Is the Specialist Executive Dead? with Mahesh Guruswamy, COO at Kajabi

(Intro Narrator at 00:00:00) Today, we're catching up with past guest Mahesh Guruswamy, COO at Kajabi, about the evolution and potential death of the specialist executive. You're listening to Joel Beasley, Modern CTO.

(Joel Beasley at 00:00:17) That's kind of where I want to start. All right, so you were at Kickstarter, now you're at Kajabi. How did that transition go?

(Mahesh Guruswamy at 00:00:24) So Kickstarter, I was CPTO, Chief Product Technology Officer, managing product, design, engineering. At Kajabi, I tacked on customer support and success to my org as well. So I'm COO managing product, engineering, design, customer success, customer support. And it's going great. You know, I decided to come back with the founders when they decided to come back to join the company. And having one person making all the decisions across these business organizations is actually, I find it increases velocity dramatically, for sure.

(Joel Beasley at 00:01:02) And when you reached out to me on LinkedIn, you said something that really caught my attention about the specialist executive being dead, because we've been told forever, specialize, specialize, specialize. That's, you know, the riches are in the niches, all those types of things.

(Mahesh Guruswamy at 00:01:16) That's right.

(Joel Beasley at 00:01:17) What do you mean by the specialist executive is dead?

(Mahesh Guruswamy at 00:01:20) So if you sort of rewind back what I would call the golden era of tech, say, early 2010s, right? So early 2010s is when SaaS was starting to climb up. There's a new business model people figured out. There was the growth of mobile, growth of services in cloud infrastructure, and it was just an explosion of technology. And my suspicion is that that's what led to the creation of these specialist leaders. Right? So, oh, we need somebody to manage cloud infrastructure, so we need a person to run that. We need somebody to do product management for this specific thing, so we need one to run that. And also, the other overlay on top of that was it was ZIRP, right? Like, money was free. So essentially free for a company—not for individuals like you and me, but for companies, it was essentially free. And so combine all of that together, that's where the specialist leader came from because it's like, yeah, we have money to burn, and we think having somebody go deep into a specialty will be beneficial for the company and will help us compete better. And that kept growing, growing, growing until we had the plateau of COVID and the reset. And the point we're in right now in terms of the evolution of tech, we're just becoming commoditized. Right? It's not like rocket science anymore to build a tech product. Ten years ago, fifteen years ago, it somewhat was. So you need specialist leaders to build a specialist product that not everybody can build. But now, I would say everybody can build tech products. It's fairly easy to stand up a new tech service or a product that you can sell into the market. And so tech is commoditized. We're all doing the same thing to a certain extent. And that's where I think boards and executives are looking around and going, do I really need a CPO and a CTO to kind of butt heads all the time and try to come to some sort of consensus, or do I just need one person who can just make the decision quickly and move on? At least that's my hypothesis on why the searches for these combined general manager dev executives have gone up. I think the stat that I read was the search for CPTOs have gone up 110% compared to last year. That's telling. That's not a small number, right? So everybody's going, yeah, a technology, a product-focused engineering leader is good enough, or an engineering-focused product leader is good enough is what companies are saying. And I think that's going to continue until there's some sort of other innovative technology that people can't understand well. Right? I think until then, this is going to be the trend.

(Joel Beasley at 00:04:12) Well, the business's job is to drive outcomes, right? So the more friction you have, the lower the quality or the longer the outcome is away. You can reduce that friction, improve the quality, and achieve the outcome quicker if it's one person handling it from end to end. But it wasn't that possible. It's like we're seeing this consolidation across every role at every level. Are you seeing that within your organization?

(Mahesh Guruswamy at 00:04:35) Oh, yeah. I mean, I'm a good example, right? So if Kajabi—well, in 2010, if Kajabi existed, there would be a CTO, there would be a CPO, there would be a Chief Customer Officer overseeing customer support and service, and there would be three execs. And now it's just one managing all these three functions. And I'm not saying that I am a deep expert in each of these functions, but I know enough depth so that I can make the right decisions for the business. And I'm seeing it specifically in mid-sized companies like Series C, D, E companies. And even the companies that I'm advising, their question to me is, how do I find a CPTO, or how do I find a combined CRO who can do marketing? Consolidation is happening everywhere. And the barrier to entry and learning has dramatically dropped with AI. You go to Claude and say, what is the brand management strategy that Google uses? They'll just spit out whatever the internet has in store, right? It's just easy to understand these things. So, yeah, it is happening everywhere. But I don't think every specialist leader understands that, you know, intuitively that that's the change that's coming. So, yeah, they'll have that tab soon is what I would say.

