Latest / Fractional Frequency / The AI Workforce Reckoning with Kirk Havens

Transcript

The AI Workforce Reckoning with Kirk Havens · Fractional Frequency

• speaker-0: The speaker-1: Mm-hmm. Today we are joined by Kirk Havens, ⁓ from the US Army Special Forces to VP of Cybersecurity at NIQ. Kirk has seen a lot and turns what he's experienced into really smart and practical advice, and that is why we love talking to Kirk so much. Welcome, Kirk. speaker-2: Thanks for having me. Happy to be here. Excited to get started and ⁓ hopefully ⁓ shed some light on some of my experiences. speaker-1: Yeah, I'm excited to talk to you. ⁓ okay, so let's start with the basics. Walk us through kind of your career journey as it's been so far. speaker-2: ⁓ in terms of starting fr from special forces and kind of moving my way up. Yeah, so I, you know, I actually joined the military pre ⁓ you know, prior to nine eleven, straight out of high school, went to college while I was in the military, and then a few years into that journey had decided that ⁓ you know, I wanted to do more, right? I wanted to experience more, I wanted to be challenged more. And that's when I had ⁓ assessed for special forces and then spent really the bulk of my career military-wise there up until 2013. ⁓ And that really afforded me a lot of different experiences, a lot of different exposure. That's where I had my first real, what I would call like technical exposure and kind of marrying up our mission set with technical capabilities. And so ⁓ it was interesting because it was taking somebody who, you know, maybe had a ⁓ more crude set of skills and refining them down and getting taught things that were very hyper technical, and then learning how to leverage them in that in that entire mission set. It was ⁓ it was very eye-opening and it played a major role in my transition later on to ⁓ the corporate world where I was able to leverage ⁓ you know, business education with the military experience with some of that technical foundation capability and the intelligence work that I was doing. And really you know, kind of parlay that into a role providing cyber intelligence to a ⁓ Fortune five hundred financial institution, which was really different and groundbreaking at that time because it was ⁓ was not something that people had heard of. it was not a role that many companies had, but it was at the turning point in which cyber attacks became, you know, really especially prolific around the time where TJ Maxx and Target and places like that had ⁓ just experienced their own you know, major breaches and they had major exposure. So it was a it was interesting time, from an experience standpoint and then, you know, entry into the ⁓ civilian sector. So pretty fortunate to have lived through those things. speaker-1: Yeah. And how do you think your experience in special forces kind of shaped how you assessed risk and threats in the corporate setting? speaker-2: I don't think that there's a really a lot of difference. Really at the end of the day, ⁓ it's all about the impact, right? So the difference of the impact is, you know, one end you got lifeline mindsight, you know nation state, ⁓ strategic interests. The other end you've got strategic interests of the company, assets that are at risk, right? So it's really the same framework and the same formula, just in a bit of a different format in terms of what you're protecting in how you look at the impact of the risk at the end of the day. And in s in a lot of ways too, that helps you stay grounded, I think. ⁓ because I see some people will come, even even non-military personnel, will come into the risk management space within corporate America and they'll look at things as, you know, very ⁓ catastrophic in nature. And that tendency to catastrophize is you know, closely tied to anxiety and I think you know sometimes angst gets caught up in the in the mix. speaker-1: Yeah. Yeah. Especially when, you know, obviously with the work that is done in corporate settings is very important, but normally there's no lives in danger. Like there could be there could be lots of money in danger, but generally no generally no lives. so it wouldn't be a conversation with anyone these days if we didn't talk a little bit about AI and how that's changed the landscape ⁓ of the work that you do. speaker-2: Yeah, I mean it's changed substantially. AI has changed things substantially for us. ⁓ you know, if you look at it purely, and I'm sure we'll dive into you know deeper questions, but you look at it purely just from the cybersecurity lens and the threat landscape that exists, you see now ⁓ AI enabled attacks through penetration testing tools, ⁓ penetration testing capabilities, exploitation of vulnerabilities that are out there in the wild. speaker-0: at it. speaker-2: With a mean time to exploit that goes down now into the measured into seconds, right? In some cases, there's exploits being developed and weaponized before they're ever published, right? In in much easier fashion. So you see some news recently tied to things like Mythos. ⁓ if you're familiar with Anthropic and you know what they're talking about, and the ability to have success in developing exploits in ⁓ different systems, different software. That creates a a whole new set of challenges. ⁓ but what I like to remind people of is that we're all faced with the same challenge. And so that creates a situation where, you know, now we're forced to change our behavior, which can oftentimes be uncomfortable. speaker-1: Mm-hmm. And why do you think that is? Why do you think why do you think it's having such a polarizing impact on people? Not just the threat, but even the weaving it into what into our work. Why why is that change, this particular change, so hard for people? speaker-2: Yeah, it's a good question. ⁓ what I would say, yeah, I'd counter back with a question to you. Is change comfortable? speaker-1: I mean, but it's consistent. Like, you know, I can't think of a time that we haven't had that we haven't had some big event or something come into play that's impacted business ⁓ or the world, you know, so it but the reaction for AI is is pretty extreme. On both sides actually, like the people that want to be perceived as very