t be the only person in the. . We as a team can’t be the only team who hasn’t solved this problem that we’re manually chasing these efficiencies and trying to bring our bills down.

[00:19:36] I don’t want a report that tells me what to do. I want something to do it for me. Right. So we made a bunch of these assumptions around the industry, and then I actually switched my focus from cybersecurity to infrastructure when we started cast. and our goal is to create an autonomous computing environment.

[00:19:53] Leon Kuperman: and we have a lot of different like sub goals under there. But the first one, which is the most prevalent for today, is bring that cloud build down automatically. So you are onboard with us and we just shed away the excess waste in your environment to get you to a solid state, and then you can build from there.

[00:20:09] and then by the way, a lot of those. , algorithms that we run. Some of them are heuristics, some of them are machine learning driven, I think the combination of smart heuristics and machine learning, that’s kind of the, the bubble that I call artificial intelligence cuz it creates the perception from the outside world that there’s a human-like entity controlling these systems.

[00:20:29] Zach White: I love that. And put real quick, in timeframe, when was that switch for you? In launching cast? It was in 20. in 2020. So, so literally right in the middle of, we’re still in lockdown, this idea is coming to light and we launched cast. And where, where is the organization at today in terms of its maturity? 

[00:20:51] so myself and our, our founding teams, ERI Laron, um, we seeded the company with our own money at, which was a luxury that you could have later in your career, right? Yep. Can’t let it check at the beginning of your career. To give us a little bit more control. So, and then we had a group of investors that seated the company.

[00:21:08] Leon Kuperman: We’ve already gone through series A with a dc and we are going through a series B process right now, which is a growth phase. without giving you specific numbers, um, we’ve hit product market fit. So we’ve got well over 50 customers now we grew at a 10 x rate last year. So 2022 was a 10 x year.

[00:21:26] Congrat and we’re gonna try for some similar multiple in 2023. Although it gets harder as you always , 

[00:21:33] Zach White: fine, right? Yeah, no doubt. All right, so I, I mean, congratulations and that’s amazing, and I do encourage every engineering leader to go check out what’s going on with cast. There’s some really, really brilliant technology behind what you’re doing, and it makes sense that every organization that wants to build on these cloud platforms needs to pay attention to this.

[00:21:50] But it, opens the door then with where, , ai, machine learning, all of these trends that are so hotly discussed right now. Where is it going? What does it mean? And so maybe to start with, can you put a lens on what you see is happening in that discussion? What is the narrative today, Leon, that people are hearing?

[00:22:15] Let’s just sort of start with a baseline. What are the different perspectives? What’s the conversation 

[00:22:21] Leon Kuperman: I think here’s what changed in the. Well, let’s call it year and a half, Zach, is that people used to think, well, computers could do relatively mundane jobs pretty well, but they can’t do creative work, right?

[00:22:34] So you can’t draw a picture, you can’t write a symphony, you can’t write a piece of code cuz coding is considered to be creative. You can’t design a machine. Those things are the realm of humans. if you wanna get a, a computer, Do a structured process like a business process automation.

[00:22:51] Sure, they can do that. No problem. What’s changed, and I would say this is a kind of a multi-year trend, is these neural networks have been repurposed Yes. To simulate creativity, and it’s, in some ways it’s more than simulation. So it started with. The team that Google bought, I would say, I’d have to go back to exactly, but have you ever seen this thing on Netflix?

[00:23:14] It’s called Alpha Go. It’s where the computer plays the world’s Best Go player. Yes. Uh, in, I think it was in South Korea and the analysts that were wa and this was a game watched by 200 million people. And the analysts watching this game believe the computer makes this massive mistake. They’re like, , game over.

[00:23:33] This computer’s just fallen apart and maybe 20 moves later, they realized the genius and the creativity of the move and everyone has just blown away that this was an original creative strategy that the computer developed to play against this player. that’s been an a recent evolution.

[00:23:50] So you see now with Dolly is the image generator that can create cr aesthetic artwork that we find as humans, pleas. that is completely unique in original art. The same thing for chat G p T based on the same tech for narrative and prose. And now we see, uh, these models that have been trained to write code.

[00:24:12] So you can, for 10 bucks a month, install a plug-in into your id, write a comment and I’ll give you a mechanical engineering example. Please, please a second, write a comment about what you want the code to do. and it’ll spit out the code and you just make a few edits and you’re good to go. So you wanna hear about this example, right?