(Joel Beasley at 00:06:01) When you arrived at Kajabi, were they separate positions and they consolidated, or did with you joining back, it consolidated at that point? Or how did all that go down?

(Mahesh Guruswamy at 00:06:10) Good question. So there was a head of engineering, there was a CTO, and there was a head of customer support. So the only executive role that existed in addition to the other two leaders was the CPO. And when I joined, unfortunately, we had to let the CPO go because it's like, you don't really need, like, you don't hire an executive and then have a bunch of other executives report to this person. It just doesn't make any sense. So we went our parting ways, not in a bad way. And he was like, yeah, that makes sense. You know, you can do product too, so it's fine if you just run this function. And, yeah, we consolidated it.

(Joel Beasley at 00:06:52) And then customer support, did you have experience with that? Obviously, you've had to talk with customers and to be a great product developer, designer, engineer. But had you ever managed customer support before, and was it different than engineering, or is it just people?

(Mahesh Guruswamy at 00:07:10) The people side of the house is the same. I mean, obviously, you're dealing with a much cheaper talent pool because it's customer support. But I had sat in on so many customer support ops reviews that I can understand what are the metrics they track, how do they track success, how do they figure out what is working and what is not working, what are the common problems you'll encounter in customer support organizations. So it was not a huge stretch for me to go, okay, I can just run this function because the customer support organization is the same across all companies. It's not like it's anything unique to Kajabi. So it was not a huge, huge stretch. So when I came in, I was like, okay, let me just run this function.

(Joel Beasley at 00:07:53) What's the delta between your expectation and reality? Did it match up? Was it really far off?

(Mahesh Guruswamy at 00:08:02) In terms of customer support?

(Joel Beasley at 00:08:04) The whole transition, the whole role, you coming in, running these three areas.

(Mahesh Guruswamy at 00:08:09) I think my expectations were very low. I was expecting a ton of friction, you know, in terms of being able to run these organizations. But after the initial hurdle of somewhat cleaning house and making sure the right people are in the right roles, it was just smooth sailing. So the hardest part was in the first few months when I had to tell the CPO that, oh, you know, sorry, I don't think we can make it work because it'd be too expensive on an org. We also removed the management layer in engineering as much as possible because with these AI tools, an engineering leader can manage twenty, thirty people. And one of the, again, another side effect of the ZIRP era was we treated employees like children, but they're not. They're all adults. They know what to do, and you don't have to give them deep instructions and you don't have to take care of them. They're not babies, right? But we have taken care of them like that for the longest time and the engineering managers and directors who oversee them and then senior directors who oversee the directors. So when I came in, I basically flattened that org as much as possible to give the engineering leaders at the top direct access to the engineers who are actually working on the problems on a regular basis. Just the side effect of that was it increased decision-making velocity. The decisions don't have to go up and down the chain. You just decided amongst three people and you're done. So, honestly, the org changes were the hardest to navigate in the beginning. Once those changes were done, everybody was off to the races.

(Joel Beasley at 00:09:49) Have you been playing with building small utility apps and stuff yourself recently?

(Mahesh Guruswamy at 00:09:56) Yeah. I have. I have an app that I just built it for myself to track productivity across the org. Okay. So I take the data from CS, I take the data from product engineering and just stick it into my little cloud app and then see what is going on across the org. And it's like, that would have been a product three years ago.

(Joel Beasley at 00:10:17) I know. See, that's why I asked because I'm talking to all these leaders. I would say 70% are—because they have an engineering background—and they don't have the cognitive overhead now of, you know, booting up an environment. I mean, that could take a whole day just to get your environment set up if you want to build something. And now we can just do it with tools like Claude and Cursor and all of that. I've been building, I've built maybe three or four utility apps. I'll just—you know what I noticed, Mahesh? I have actually figured out that there's often this situation where the time it would take for me to search, test, try, trial and error, different products in the marketplace versus building a single-page utilitarian app that would do the thing I need, it's shorter to build it.

(Mahesh Guruswamy at 00:11:05) That's right.

(Joel Beasley at 00:11:06) I can describe it exactly. It would only build what I need, do exactly what I need, and so I've got all these little single apps for all these things I do in my life now.

(Mahesh Guruswamy at 00:11:15) That's right. Do you use Whisper Flow?