pro-AI, like they're they're in all in. Like no thought, just all in. Right. And then you've got the people who are anti-AI and they they won't look look at any of the benefits. So I'm like, it's it's it's jarring to me that there's so much extreme belief. Yep. Rather than like a measured application of a t of a technology. speaker-2: I agree. And and I think that again, a lot of it is anchored in not just the change, but I think it's the rate of change, right? And then so you take that rate of change and you match it to some of the hysteria that you hear in the media that are that's aligned to both of the both sides of like what you just described. And so you get then you end up with really what is just people kind of what I like to define as like tumbling through the sky, just trying to find something to brace themselves against and some reality to anchor against, right? And ⁓ and sometimes, I mean, you even see it where it starts, you know, feuds like internal disagreements, you know. You don't know what you're talking about, you don't know what you're talking about. It's like, okay, well maybe none of us really know what we're talking about. Right. And like we need to settle on facts and not be so, you know, dramatic and use all this hyperbole, right? speaker-1: Yeah. speaker-2: There I think that there should be some merit given to some of the suggestions that have been made that sound like hyperbole. and there needs to be some consideration for how we bring people along and how we bring companies along and how we bring governments along. ⁓ when I say governments, I mean, you know, down to local level, you know, public education, you name it, right? ⁓ I think what is you know, to borrow from your words, change is constant. And what would be wrong for us to do is to try to ignore this and think this thing goes away. which I think is what unfortunately a lot of people prefer. And I understand, I'm I'm actually really sympathetic to it. Like I understand why people feel that and I understand why they want that to be the case. But I also worry that they're missing the bigger picture. And By not adopting, it could have bigger effects than they are anticipating against themselves, their family, children, spouses, etc., friends, you name it, right? Yeah. ⁓ but I think again, the uncomfortable part about all of this is that it requires a behavioral shift in how you do business and how you live your life on a day-to-day. And so the more that you become acquainted with how these things work and how to leverage them and you build a level of comfort and your teeth are cut and it's like, okay, now I can bring some perspective into the equation. So now I understand both sides of the equation and perspective, you know, I think is always a good antidote to h to hysteria. speaker-1: Yes, I would agree. There's a couple of things I just want to double like click on that you said that I think are really important. One thing is the the misunderstanding of what people are trying to say in conversations. And I think this is because every day there's a new like catchy buzzword for something, like whether it's like an agent or whether it's a, you know, a type of engineer that's, you know, an AI expert as of two weeks ago. ⁓ you know, there's all these different things. And I think what I recommend to people when you're sitting down and having a conversation about an AI implementation or an idea to have a forward deployed engineer, like what does that even mean? Like, let's let's break down like what we're actually saying. Let's get clear on the definitions of what we're saying. ⁓ so that we're all understanding each other because everybody's got their own idea of what they think it is. Yes. And then they're come to the table with that, and then there's no curiosity of the other person's understanding. And so that's why it gets a little bit tense. and then I think as well, just the the enormity of it, like. People are doing that to themselves. Like you don't have to have all of the answers tomorrow. ⁓ you can take your time to figure out what makes sense for your business. And you know, just rewarding the the the application is missing the point. speaker-2: That's right. That's all. speaker-1: There's people that are building AI processes that take longer than if you just use the original tool. And so what what's the benefit? Like, okay, you get to say you're using AI, but what was the goal, isn't it, to just like be efficient and get things done and make progress? And so I I personally think there needs to be curiosity as with the passion. And then if you're not passionate about it, Like I'm afraid you're gonna have to lean in a little bit. You know, not everybody is comfortable with the idea that jobs could be impacted because you can automate something. but I think the you should only be worried, and this is the last piece and I got my soapbox, but you should only be worried if you have nothing in your head of value. Like if you if you've learned nothing in your career, then yeah. Probably you're gonna get replened. speaker-0: Thank you. speaker-1: But shouldn't you anyway? Like you know? speaker-2: The grift is over, is that what you're saying? speaker-1: Yeah. Kind of. It's like okay if you have no original thoughts or you can't draw on your perspective to give a point of view and direction or make a decision. Okay. Maybe your role will get impacted. speaker-2: Sure. Sure. Yeah, I I think that it's ⁓ it's inter like what you're talking about and describing to me is like there's a there's not a lot of critical thinking that goes on when you think about the application of AI and like what it means to you personally, professionally. And that would be like something that I would challenge people to think about, you know, like I I'll give you a like a personal example. ⁓ you know my own kids, teenage kids. You know, they'll they'll make fun of me at this point and they'll say, ⁓ did you get that from AI dad? And I'll say to them, Do you say that like it's a bad thing? And then it's that pretty much ends the conversation, right? Because it's like, is it a bad thing? And I'm like, I don't know, is it? I'm ⁓ I'm leveraging something that is tapping into every single available