[00:24:32] So yeah, please. So my son’s working on a, he’s an environmental engineer, so it’s all engineering in the family, right? and he is working on a system for a laboratory to help understand, plant. Drought stress. And so that requires some automation of watering and temperature controls. all of the environmental controls you’d use to Yes.

[00:24:52] Classic collecting data from. Right. And so he needs, he’s got this little Arduino and he’s gotta connect to it over a cereal port, but he’s not a programmer. Right. So, Several months before we sit down and I’m showing him how to write this all in Python and we’re looking it up in the documentation and it takes us weeks, right?

[00:25:10] so he tries the same experiment with AI generated code, and he pipes in the comment that says, give me a piece of code that connects toward on Arduino device over a serial cable is able to accept commands from the device and then execute submodules for every different. and he gets half a page of code.

[00:25:30] Leon Kuperman: That was pretty much what we wrote together. Oh, wow. and then he, we were reflecting on it and he goes, dad, this is great for me because I’m never gonna be a coder. I’m not gonna go down that career path like you did, but I need to have these skillsets. So for me to casually be able to generate a set of code that that helps me with machine control is unbelievable.

[00:25:51] And I don’t have to spend the 10,000 hours there. I can spend it elsewhere. 

[00:25:55] Zach White: Yeah. Or rely. another colleague to come over and you, you know, you put that in the queue and you’re waiting for someone else to get it done. Exactly. Which was, I’m thinking back to my engineering days, Leon, that’s exactly what would’ve happened in that situation.

[00:26:08] You know, as soon as I get to that threshold of my ability to write code, especially as a test engineer, when I was in the lab and working with these PLC controllers and all the stuff that I didn’t have a clue, it’s like, where’s matlab? You know, I wanted to go back, go back to Purdue and write something in matlab, but.

[00:26:23] Yeah. It’s just, you gotta get help. You’ll reach out to somebody who’s a, who’s already the expert, and a lot of times they’re busy. They’re not just sitting there waiting for your problems. So that’s 

[00:26:31] Leon Kuperman: huge. Yeah. And Zach, you were using these Alan Bradley devices, these PLC devices. Yeah. So that, that ladder logic is super hard for humans to understand it.

[00:26:39] I like, I still struggled with that, kind of, with that kind of machine level code. So you’re absolutely right. It’s much. To have someone help out in those contexts and the machine is just as good as as a person in this case. 

[00:26:52] Zach White: Interestingly enough, you mentioned all these creative examples, and I have a client right now who’s applying AI into music, writing music, really?

[00:27:03] Impressive music and, and he’s being sought after by different companies that want to leverage this technology to, create music on demand, through ai. It’s just amazing. So, okay, this, this major swing in belief. From creative work is the domain of humans and executing with extreme precision and speed, things that we’ve created for computers to do has now been disrupted.

[00:27:29] We’re, we’re all looking at it, playing with it, asking new questions, it has absolutely stirred up fear in many people. What does this mean for this industry? This industry, these jobs, those jobs? You, you know what is gonna happen here? So, Walk us down the path a little bit. Leon, what do you see unfolding in the near term?

[00:27:50] Maybe then we could talk longer term, but like let’s start in the next, you’ll call it one to four years, like in a near term timeframe, where is this going? 

[00:28:00] in the very short term, there’s still a lot of unclarity around the commercial models that surround these AI innovations, right? And so I’ll give you a couple of examples.

[00:28:10] these models are developed with open source data streams, right? So like in chat GT’s case, they’re scraping the web,to, to pull down articles. So when you can give a user an answer without referencing the original source material, there’s gonna be a legal challenge there. Yeah. And this is what OpenAI, who’s the company behind chat?

[00:28:33] Leon Kuperman: G P t has a, and I’m sure they’re thinking about this. . They have to figure out how they’re going to create attribution for the input documents. Right? So Google does this because you click on a link and there’s a revenue and a monetizable way of getting back to the source document. Sure. Yep. But with chat G P T, you get an answer that is a derivative work.

[00:28:56] So what’s the legal ram? So first of all, to get a commercial, a viable commercial model that makes money, we have to figure out this legal hurdle. And I, predict that it’s gonna go all the way up to the Supreme Court. It’s gonna be adjudicated at the highest levels because it has a direct impact on our society.

[00:29:12] Yeah. but let’s assume we get past that hurdle. I think certain jobs are going to be harder for humans to do than machines. And I’ll give you a couple of extreme examples. , the legal field is extremely complex with huge amount of co-dependencies. I think that a big portion of the work that humans do to prepare for legal arguments mm-hmm.