(Joel Beasley at 00:11:18) No. The only thing I'm—my current stack is I make it write everything in Rails because that was my language of choice, or framework of choice as an engineer—Ruby language, Rails, the framework. So I can look at it, but man, I don't even have to look at it that much. And then I use Cursor with Cursor Agents. And then I multitask with the agents, which is fantastic. And so that's my current stack. Whisper Flow is like a text to speech that's better than normal. Is that correct?

(Mahesh Guruswamy at 00:11:48) It is phenomenal.

(Joel Beasley at 00:11:51) It's for getting—

(Mahesh Guruswamy at 00:11:52) Yeah. When I built this utility tool, I installed Whisper Flow because my all my teams are using it. Because, you know how I could—so what they would do—I can look outside my window, inside my office, and everybody's hunched down whispering to their mics. So one day I walked over and said, what are y'all doing? And he's like, yeah, I'm just using Whisper Flow. So they use Whisper Flow to write Slack messages. They use it to talk to their Claude instance when they're coding. They use it for writing down notes for themselves. They use it for everything. Instead of typing, they're just whispering to their microphones. So I started using it, and it's very, very good. So I just combined Claude with Whisper Flow. So I just talk to my Claude instance and say, I want to build a single-page utility app that tracks these metrics across these departments, and I want this and this and this and this. And it has inbuilt grammar correction, inbuilt all the things that you need to do to just describe what you're trying to do. And our CEO, he uses it in his car when he's just talking to Claude. He's just brainstorming with it using Whisper Flow.

(Joel Beasley at 00:13:02) That is wild. So I could probably use it with Cursor then.

(Mahesh Guruswamy at 00:13:06) What's that?

(Joel Beasley at 00:13:07) I could probably use it with Cursor.

(Mahesh Guruswamy at 00:13:08) You can use it with any product where you can type, basically.

(Joel Beasley at 00:13:12) Okay. Yeah. All right. Well, you know what? This reminds me of Superhuman, like the internet product. Where I hear about it from very smart people multiple times, but I'm just like, I'm not going to. And then when you hear about it enough, you're like, okay, I'll try. Dude, Superhuman has—do you use that or no?

(Mahesh Guruswamy at 00:13:31) No. Fun fact, I interviewed at Superhuman. I didn't get it. But I did for a CTO job. This was many, many, many years ago. I forget which job it was. It was a head of engineering, engineering leader, or something like that. But they're a really cool company. I think it's really cool. I haven't used the product, though, because the same inertia like you described. I don't want to pick up a new email client. Do you use it?

(Joel Beasley at 00:13:57) I just started three months ago, and I wish I could go back in time and have used it forever.

(Mahesh Guruswamy at 00:14:02) Okay. Tell me more. What does it do better than your standard Gmail?

(Joel Beasley at 00:14:06) So, well, first of all, it's got these—you can do this thing called split inbox where I can tell the AI, hey, anytime someone's reaching out to me for these reasons—the AI reads every email—put it into this special inbox so it pops up. So you can do all of the rules. It's basically like the rule set is AI prompts. And so you can—so it's got a bunch of good stuff by default as far as categorizing everything. But then it's got the best snippet system ever. So people hit me up to come on the podcast every day, so I've got snippets to respond to that. I've got my sales snippets, and it's just a really solid, stable application. And then you can talk to it, which you can do a lot of this stuff in my paid Google account. But the overall experience of the concept of having—for you, for example, if you had a leadership one, you'd be like, hey, all my leaders who are talking to me put their stuff in this inbox. Then you can go get that to zero inbox because you can reply to them and save them for later and remind you later. It's just—everything you can do with Gmail and all that stuff is just more usable. It's just better, and it's beautiful.

(Mahesh Guruswamy at 00:15:26) Got it. Got it. Yeah. Do they still do the personalized onboarding when you buy a license?

(Joel Beasley at 00:15:32) I did self-serve. It was like $40 a month.

(Mahesh Guruswamy at 00:15:34) Okay. Yeah. Okay. Okay.

(Mahesh Guruswamy at 00:15:36) Back in the day, when you signed up for it, you had to go through the onboarding with a person before they would let you in.

(Joel Beasley at 00:15:43) I hate that. Yeah. No. Maybe that's what I'm saying. Were that the case, I would not have used it. No. I'm not doing that. No. Yeah.

(Mahesh Guruswamy at 00:15:50) So yeah. So I'm going to check it out. It keeps popping up, and they got acquired by somebody, didn't they?

(Joel Beasley at 00:15:56) Oh, they did? Oh, I don't know.