source of information on whatever topic. speaker-1: Yeah. speaker-2: And I'm inferring like what the next best step might be to tackle. Does that make me bad, good, imperfect? What does it make me? speaker-1: Right? Yeah. speaker-2: So you wanna label me because of an insecurity, but have you critically thought through this? speaker-1: Yeah. Like is it humanly possible to retain that amount of data that you can pull? No, that's why we do machines. Right. ⁓ I'm curious, and this is a total side question, that's just a one thing, but what's the what LLM do you like to use the most? speaker-2: ⁓ so I'm I'm ⁓ I'm partial definitely to using Chat GPT. I'll and and that's just because most of that has to do with exposure and what I decided to like teach myself on. ⁓ but Claude is definitely ⁓ very good. I've leveraged it as well. I would say that from an extensibility standpoint, Claude has to be number one. Chat GPT close follower. ⁓ you know it depends i the application of like what you want to do also will drive like what you're using it for. But from a universal standpoint, I think Claude and ⁓ Chat GPT are you know the top of the the top of the pile. speaker-1: Yeah, yeah, makes sense. speaker-2: But that is again, that's very I just want to be very specific because you know you could publish this and people will wanna sharpshoot this. Like there are niche use cases for many, many things. But if you're asking me from a like a blanket speaker-1: Is it for you? Yeah. What do you mean by using? speaker-2: That's that's how I look at it. speaker-1: And then just to kind of wrap us up on the AI, well, from from my perspective, AI and cybersecurity, what do you think is currently being missed ⁓ or misunderstood about cybersecurity needs in the AI era? speaker-2: ⁓ boy. ⁓ this is a very ambiguous question with a probably a very lengthy response that I'll try to truncate. ⁓ there are a few things. So the the best way to think about AI in terms of cybersecurity is it can be an amplifier. Both from a ⁓ from a productivity and effectiveness standpoint as well as from a d like a an attack. perspective, right? So for in the attacker's hands. So an example would be if you're really good at you know A, B, and C with AI, you're going to be awesome at A, B, and C. If you're not so good at A, B and C with AI from an attacker's perspective, you're going to get buried underneath A, B, and C. And so really like what that's a signal to do is if you're going to, and there's there's there's It's multifaceted, but again, I'm gonna make this very simple. If you look at things from the perspective of AI for enablement, enabling the cybersecurity the team and defenses, versus AI from the attack perspective and what you need to do ⁓ you know, from defending against that, what you end up seeing is is that you have to be really good from the internal standpoint at the basics, and then you can talk about enabling people with AI. Because, back to your some of your earlier points, if you don't have very clear ownership and distinct outcomes defined, you are going to spend a lot of time and likely a lot of money and be no better for it. And then when you're designing your defenses to protect against existential AI-related threats, you have to understand yourself and the existing control plane that you have in your environment and how well that is set up. speaker-0: No. speaker-2: and some of the challenges that you may face as a result of its disposition. And I think that those those are really the guiding principles or should be guiding principles for cybersecurity teams as a whole. speaker-1: I almost think that needs to be the guiding principles for like all applications of like be clear on like what your outcomes are supposed to be. ⁓ yeah. speaker-0: All speaker-2: It it is and I'll ⁓ you know, I know Aaron has some questions and I'll refrain from elaborating 'cause I wanna be courteous to ⁓ to Aaron. But I think we'll I think you know, we'll go down that rabbit hole a little bit. speaker-0: Mm. speaker-1: Yeah. I'm curious, and I'm I'm I'm gonna go off our sort of preset questions just a little bit, Kirk, because this sort of popped into my mind as as you and Amy were talking. But you have really been on the front lines of seeing major changes in the world of work and defense, you know, both with the dot com era and in the military, and then now with the entry of AI in in the corporate world. And we've talked a little bit about change, ⁓ but kind of curious from your perspective, in particular in these large environments, what has to be true in order for a a change to be embraced or or you know, something like AI to be deployed and and for it to go well? speaker-2: so that is the that is the number one thing to define what has to be true for it to go well. ⁓ and that is where That is where I think a lot of places are a little bit or not a little bit, maybe a lot bit wayward. ⁓ and I think I understand ⁓ why things are the way that they are. You can see in the last couple of weeks, the the CEO of Coca-Cola and the CEO of Walmart both said, Hey, you know what, we're gonna step down. Like we're not ready to go through this transformation. Like we've seen enough, we're we're done. ⁓ and and you know, two I actually give them a lot of credit for that. I think that that's a that's a bold statement, ⁓ very honest, and I think I understand why they arrived at that conclusion. ⁓ what has to be true is you have to be open-minded and understand. Again, I've touched on this already once, maybe twice. ⁓ you have to you have to be very understanding that this requires a wholesale behavioral shift. Wholesale behavioral shift. And what I mean by behavioral shift is you're gonna have to change culture, you're gonna have to change systems, you're gonna have to change the things that people were comfortable with, again, change, right? The things that people were comfortable with will no longer look the way that they used to look. And what's difficult if I talk about it just from the cybersecurity perspective is if you're not you're a in in unless you're in a company where you're selling cybersecurity products. You're a support function. You're not what I would characterize as the main effort, right? The