[00:29:35] will be automated by, by machines. Same is true in in gap accounting. Like I don’t see a hun uh, a future A hundred years from now, let’s call it, where there’s a role called an accountant, unfortunately. But by the same token, a hundred years ago, you know, it’s, call it, 70 years ago we had an elevator operator.

[00:29:55] That was a real job. But we don’t have elevator operators anymore. You know, the equestrian population in the United States was, you know, through the. , everyone had horses. We, we don’t, you know, so society evolves. It’s not a negative thing. You see, I think trying to have a resistant strategy where you say, no, no, no, we need to keep these jobs.

[00:30:16] We, we can’t have automation cuz we need to get, no, we need to educate people from, as a society, it’s our job to make sure there’s access to education so that people can elevate their creativity and their. and do more interesting things versus doing the low level on mundane. And Zach, I faced that all of the time with engineers, right?

[00:30:35] With DevOps engineers. Cuz my platform is designed to take away some of the mundane work. And there’s this natural fear of, no, no, no, don’t touch these settings. These are settings that I’ve hand tuned over five years and only I know how to get them. Right. Like, well it is a bit of a protectionist.

[00:30:52] That is ultimately gonna have to fall away. 

[00:30:55] Zach White: I agree with that. And well, lemme just leave there. I totally agree with that. And so maybe further our time today, Leon, because we’re not gonna, you know, solve the challenge of how to prepare and educate everyone on the planet. Uh, let’s focus on engineering and maybe specifically in the world of, of software and, you know, full stack architecture and the roles that are closer aligned to the work that you do.

[00:31:20] What would an engineering leader. At whatever stage of their career want to be thinking about in how to prepare themselves to create value in a world 5, 10, 20 years from now. 

[00:31:33] the really important thing to understand is in the fields that we’re talking about, we have a general shortage of human capability and skillset.

[00:31:42] Leon Kuperman: You know, we’re millions of engineers short globally, right? There’s no shortage of work we should be looking 

[00:31:47] Zach White: at. Let’s, let’s say that again, this is a real, within weeks of this interview, there’s been tens of thousands of engineers laid off in, you know, in the world. It’s, a lot of people may not necessarily like or agree with that statement, so unpack that for a little bit.

[00:32:02] What makes you say that we’re millions of engineers short, at least over some time horizon. . 

[00:32:08] Leon Kuperman: So just in cybersecurity, and this is like a well understood stat, we’re like something like 6 million cybersecurity analysts short, we’re kind of an adjunct to, to the engineering field. So from a technical work perspective, you have to separate the fact that Google, Amazon, and all of these folks overhired in anticipation of ever expanding growth.

[00:32:32] Yes, yes. what they’ve really been doing effectively is they’ve been taking players off the board. So when you pay an engineer so much money that it is not worth it for them to do a startup like thing or take higher risk. Yes, yes. I’ll get paid this, you know, ungodly amount of money and do work way under my, what I, what my potential is versus going out there and hunting for the ne next Great idea.

[00:32:58] Yep. Yep. This is a mechanism for very large kind of what I’ll call oligopolies to take players off the board. Right. they’re realizing, hey, money is no longer free. Yeah. We can no longer afford to take all these players off the board. We’re gonna have to compete and let them go into the world and find their, true calling and.

[00:33:18] it may be that these 10,000 people are the next geniuses or some combination of people that will exit those businesses are the next big entrepreneurs. It doesn’t matter. The market has to readjust. Yes. So, yes. we’re gonna have a temporary blip of unemployment. , will people get paid as much?

[00:33:35] Leon Kuperman: Probably not, but that’s okay. the salaries are indicative of what Overp employment looked like. Yeah. Right. There weren’t enough people. Yeah. 

[00:33:42] Zach White: Thank you for sharing that and, and I really agree with that sentiment. I think it’s encouraging. I hope that, Every engineer I get to talk to will receive that message in a positive way.

[00:33:52] And it’s no insensitivity to the challenge of being let go. It’s a huge challenge. It’s very disruptive to family life and everything. And my heart goes out to people who are in that situation. But there is, and I agree with you so much opportunity. So that said, huge opportunity in this space. Where would someone aim their growth, their education, their readiness for this new.

[00:34:14] a lot of it depends on where your passion lies. Like, so for example, folks that are really, and I’m speaking about software specifically cuz it’s, yeah. Perfect. So, and, and there’s a couple of areas for folks that are specifically geared towards data, we’re gonna produce more data by orders of magnitude every year.