(Mahesh Guruswamy at 00:15:57) They got acquired by Grammarly, I think. I think Grammarly acquired them. So now they're part of that family.

(Joel Beasley at 00:16:06) Yeah. The thing I was doing before is I had like a Zapier workflow reading my emails and drafting replies.

(Mahesh Guruswamy at 00:16:13) Yeah. Yeah.

(Joel Beasley at 00:16:13) And the fact that it can do it natively was just way more stable. You know? Okay. So anyways, worth—I would look into it.

(Mahesh Guruswamy at 00:16:21) Okay. Good.

(Joel Beasley at 00:16:22) But don't listen to me. Wait till 20 people tell you and then do it, because this is how we deal with product feedback.

(Mahesh Guruswamy at 00:16:28) That's right. That's right. Well, you know what? If Claude tells you, it's probably true. Right? It's become like that. You know, if your AI tool is telling you you should use this product, you're probably like, yeah, you know? Maybe I should try it.

(Joel Beasley at 00:16:40) Yeah. I know. And as a human, I love it and I hate it because sometimes I'm talking to a human in real life, like in person, like, "Oh no, Claude said," and I'm like, oh, this is like the Google thing I told it, but 10 times worse.

(Mahesh Guruswamy at 00:16:55) That's right. That's right. That's right. That's right. I mean, on that topic, the one thing I'm noticing with myself and even with my team is they categorize it as AI fatigue. But if I sort of dig deeper into specifically, like, what is bothering you in terms of using this AI tool all the time, it comes down to they feel like their core skills are somewhat eroding. So, you know, we are a Ruby on Rails shop, so—

(Joel Beasley at 00:17:21) Oh, really?

(Mahesh Guruswamy at 00:17:22) Yeah. We're Ruby on Rails. And they're like, oh, I used to be able to just write this without consulting Stack Overflow, whatever. I just memorized all these functions. I could just power through this. But now after using Claude for a bit, and they use Agent mode, which means no typing of any code, right? And they're like, oh, I feel like some of this joy is gone from coding. And I feel the same with me. Like, when I write—I lean on AI quite a bit for content, just like writing emails and social content, et cetera. And I read my last one that I posted on LinkedIn. I was like, oh, this is crap. This sounds exactly like everybody else's content. So I took a breather, and now I'm like, okay. Maybe I don't start with AI when it comes to doing anything. Maybe I just do my own version first and then give it to Claude and say, what do you think about this or make it better or something. Going back to coding, especially with PRs. So Claude generates these PRs, which looks like it's written by Shakespeare. Right? And the comments are like Shakespeare. So my team is like, I know AI generated these PRs and these comments. Should I even bother responding to it, or just sick my AI at it? And because that was the fun of coding. Right? Like, you pull a PR and then you talk about the changes. You discuss it. Now if that discussion is with an AI, then the team doesn't feel like they should engage in it because it's not worth their time. So what an interesting inflection point. I think we might have to say no more AI PRs. Like, you just do it the old-fashioned way because that's not going to take that much of your time. So let's see.

(Joel Beasley at 00:19:15) Have you seen the companies that are like, we don't let humans write code anymore?

(Mahesh Guruswamy at 00:19:20) I've heard a lot of them. My suspicion is I don't think that's going to last long. Right? Because how do you—it's like saying whatever expertise you learned in college and then your experience, you just don't have to use it anymore. So I just—I don't know. It just feels like maybe at some point, like, handwriting code will become like cursive and nobody will care about it anymore. But I still feel like there has to be some aspect of software development where you can apply your critical thinking. Now maybe it moves to some other part of the stack or some other part of the process besides the coding. But that stuff scares me a little bit.

(Joel Beasley at 00:20:07) I think you nailed it too. You've got it. You're—they still have the critical eye. You're still reviewing the code. You're still correcting it, but you're not like hands-on—you're not writing the functions out. You're not writing the variables out. You're giving it an instruction, and you have to be—look, I can tell you, I played over the past couple months with this on how sloppy I can get. Sure.

(Mahesh Guruswamy at 00:20:31) So—

(Joel Beasley at 00:20:32) I'm like, all right. I'm going to go in there, and I'm going to pretend in my head I'm a hairdresser, and I know nothing about technology. I'm just going to try to tell it.