main effort is we build the products, we sell the products, we generate the revenue. Very difficult from the support function side of things to ⁓ generate change because the main effort is really the main effort, right? And so you need all things to be aligned in a co-equal way in order for it to be seated and function well, right? So that means like Hey, you have to acknowledge your warts, set aside your differences, and align on a new way of doing things. Which if you think about that just from a ⁓ an enterprise transformation perspective, forget AI for a minute. That is difficult in and of itself. Yeah. Right. Many companies have tried that and many have failed at that, and many have failed because of trying that. ⁓ you know, so like that's a and people are starting to see that, ⁓ you know, and and I can elaborate on that more. ⁓ but again I wanna be, you know, courteous to you. People are starting to see that wait a minute, we can't just say, Hey, we're gonna do AI things. That's starting to become evident now. Yeah. speaker-1: Yeah. Yeah. No, it's so interesting. ⁓ go ahead, Amy. I was just gonna chime in because it just made me think about something that is so obvious but just like popped into my mind. The bigger the organization is, the almost like they're at a disadvantage. speaker-2: Yeah, yeah. speaker-1: If you're a big company, getting that thing to change is a huge effort. speaker-2: Correct. And so that is ⁓ and I I love that I didn't have to say that to you, that you drew that conclusion. ⁓ because that's more or less I was going down that path. And so there is ⁓ this is kind of like a thing as of late that I've had this maybe like an epiphany of where you know you start to go down this the implementation path and there's a there are a lot of Considerations that have to be had. And what you realize is like, wait a minute, there's correlation of the size of an organization and the implementation. And so there's there's there's actual like parochial knowledge that's tied to this. It's just it's simply systems economics. And then inside of systems economics, it's complex complexity economics. So when you look at systems economics and complexity economics in contrast to you trying to deploy AI, what you realize is is that large organizations can be at a I say can be, not are, can be at a disadvantage. But I my my belief is is that the the longer you wait to do the adoption and you let this play out, if you do it too slowly, that is when you're gonna start to see negative impact. And in and I in On the flip side, if you do it too quickly without the right guardrails, you're also going to see negative impact. ⁓ and that is something that I a lot of organizations are grappling with right now. So you'll see some stories that'll pop up every now and again, ⁓ where it's like, ⁓ we implemented AI and we got rid of all these people, but the AI cost us more than the people did. It's like, right. Exactly. speaker-1: That's probably gonna happen for a little bit. speaker-2: Well, what I always tell people though is is like it this is it tends to be these conversations tend to get emotive like very fast. But I'm more of a fan of disarming and getting back to like the core of the problem, which is did the AI do that or did we do that? speaker-0: Mm-hmm. speaker-2: And it's like the AI wasn't wrong. It did exactly what you wanted it to do. Right. And then it's like, ⁓ right. speaker-1: But there is gonna be, you know, a time period that the human automation mix is in flux. And the the cost up front probably is gonna be more than a wage bill. ⁓ but again, like that is going to change. What concerns me, I mean, because how much money is wasted in corporate America every day. So watching this happen is no surprise. but what I think amazes me is just that like nobody's really clear on like what the goal is, what they want the people experience to be of their organization. You know, it it boggles my mind. Like, what's your long term strategy for your business? Okay, you've got tools and technology. And then they have changed exponentially and gotten way smarter. Great. Yep. So what's your business goal though? Like stay focused on that and then you can work from there in terms of what makes sense from an automation perspective. But don't sit there with like your process that you've had since you started your business, and then just think if you buy a tool that you're gonna 10x or 20x your output. Like that's just not. speaker-2: Especially if it was a bad process. You just ten you just ten times a bad process. speaker-1: Fail faster. There's so much conversation about what the workforce will look like. And the aspiration that at least I'm hearing is that teams will become very small and you'll have this finite number of highly skilled, highly talented people that are then enabled with these AI tools to do what was previously maybe the job of a team of 10. Do you think that's realistic? And speaker-2: But you know. speaker-1: You know, yes or no, maybe. And then it tell us a little bit more about what you think the best case scenario is for like an integrated sort of AI or agent and human workforce. speaker-2: Do I think it's realistic? A million percent yes. ⁓ I actually think that it's the biggest risk to existing organizations because your ability to now create something from scratch with minimal overhead, minimal debt, ⁓ with minimal friction, and take market share from existing organizations that do the same thing, that is on now in everybody's hands. Right. Like up I'll give you an example tacitly related. My wife has a she does this design work on the side, ⁓ interior design kind of work, kind of a side hustle thing. She loves it, she's great at it. People person, they hire her because they love her. It's great. My son gets this idea that he said, Hey, you should build an app, Mom. So okay, you know, tell me more. You know, I need the paywall, I need this, I need that. And he While we were out to dinner one night, he starts fashioning up an app on his iPhone. And he's got a functional app by the next day. Right? So, like, think about companies that do th similar things when you house and things like that, right? Where you can post pictures and do things. He