[00:34:34] Leon Kuperman: The amount of storage we need to create is astronomical. So if you can generate a skillset, organizing and analyzing and gleaning insight out of data. That could be in machine learning, that could be in traditional statistical analysis. That is a massive opportunity that’s not going anywhere, right? Like, yeah.

[00:34:52] Yeah, because it, it requires some, also some creativity that machines haven’t yet proven. , capable of, but also in the realm of physical, like when software meets the physical world, there’s also a massive demand for people with a dual skillset, people who understand the mechanical side of the world, but understand the software side of the world.

[00:35:14] And you can call it cybernetics or you can call it robotics. Sure. But these programs are burgeoning and they are exploding. So almost any kind of future mechanical, role, mechanical engineering role will require some intense knowledge of the software side of the business.

[00:35:31] Zach White: Absolutely. I, I can say, and this is humbling, it’s hurtful to myself to say this, but the truth is, Leon, I would not be in a position today to go compete for the type of job I would want if I were gonna go back into mechanical engineering right now because of how disconnected I am from that world and, you know, buy my own choice to take a different path here.

[00:35:51] But I a hundred percent agree it’s going to become essential. 

[00:35:55] Leon Kuperman: by the way, Also hurts me. my guys won’t let me touch production code . Like it’s, yeah, well, my skillset has evolved to be slightly different. I’m building an organization that can execute on a vision versus, I’ll do it for fun.

[00:36:10] But they’ll kind of laugh at me, you know, in Snicker and like, ah, he still uses Python, that kind, you know what I mean? So he still uses Python, , it’s funny, I, I wrote this little low code, um, low code is like a, something you can drag and drop things around and build a UI pretty quickly.

[00:36:25] I wrote this thing for a customer cuz it was just, we needed to get done, and the guys were like, well, you’re cheating. Like, you’re, you didn’t actually write a lot of code. But I’m like, yes, that’s true. Yeah, but we, we 

[00:36:35] Zach White: got the, we got the invoice paid. That’s the, the main thing. . so Leon, then tell me for you, and I mean I know we’re just projecting and guessing, but you have such a unique perspective in this area of your own genius.

[00:36:47] So where is the place that you see the limits? Ai, you know, what we can see versus, which is, you know, sort of the human, uniquely human element of work. Where’s the boundary in your perspective? 

[00:37:03] Leon Kuperman: I think it’s emotional intelligence and, we don’t have enough computers on the planet to generate a single brain, let alone many, many brains, right?

[00:37:13] So we still don’t understand how emotions work. It’s all a chemical process. and you need emotional intelligence to effectively come up with creative ideas that are unique. Not someone saying, do me do something in, in the realm of X, but actually, you know, synthesize an idea from nothing. And you need emotional intelligence to get along with other people and orchestrate a plan that’s bigger than yourself, that’s bigger than a single individual.

[00:37:40] Zach White: Yeah, I love that. Within the idea of creativity as a. Differentiating human element of work, do you think that holds true in the future? Or does creativity take on a new definition in some way? 

[00:37:56] Leon Kuperman: I think it takes on a meta definition where, like for example, if you are, an advertising, let’s say, right? And are you gonna necessarily work with an art, an individual artist to come up with that?

[00:38:07] Or are you going to rapidly iterate. With something like Dolly that helps you come up with iterating versions of your vision where you describe them, but don’t necessarily have to draw 

[00:38:18] video1821240549: them 

[00:38:19] Zach White: yourself, so, yeah. Yeah. It’s interesting. I have a similar bold prediction for 2023 Leon, that the vast majority of bloggers and content creators in, like on LinkedIn, for example, Built their audience by just writing thoughts about a subject and their unique perspective.

[00:38:41] Many of those people will. Driven out of any reason to read their stuff because you can ask chat g p t to write that blog and it can do it in 30 seconds and anyone on the planet can put out a decent piece of content. You know, it’s just a, it’s fascinating. And so the uniqueness of the person who’s willing not just to regurgitate things they’ve heard, but actually as a thought leader, the true willingness to dig in and create something that hasn’t been thought of before.

[00:39:08] I think it’s gonna separate the, the true thought leaders from people who are just. You know, filling the internet with content because what’s, what’s the point? Why would I go read that anymore? I 

[00:39:18] Leon Kuperman: think there, I think there’ll be a class of people who specialize in interacting Yes. With these type of prompts.