(Mahesh Guruswamy at 00:20:38) And it's—

(Joel Beasley at 00:20:38) Like, okay. And then I went in where I basically was describing the class that I needed, you know, and the functions. And I found this kind of sweet spot, and I've noticed over the past six months, it's actually gotten better. Like, I can be a little bit lazier and then a little bit lazier. Right. As long as I still review and make sure the models and the associations are linking up for things that I want to—like an application. Like, there was an uploads thing, and it just wrote an uploads thing, but I wanted it to be polymorphic so that it could've—because I know I'm going to do more of this in the future.

(Mahesh Guruswamy at 00:21:17) Sure.

(Joel Beasley at 00:21:17) And so I rejected it and said, no. I want you to make this resource polymorphic. I'm going to use it a lot more. But like, if you didn't know this stuff—

(Mahesh Guruswamy at 00:21:24) Yeah. That's right.

(Joel Beasley at 00:21:26) Yeah. Yeah. You might end up with like that feature seven times and like a bunch of bloat, and who knows how it would go.

(Mahesh Guruswamy at 00:21:32) That's right. That's right. That's right. I think, like, you know, innovation and craziness, they're like close cousins. Right? And craziness can only be provided by humans. So we just have to decide consciously where is that being provided in the process. Right? So even with coding, I've had—back when I was coding, I would receive some comments on my PRs. I'd be like, oh, this is—I've never even thought of it this way. Like, I don't know how many moments people have like that anymore. You know what I mean? Like, you know, you're like, AI is doing all the work for you. So I don't know. It'll be interesting to observe where this goes. But I was in the camp of like, throw AI at everything in the beginning, but now coding for sure. I think it makes all the engineers more productive. We can get more stuff done. We can ship faster. But anything that touches content or copy or marketing assets or images or videos. I'm like, I don't—this is like sameness, right, if you go down that path. So it makes me worry for OpenAI, so we'll see what they end up doing.

(Joel Beasley at 00:22:47) This is going to be crazy. What's happened in three years to have a conversation in the next three years, it's unbelievable. I think, you know—do you think that—are you optimistic here with the workforce being cut off? Do you think it's going to be like the mail where everyone was scared that the Internet was going to kill the postal service, but it did the opposite because it got all of these packages to be delivered through online orders. Right.

(Mahesh Guruswamy at 00:23:19) Do you—

(Joel Beasley at 00:23:19) Think it's going to be like that, or do you think it's just going to be destruction today?

(Mahesh Guruswamy at 00:23:23) I think I'm not worried about the workforce being displaced. Right? I think when the technology becomes mature, we'll find better use cases for it, and then I think it'll, net-net, create more jobs. That's not the stuff that scares me. What scares me is all the token costs are being subsidized by VC money. So Anthropic, OpenAI—not worried about Gemini because that's under Google, so they can do whatever they want. But what will be telling is when these two companies go public. I'll be very surprised if OpenAI goes public this year. I think Anthropic will go public this year. If they—so we're—I'm really waiting for that to happen because if the market goes, your token to use—your $200 per month plan, it costs you $3,000 to build it and ship it and support it. The math doesn't make any sense. So then the market's going to react to that, especially if it's a public company. Right? So that is the piece I'm worried about. Like, we have all this optimism around AI, and I think the utility of the product itself, I'm excited for it. And I'm excited to see where it goes. But the way the economics of our market is attached to the success of these AI companies worries me quite a bit. So we'll see.

(Joel Beasley at 00:24:43) Like, that your access to them, the cost would go up? Like the drug dealer thing, they give it to you cheap and then the price goes up. Is that what you're getting at?

(Mahesh Guruswamy at 00:24:52) So when you—when the doomsday scenario in my head is Anthropic goes public, and then the market's like, well, you're charging $200 but it costs you $3,000 to build. So it's not sustainable. Like, what are your plans to bring this cost in line with, you know, what is going on? So let's say they react to that and token costs go up, then not everybody can afford whatever they're doing right now. So usage comes down, and the valuation of Anthropic takes a hit. So that's my worry, and that'll affect everything else that is connected to that company. Anthropic—last time I—this is like, I don't even know. These are not public numbers, so I don't even know if they're real. I think they're close to like breakeven, but OpenAI is deeply in the red. I mean, they're burning more money to provide whatever services they're providing, which is why I believe they don't—I don't think they're going to go public this year because the market is just going to destroy them. So if overall usage of AI comes down because the token costs are misplayed or not accurate, then everybody in their orbit also goes down. NVIDIA, which is going through the roof because of the dependencies on GPUs. They come down. Anthropic comes down. Oracle comes down. All the other companies in the orbit come down. And if this correction happens over many years, then great. Then it doesn't matter. Right? Because it's not going to affect the average person. But if it happens in a matter of months, then everybody will feel it. And that's my worry.