just basically replicated similarities. Now it's not perfect, right? It's not ready to, you know, make a million dollars. But the idea that a 14 year old kid can spin something like that up. While having pizza with his family. What does House have to think about that? What does Pinterest have to think about this, right? ⁓ in terms of like what would have to be right for adoption, there's a new concept. ⁓ and I think it's almost like ahead of itself, ahead of its time. There's a new concept in technology that's out there, and I think that it's it it's perfect. It serves speaker-1: Yeah. speaker-2: It should serve as the anchor point for all business functions. Anything all your your your ⁓ universal services within a company, right? HR, finance, legal, technology, and it's this concept of ⁓ observability pipeline, right? So if you think about it, AI and a agentic AI, and the term agentic agency means like To have agency to make decisions. If you if you think about it, it's basically a person. Right. I'm not giving it all like dissentient qualities, but it is a person. So let's think about it in those terms and how we manage it. An observability pipeline is set up with policies, provide standards and how to build something, provide standards on the identity, what it can and can't do, what this person can and can't do, right? ⁓ It's governed, it's monitored, it's all of those things, just like you would for one of your employees, right? I think if you anchor to that observability pipeline concept and you define in each one of the the verticals of the the shared service verticals, ⁓ what is and isn't acceptable, you're gonna be leaps and bounds ahead of where anybody else is. Because that right now, ⁓ you know, this is This is industry wide. What I see is you see a lot of people that are like getting these, they're getting these mandates, do something, right? A lot of that is pressure from investors, ⁓ you know, pressure from the market because hey, this buzzword means, you know, n percent ⁓ increase in value, right? ⁓ and then unfortunately a lot of employees hear it and it's like, they said do something, so we're gonna do something. And then they go and do something not realizing ⁓ hey, there's really not a lot of value to this, or hey, you just expose us to a lot of risk by doing this, you know, and then it just it runs until it doesn't anymore. And that's where I I think a a lot of companies are starting to like the checks coming due now. A lot of companies, and they're starting to see these tabs, and it's like, wait a minute. We didn't count on that. We thought we turned this on and we gonna we're gonna get value out of it. and that's just not been the case over and over again. Again, that's not AI's fault. speaker-1: It's the application. I wonder as well if we'll start to see a rise in people thinking about governance or even ethical application or not that not that ethical in terms of like the big picture of wiping out jobs, ethical in terms of being transparent when it is AI that's generated the work or being mindful of where a human interacts with your product and what the best solution for potential challenges might be and where risk could heighten. ⁓ I wonder if we'll start to see a rise in like jobs that would help with that or if the expectation is gonna be that that's what your C level does. They govern the they govern the application. But I don't know if that's the right fit. What do you think? speaker-2: I think you're starting to touch in the complexity economic side of the equation. ⁓ this this gets really what so this gets really complicated really fast. So what you're describing is is you need like you need an agent to watch the agents. It's a real thing, by the way. It's a real recommendation, a real application. But like the more you layer that complexity on, the costlier. speaker-1: Okay. speaker-2: more costly things become right and so that is a very that's a very difficult problem to solve ⁓ from an ethical use standpoint that is where there's there's there's ethical use by the user within the constraints of a model and then there's ethical applic or there's ethical model and is the model itself ethical right so certainly you've seen Some of these things that have gone around where it's like, well, I'm gonna ask four different models, four different questions, and you're gonna get four different results. Or you might get two or three that are similar, and one is just radically different, right? So, what is the source of truth? But these are questions like that, you know, those are regulatory oversight questions that need to be asked at a much higher level. speaker-1: Yeah. When do you think that'll Yeah. How quick do you think they'll get to us? But like is everybody just kind of sat watching waiting to see what could go wrong? Like what speaker-2: Happen. When It's weird because I I think from the so I think from the governing and and regulatory oversight, like through government sense of things, I think what we're likely to see is glacial change. But the the reality is like internally, you know, s Department of Defense, their you know, intelligence community, they're they're they want to leverage this stuff and are leveraging this stuff as fast as humanly possible. But it it is in in the same way that that businesses do, 'cause it is powerful and it is very useful. However, unchecked, like you start to s you start to have or you should, I think we should, question, where does the concentration of power start to sit? speaker-0: Mmm. speaker-2: And now it's concentrated truly within these companies that are the frontier models. And it's like, Hey, who's really driving what happens next year? Like who are the real power brokers and then who's watching them and who's watching what goes on there? speaker-1: I mean, the challenges as well is that, you know, you've got the companies that are creating the technology that will benefit from it being widely adopted and they're taking charge of the narrative in a lot of ways. But it's not like our government is set up to deal with this kind of problem. I mean, it wasn't that long ago that There was a they were interviewing Mark Zuckerberg and they didn't even know how Facebook Messenger