[00:39:26] Yes. That are really good at it. Like, so the difference in content will be I know the questions to ask. Yes. This is my competitor 

[00:39:33] Zach White: doesn’t really Good point. Ah, Leon, I really appreciate this and I, I. I’m energized by the positivity of where this leads. And yes, there’s gonna be destruction of certain areas of industry, but the creation of entirely new types of work.

[00:39:50] To your point about the elevator operator, we’re entering a really exciting time for every engineer to consider where I fit. In that picture and I wish we had two more hours cuz we could just keep going all day on this. But I wanna wrap it up and be respectful of your time. So Leon, where can people find out more about what you’re doing with cast, connect with you and understand the amazing work that you’re doing.

[00:40:15] Leon Kuperman: Zach, just before we go, I did wanna point out one, one kind of interesting nuance. I think we’re gonna have a, a educational crisis because of chat. G P T I. people are gonna be plagiarizing through that type of like ask any student, Hey, do I want to write a paper or do I just want to, like, if they’re ask, kids are gonna get really good at this and it’s gonna get really hard to catch ’em.

[00:40:40] that is gonna be a societal bubble we have to get through before we can figure out how to get students to learn to write without the aids. Most people coming out of university don’t know how to write well, like their narrative skills have been. And especially if you get into corporations with PowerPoint presentations, everyone’s just ing everything out.

[00:41:00] the art of writing is being lost in our society, and I’m worried that this type of technology actually leads to the. acceleration 

[00:41:08] Zach White: of that loss. Absolutely. That’s a probably a whole nother podcast. We could come back and record ar around two Leon on what this means for educating period, you know, engineering being the focus of, of my passion.

[00:41:21] But, but yeah. Wow. to your question, um, you can get me on LinkedIn, I’m just, look me up, uh, Leon Cooperman, or you can just go to cast ai. We have this. so our product is free for anyone who wants to install. A separate conversation as to why, um, but you can install it and run it and get all kinds of recommendations for $0.

[00:41:42] Leon Kuperman: And we also have a Slack community chat, uh, chat channel where you can go and talk to community members. And then all of my staff are on there helping customers all the time. So those are the best ways to reach us. 

[00:41:54] Zach White: Brilliant. , we’ll put those in the show notes. I encourage every happy engineer out there to go check out what Leon and his team are doing.

[00:42:01] Leon, to wrap it up today, one of the things I believe, and I’m sure you’ve experienced this, you know, great engineering, great coaching, and maybe the future of using these AI tools has in common that questions lead. The answers follow. And if we want better answers in life and in career, we need to ask better questions.

[00:42:20] So for the engineering leader who’s out, Maybe they’re fearful of the future of ai. Maybe they wanna be a CTO and get into entrepreneurship or do the things that you’re doing. What would be the question that you would lead them with coming out of our conversation today, out 

[00:42:37] Leon Kuperman: of our conversation? That’s, uh, in general, I love the question.

[00:42:41] Why? Because you can ask it iteratively, just like a child would, 5, 6, 7 times. Yeah. You usually get to a root cause of what you’re looking for. Right. there’s this fa I forgot Simon is his first name. I forgot his last name, but we can find out and put it in the notes. Um, he has this great Ted talk about, start with why, and then you form your Yes.

[00:42:59] Simon Sinek. Simon. Simon, that’s right. Yes. That’s right. I didn’t wanna get, I didn’t want to butcher the last name. No problem. Um, but, let’s make sure that the question has the word why in it as. , why do I love what I do? Ooh, let’s leave it to that because that opens the the gate to all of the subsequent answers, like as to what you’re gonna ever do in the future.

[00:43:23] Zach White: Why do I love what I do? And then you can ask why five more times? 

[00:43:31] Leon Kuperman: Yeah. And what if the answer is, I don’t love what I do? Well, that’s an. Do something else cuz you guys will understand that A, it’s a really short gig. It’s really short, right? Like, um, we don’t have any future longevity that gets us beyond, you know, what human mortality is.

[00:43:49] So just spend the time doing what you love. 

[00:43:52] Zach White: Amazing. Leon, thank you so much for your generosity today. Just wanna acknowledge you for the incredible work your team’s doing and for sharing your wisdom with the happy engineers out there. And we’re gonna have to come back for round two eventually. There’s a lot of topics we didn’t get to pull the thread far enough today.

[00:44:05] So thanks again, man. 

[00:44:07] Leon Kuperman: Thanks Zach. Great questions. Really appreciate the conversation