(Joel Beasley at 00:26:31) Interesting. So you're worried like macroeconomics—

(Mahesh Guruswamy at 00:26:34) Right.

(Joel Beasley at 00:26:34) Of these entire industries. You're not worried about your access, the cost of your access going up and it being unusable for you. You're worried about macroeconomic things.

(Mahesh Guruswamy at 00:26:44) That's right. I think that, you know, even if the token cost becomes expensive for a business size company, they'll be like, you know, okay. We'll use it a little bit less or absorb the cost. Smaller companies will be going, why don't we just use the open source models, and we'll just do it ourselves? Right? They can just do that. But there's like so much optimism. You know, they call it the Magnificent Seven. Right? So much economic optimism is attached to the success of these seven companies. If these seven companies don't do well, then we're all going to feel it. It's my worry.

(Joel Beasley at 00:27:18) Yeah. Well, I was going to go into a SpaceX conversation, but it's not an investment podcast.

(Mahesh Guruswamy at 00:27:24) Let's—let's take SpaceX as a good example. Right? So much optimism, but it's now down, whatever, 15%, something like that. Right? So I didn't—it's—I'm not diagnosing a little bit, but I'll go ahead and talk back. But the only money-making part of SpaceX is actually Starlink. That is the only thing that makes money in that entire company. So I don't think it's $200 billion worth of investment. So we'll see what happens if that comes.

(Joel Beasley at 00:27:57) I'm just curious. Are the data centers rolled into that or not? I don't know.

(Mahesh Guruswamy at 00:28:01) Everything is—everything is rolled up into that.

(Joel Beasley at 00:28:05) Right. But the data centers aren't profitable.

(Mahesh Guruswamy at 00:28:07) Right? None of them are profitable. The only thing that makes money is Starlink, which is a great product, by the way. It's phenomenal. Amazing. And everybody is buying it. So that's—if you look at it purely from the lens of markets and investing, then I would look at it and go, why all the optimism? Right? None of this makes any sense from a mathematic standpoint. But there's also the Elon halo effect where he can just basically explain to everybody why it's going to be great. So maybe it'll keep going, but we'll see.

(Joel Beasley at 00:28:39) Or like when he told people not to buy Tesla stock. He was like, don't buy it. Like, it's too expensive. Don't buy. And then that just made people buy it more.

(Mahesh Guruswamy at 00:28:46) That's right. That's weird. That's right. That's right. That's right.

(Joel Beasley at 00:28:49) Oh, goodness. Okay. This AI fatigue thing and loss of core skills, you know, you could say the same thing about compilers and understanding what's going on there. Yeah. And all the way up the stack. Do you think—do you think where it's the same thing just repeated over, and this is just us getting to experience it professionally, what's happened the generations before?

(Mahesh Guruswamy at 00:29:13) Actually, that's a really good point. So it's a really good point. I think it comes down to if it ever comes to a point where you don't understand the different moving pieces that you—okay. Let me rephrase. If it ever gets to a point where you don't need to understand it, then it doesn't matter. Right? Because, like, now, you know, I think nobody needs to really understand how compilers work because you're not going to go debug a compiler. Like, those days are gone. Right? You're never going to log into a box and plumb an IP. Those days are gone too. So I'm trying to think, is that knowledge useful? I think it's useful in some cases, but for the average company that builds and sells SaaS, probably not.

(Joel Beasley at 00:30:02) Right.

(Mahesh Guruswamy at 00:30:04) So play this out. Okay. So you don't really need to understand these details. Does that mean that universities don't teach them anymore? Does it mean—wouldn't that mean that you don't need to be computer science smart to build software anymore? Does that depress everything, like salaries and barrier to entry? And I don't know. Maybe—

(Joel Beasley at 00:30:26) It does. It makes it so basic people—it's giving more access to it.

(Mahesh Guruswamy at 00:30:31) That's right. That's right. That's right. That's right. I think a good analogy is the automotive industry.

(Mahesh Guruswamy at 00:30:38) So I think it was hard in the beginning. It was like specialized skills, right? This is building cars, et cetera. But now everything is automated, so the barrier to entry is much, much lower, but it's like depressed wages.

(Mahesh Guruswamy at 00:30:52) You know, this is what, like, in the fifties you could work at GM and buy a house, right? Just work the line. Not anymore, because those wages are low. So maybe that's what's gonna happen to software engineering. I don't know.