worked. So how can you expect these people to come up with a regulatory philosophy on something they could they probably don't understand? speaker-2: Right, and I I I would say in addition to that, it's not an accident that the folks that are responsible for these frontier models are also volunteering to help draft regulatory oversight frameworks. Right. There's there's some value in being part of the solution here. speaker-1: Yeah. And who else would who else sorry? I'm just like where else would you go though? If you don't go to, you know, your tech leaders, you can't ask your government what who where do you go? speaker-2: Yeah, I mean, so these are the hard questions. So you start this is where I think people i and and I think it's actually healthy to kind of go down this path so long as you stay grounded. But this is where you start to get down these paths of like eighty percent unemployment. Everybody's losing a job. The everything's gonna be free, like universal basic income for all, universal high basic income for all. Like all these concepts start, you know, coming out. This is another one. Like I I've brought this up many a time. Like, how do we how will we govern inside of a country where you have the problems that you just, you know, so eloquently laid out? And in a situation where the application of AI, if you look at it from like let's just look at it from a ⁓ nation state to nation state conversation, AI creates a situation where A small organization or a small country, rather, small nation state, can have a disproportionate impact now because of their application of AI, their application of asymmetric intelligence gathering and asymmetric warfare associated with AI. Now a country that is a fraction of the size can project a tremendous amount more power than they previously could. Right? So now you're talking about difficulties in governing. Not just from a defense side of things, but how do you govern ⁓ the defense side, also the economics, the economic side of what's going on inside your own country when you have a very high concentration of power centered in a you know a handful of companies that are effectively dictating what will unfold and how things will unfold, and in many cases at what pace it'll unfold. ⁓ but what you end up speaker-0: and speaker-2: seeing is what I think if you if you believe anything that you know that Elon Musk says Musk says, you end up in like the singularity conversation, which is you can't really see over the horizon of like what's next. Now you can iterate through, but I always try to and I I think there's truth to that. I try to ground myself in reality though and say like, okay, we're worried about 80% unemployment. speaker-0: Peace. speaker-2: Have we ever considered what forty percent unemployment looks like? speaker-0: It's like, wow. speaker-2: ⁓ like right. Okay. So if you know the bank forecloses on your home and they foreclose on everybody else's home, does the bank exist? ⁓ and it's like, yeah, we never make it to 80%. Okay, well, who's responsible for you know solving for that problem? Universal gaining high income? I don't know. Right? So my my grounding in that is is to really say, hey, you know what? I'm choosing to be aware of what's happening and how to leverage these things and how to apply them to be beneficial, ⁓ what the harm could be in in in using them ⁓ for nefarious purposes and trying to make sure that I'm living in a way where I'm prepared and not necessarily caught off guard, right? Yeah. So I'm embracing the change, but I'm also aware of like, you know, the ups, the downs. And whatnot associated with it. speaker-1: Yeah, more change more change is on the horizon. Yeah. I think that's you know, if there's one like really key takeaway, I know we're gonna transition into some into some takeaways in a moment here, but that that feels like it to me. I was reading an article that Sherm put out yesterday that talked about the disparity by company size with respect to AI strategy and AI deployments. And you know, no surprise, right? Like the ones who are speaker-2: More thing. speaker-1: Are doing it are the big companies and then the small segment of small companies who are tech startups. And in between, the they surveyed this population of you know smaller and mid-market companies to say what what's getting in the way. And the answer is we just don't know where to start. And so I think, you know, the the guidance that you just shared, Kirk, around like You you have to start. You have to sort of inventory what you know to be true and start to make, you know, kind of one step, one foot in front of the other steps towards transforming your business before it's obsolete. ⁓ that's where people start. And and I think it's ⁓ you know, it's a a great message of, you know, maybe instead of thinking a hundred steps ahead all the time and just being consumed by that, how do you how do you start to make measured baby steps in the right direction or at least in the direction that you feel is right, knowing what we know today. speaker-2: Agree. A hundred percent agree. Yeah. It's fun to it's fun to go down that rabbit hole and ⁓ you know theorize behind what or or what's on what's gonna happen, but again, that's where I I think that's where a lot of at the outset of the conversation, Amy, what you described, I think that's where a lot of that angst comes in though between people. I think this or I think that, I'm not touching this, I'm not touching that, I'm gonna do it this way. It's like okay. speaker-1: I mean, the the reality is though, whatever's gonna happen is gonna happen. There's no there's no stopping it. So speaker-2: I know. That's the that's that would be the worst mistake at this point, you know. And ⁓ I have people tell me frequently, speaking to the government, I have people tell frequently, Well, we'll just we'll regulate it, we'll stop it. And it's like, no, there's that's just the the government knows that's not possible, right? Like we're in a we're in a a literal arms race with near peer nation states. Like they're not stopping. There's no way. We can't afford to stop. Like our economy now is way too dependent on it now. Right? So like stopping is not an option. So people should really flush that from their vocabulary. You know? I I worry more for I you know if if you want to talk about concern there's gonna be these this bridge time for what is it Gen Z I guess you would say like there's gonna be a bridge time where it's like you were you're educated in a time where it was still very public school, you went to sch you went to school, you got the books, you learned, you know, reading, writing, arithmetic. Like I think that that is going to be so antiquated in the next 10 years that the kids that were learning what they learned in their first, you know, five, 10 years of school would be semi-obsolete. Now, not entirely, because I think that crystallized intelligence is is very important. And it makes you a better operator utilizing AI. A hundred percent believe that. But I think the delivery method for education is gonna be the hardest shift, but it's gonna be monumental, I think. speaker-1: I mean some would say it needs something because it's not like the US have a great reputation and the world for having a solid education for under eighteens and then college is very de determined on your financial situation and what type of education that you get. So it's kind of crazy. But that's not That's not true anymore necessarily. It's funny, I was just having this conversation ⁓ the other day, because I was with a client and we were taught looking at profiles on LinkedIn and ⁓ instantly they were like scrolling down to the education, scrolling down to the education. And I'm like, that is so irrelevant. Like with this person. What you want this person to do, like, didn't exist when they were in school. So what does it matter? speaker-2: Sure. Right. Yeah. Like you're hiring for aptitude. speaker-1: Father. Yes. Yes. And that's a big shift for some people. It's like forget, you know, how you used to assess a resume. Now you're gonna have to talk to the person and understand how curious they are, how they think through a problem, what their appetite is for learning. ⁓ and that You know, people haven't had to assess people like that before. It's been like, well, I've done this job for ten years in different companies, so of course they're going to be good here. And let's just, you know, make the higher. And plus I know a few people that went to that school and they're super smart. So I think we'll be fine. it's not relevant. It's totally different now. The people that you want are probably like hanging out, building speaker-2: Yeah. speaker-1: crazy things on their own with no education. Like they're just they're just cu they're just curious. speaker-2: possible yeah it's that I mean that you know that is the that that is one of the other like points of consternation is like what does the education even look like like what's the relevance of of you know going to college ⁓ yeah you know a lot of people are are starting to ask that a lot of parents I hear are even starting to ask that question you know it's like what are you gonna teach them that they can't otherwise learn themselves speaker-1: Right. But then there's a there's a real gap in communication, in just the basics of being a human on earth that people struggle the if the longer you spend in education, I think the harder you find life. speaker-2: Well you're taking a I mean it's like the the it's the name academia, right? Like taking a very academic approach to a practical issue. You know, you can you can sit there and hypothesize and teach to like how to tackle a problem, but then ask the person like, Okay, well how many times have you faced that problem? I've never actually faced the problem. But I thought about Okay. Thanks. speaker-1: Yeah, 'cause when you're in the hot sea it changes. Everything changes. speaker-2: Yeah, of course. You know, there's you you're there's another I I thought you were gonna go a slightly different direction. There is a ⁓ there's a shift there this is not like the catastrophizing s sense like side of things, but there is a shift that I see happening, and my wife actually sees it happening too in her line of work, in the mental health profession. There is a shift where That communication, so I I would ⁓ first just to put on the table, hundred percent agree, being able to communicate effectively is important, but its relevance to like business and what you do as a job, I think is gonna become potentially less important. I I'll qualify that. I know that'll that'll sound ⁓ you know, challenging to some. I don't think that's a universal like ⁓ like an absolute, by the way. I just think in many cases because the more that we leverage AI, the more it depersonalizes things. And so what you see now is, and I even find myself doing this to a degree. What you see now is is you'll see people, ⁓ they will have very personal relationships with the AI that they're using. Yeah. Right? And so it's ⁓ in speaker-1: Yeah. speaker-2: in in in some cases it gets unhealthy. ⁓ but you see mean you see the establishment of ⁓ like a real relationship with some of these things. And you can if you if you're using the tools long enough like you could recognize why. ⁓ but it's dang I think it's dangerous in the sense that if your your input into that mechanism has a a dramatic impact on what the output is. So you end up in this ⁓ speaker-0: Mm-hmm. speaker-1: Like an echo chamber you're after. speaker-2: Yes, yes. You know, so you have to be very careful about like what you're doing and how you treat that. But I can see already how it's depersonalized a lot of engagements. But I I think that, you know, if you're talking about the level of efficiency that I think can be brought to bear, that may be your one of your greatest touch points. I mean truly. speaker-1: Yeah. Crazy. ⁓ are we ready to do the wrap? speaker-2: Let's do it. speaker-1: Okay, what is one company type that is most at risk right now? speaker-2: One company type most at risk. ⁓ so the consulting firms I think are majorly at risk and they're they are definitely scrambling to ⁓ change their business models to find ways to integrate AI. ⁓ you hear a lot about SaaS companies and there's there's some back and forth. ⁓ I'll tell you that it's there is truth, a lot of truth to the SaaS companies being at at risk. Excuse