(Joel Beasley at 00:31:06) Well, that's based off of the beginning of the conversation where you say that this is becoming commoditized. That would follow the trend.

(Mahesh Guruswamy at 00:31:12) That's right. That's right. That's right. That's right. I hope to be retired by then, so yeah.

(Joel Beasley at 00:31:17) Well, you know, I also think that for people who try, who care, who wanna stay ahead of the curve, I don't think this is the last industry ever. You know, we're gonna have new industries where that's what the competitive advantage is going to be in, you know, space, or it might be in biologics and like physical things.

(Mahesh Guruswamy at 00:31:39) Yeah.

(Joel Beasley at 00:31:40) You know, new advancements that will, like, for me, you know, I don't have a big background in biology, but the ability for me to get taught right now in real time by Claude is just me asking it, and I can get the basics. Before, you would have to apply for a school, wait for a semester, get a teacher who actually cared and was good and who could actually communicate. There was all of this fluff in between.

(Mahesh Guruswamy at 00:32:02) That's right.

(Joel Beasley at 00:32:03) But now we have just such a high quality educational environment, and everything's pushing forward with autonomous agents, and it's just, it's a beautiful future. I think it's a beautiful future.

(Mahesh Guruswamy at 00:32:12) I would agree. I think the, if we assume that the, not assume. I think the smart people who can reprogram themselves to where the future is heading will always be successful regardless of the industry that they're in right now and regardless of how the world changes, et cetera. So that was true before. I think that's true now, and that'll remain true in the future. So I agree with you. I think you're right.

(Joel Beasley at 00:32:45) You built a course. Tell me about the course that you built.

(Mahesh Guruswamy at 00:32:48) Oh yeah. So the course is called Rewired for the Future, and it teaches specialist executives how to become a general manager. So as we started the conversation talking about how tech has become commoditized, and that is leading to specialist executives being rolled up into more general managers, I believe that's the future. It just happened at automotive industry. I think it's gonna happen to the tech industry too. And that course teaches people to make that change, to go from, okay, I'm a specialist executive. Here are the skills I need to learn to become a general manager. So that's the course.

(Joel Beasley at 00:33:23) That is awesome. And where can people find that? We'll put a link in the show notes, but is there any easy URL to type out?

(Mahesh Guruswamy at 00:33:29) It's like learn.maheshguruswamy.com is the link to it. But I think it's easiest to just get it off of your show notes.

(Joel Beasley at 00:33:37) Yeah. I don't think most people know how to spell your name. Just being real here.

(Mahesh Guruswamy at 00:33:43) Yeah. That's right. That's right.

(Joel Beasley at 00:33:47) Now, do you have kids?

(Mahesh Guruswamy at 00:33:47) I have one son. He's 12 years old.

(Joel Beasley at 00:33:51) What are the 12 year olds, how are they thinking about AI right now?

(Mahesh Guruswamy at 00:33:56) So he doesn't consume a lot of AI content. The only time he uses AI, he uses Claude is to double check his work, his math work. So that's the only time he uses it. The school has gone hard into not using AI for any other work that the kids need to do, like writing, reading, narrative, math, science. They're like, no AI. It's okay to do research using AI, but then they want the original sources linked out when the kids write their essays, et cetera. But purely from a social standpoint, the 12 year olds are not using AI. I mean, they'll use like TikTok to stitch things together, make videos and whatnot. But other than that, they're not using it at all.

(Joel Beasley at 00:34:57) Yeah. What are your thoughts as like personal preferences? I've got three kids. They're all 10. So they're not really there yet. But what are your personal parental preferences on letting them have, like, if he came to you tonight and he was like, hey, dad, I've got some questions about the economy and some of these topics. I wanna talk to Claude about them. Would you encourage that? Would you put guardrails? Like, how do you handle that as a parent?

(Mahesh Guruswamy at 00:35:25) I think the way I've done it so far, I think I'll continue to do that is, hey, look it up on Claude, but let's have a conversation about it, right? Because I just don't want him to take whatever Claude says as gospel. I just want, let's have a conversation about it because there's always like two sides to any question. In terms of, he has never, he hasn't asked me yet, but in terms of discussing personal situations with Claude, I'll be like, hard no. You come to us. Go to one of the parents or, you know, prioritize talking to your friends over talking to your AI agent or whatever, right? So, but so far, the, at least as I said, the kids are gonna be okay. Is that the vibe that I get from the kids of like 12 year olds is that they're gonna be fine. At least that's what I think.

(Joel Beasley at 00:36:16) Does he have like open access to it, or does he have to go through, how do you set that up as a parent?

(Mahesh Guruswamy at 00:36:22) He has access to all of our phones. I mean, we all have access to everybody's phone, each other's phones. So we just like, he opens mine and opens Claude, and then he starts using it. That's the way he does it.

(Joel Beasley at 00:36:35) Does he have his own phone?

(Mahesh Guruswamy at 00:36:36) He got his own phone, yeah.

(Joel Beasley at 00:36:37) Yeah. So he has Claude on his own phone?

(Mahesh Guruswamy at 00:36:40) He does not have Claude because he's not interested in it. So, and he didn't install it there. But if you want something from Claude, he'll come use one of our phones, which is super rare. So he's not into it at all.

(Joel Beasley at 00:36:52) Okay. Interesting. Yeah. We're, we're just, the only reason I'm asking this, I know we're off topic for the interview, but like I've got these three kids. They're eight, seven, they're about to be nine, eight, and four. Okay. And we got a bunch of birthdays coming up. But they're starting to show interest. They're asking, they're coming over to my phone and saying, hey, can we look up dirt bikes? Or like, can I, and then they see the search and they're like, can I do it? And so we're at this point where they're just getting to that age where they're starting to wanna, and we have their, we have their iPads locked down to like only stuff that we approve to install. And so my wife and I were just talking about it lately, and I was just curious from you how you're handling it.

(Mahesh Guruswamy at 00:37:36) We do the same thing. The phone, his phone is locked down. Like, we have to approve the apps that get installed. But almost like 90% of his requests are just games of different types.

(Joel Beasley at 00:37:46) Same.

(Mahesh Guruswamy at 00:37:47) So which is fine by me. And, but if he is playing something by himself, then I'll be like, okay, well, you need to take a break. But if he's playing with his friends, then like, okay, it was like building, learning social interactions, because they get into fights and then they figure out how to apologize to each other and all that, which is like great. That's fine to do. But if he's just by himself, then I'm like, okay, maybe you need to take a break. The only exception is if he's playing chess with himself with the computer, then great, you just do that.

(Joel Beasley at 00:38:22) Yeah. We've been, same with us, and then we'll have a, we have a rule on YouTube where they can only watch, like, Backyard Scientist. There's like these three or four channels that we've approved that are, Mark Rober. If you have any...

(Mahesh Guruswamy at 00:38:38) Yeah. Yeah. Yeah.

(Joel Beasley at 00:38:39) Yeah. If you come up against him, that guy is awesome. But yeah, because I'm okay with them sitting down and watching the TV, but I don't let them watch like emotional drama stuff that much. Like, we'll let them watch it like a movie on a Friday night or something. But they're not gonna sit there for four, if they're gonna sit there for several hours watching TV on summer, it's gonna be scientists building things, experiments with physics, you know, chefs making food. You know, they're gonna be looking at skills that I wish they had.

(Mahesh Guruswamy at 00:39:06) You're a better parent than me.

(Joel Beasley at 00:39:10) Not by much. I'll tell you what. If we looked into it, I'm sure you're gonna find some real messed up areas.

(Mahesh Guruswamy at 00:39:17) I think the thing is, right, the peer pressure is a real thing. Right? If these kids in school are like, we're watching this anime, then my son will be like, I wanna watch that too because then they get to talk about it together.

(Joel Beasley at 00:39:32) That's important. It's important that they...

(Mahesh Guruswamy at 00:39:33) It is important for kids to fit in. So sometimes I give in, like, okay, you wanna watch this anime? It's like, is this level, like, is it too violent? So I'll watch some of it to see how violent it gets. Yeah. And then I'm like, okay, fine. This is okay. You can see this because it's like cartoon violence. It's not like real violence. But that is sometimes, like, parents take a hard stance against all of that, but I'm like, you need to fit in. So if this will help you fit in with your crew, then fine, right?

(Joel Beasley at 00:40:04) As long as the show is approved, that's fine. Just like, we're in summer right now. So, you know, when it comes to them wanting to watch stuff for long periods of time, I'm just like, nah. You gotta, no more Bubble Guppies. You need to watch something else, you know?

(Mahesh Guruswamy at 00:40:23) That's right.

(Joel Beasley at 00:40:24) Awesome. Well, we made a podcast, my friend. It's always good talking to you. How do you feel?

(Mahesh Guruswamy at 00:40:28) I feel great.

(Joel Beasley at 00:40:30) 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.