me. ⁓ speaker-0: Mm-hmm. speaker-2: The you know, to give the the description earlier about what my son had done, you know, that in and of itself is an example of that, right? ⁓ if you like, there's some ⁓ there's a fun ⁓ website that's out there. It's called Death by Claude. You put the you can enter the ⁓ domain that you're interested of the company and it'll give you some rather hilarious ⁓ information about said company and whether or not it's at risk. ⁓ I mean there's it's definitely directionally correct. ⁓ You know, there's it's a it's a bit satirical, but ⁓ it's it's it's ⁓ it's accurate. So yeah, the SaaS companies that were providing you know, seat level licensing services, those companies are very, very, very ⁓ much at risk. And then I would say the other one ⁓ the other companies at risk are the ones that are highly complex and diffuse are gonna be at risk of ⁓ speaker-0: risk. speaker-2: market sh market share erosion. So I don't look at it as like, hey, you're gonna come and be unseated by, you know, somebody that designs a thing that displaces you entirely. I think it's more likely, hey, somebody's gonna displace ten percent of what you do. Are you ready to lose ten percent? speaker-1: Yeah, no. And that kind of leads me to the next question. What you think one economic shift is taking place that leaders are underestimating? speaker-2: An economic shift in terms of AI. speaker-0: Mm-hmm. speaker-2: I think that the I don't know if this will answer your question directly. The biggest underestimation, I think, at this point is the lack of financial controls within the deployment of these solutions and the impact that will have on them economically, on the companies economically over the next, we'll call it next year. Cause I think we're in very early innings of companies that have adopted Right. So you can have like your JP Morgan's, I forget how many nine figure sum of money that they invested into AI. And you know, there's AI for we built it into our products, and then there's that carries its own problems. Then there's AI that we're providing to employees, right? ⁓ but without the proper financial controls, this has the potential to be, you know, catastrophic for some organizations, but whereas For a company like JP Morgan that's printing money, probably less impactful, right? Maybe a rounding air for them. ⁓ but the the the unchecked rollouts and deployments of these solutions, and then you talk about in terms of economics, the what I think will be required in terms of reeling that back in. Look, there's you only have so many opex levers to pull. Like you can either make more money, you can reduce ops. Like Those are your those are kind of your choices, right? Mm-hmm. And then the downstream impact is like, okay, well, what does that do to these existing AI companies and so forth, right? speaker-1: Yeah, it makes sense. And then what is one move that you recommend people make to be more secure with their AI strategy? speaker-2: Personal or or company wide? I'm gonna I'm gonna take the the company approach here. Before you roll out AI in your organization, have a very defined plan as to what you want to accomplish. That observability pipeline is key. If you work down that theory, that methodology. speaker-1: Either. speaker-2: Your application of AI is not going to just be more secure. It's also going to be financially more responsible. The biggest flaw that I see time and time again, and you see, you hear about this all over Twitter, you know, this company had this agent do this thing. The biggest flaw I see is everybody's very quick to roll these things out, absent any of that observability piece. They don't even understand what that agent, right? This virtual human. like I described earlier, is doing inside of their own environment. That's the biggest issue that I see. speaker-1: Yeah, makes sense. ⁓ last question, totally off topic. How long were you in the military? speaker-2: ⁓ almost fourteen years. Yeah. speaker-1: How was the training? Were you pretty fit before you went in? speaker-2: Is this is is this gonna be ⁓ recorded or are we gonna we're gonna include this? speaker-1: Was it was it s was it pretty strenuous? speaker-2: Of course. Yes, as you would expect. speaker-1: Walls and no no footholds. speaker-2: ⁓ I mean I've definitely I've definitely done that, yes. speaker-1: She means through the t through the nets. speaker-2: Of course, yeah. before so it was something I was always interested in. speaker-1: Because you're patriotic or because you just thought it was an interesting speaker-2: I think a little of both. Like I I I just had a a ⁓ a level of respect for people in the military. I was fascinated with military things. You know, I don't really know, like I I can't point to like one thing and say, hey, you know, that was it, right? That was the moment. Yeah. It wasn't it wasn't really like that. ⁓ it but you know, there's a lot of misnomers too. There's a lot of things that people they have these perspectives, and like I don't fault anybody because in in the US especially, the US is so big and broad, that it's extremely uncommon to live near a military installation. Like unless you live near a military installation, it's you know, you can ask speaker-1: in England you can't like turn a corner without going into a barracks. speaker-2: Right, exactly. So it's it's very different, right? So you you see and you're kind of familiar with these things, you know. So people will ask like really funny things like, Hey, did you live in a tent? I'm like, Well not usually. I mean house, family, you know, the dinner table, you know. But like I understand why they ask these questions 'cause like they don't know, right? And it's like, no, I lived a you know, fairly n people will be like speaker-0: you speaker-2: Are you crazy? I lived a fairly normal life, you know, outside of what my job was, you know. Yeah. My day to day life was, you know, I went to work, did my job, I came home. Like unless I was training or deployed. Like that's that was my life. ⁓ but yeah, I mean the the training is ⁓ like anything else in the military, I tell people, it's what you make of it. And so, you know, I chose a really difficult path. ⁓ speaker-1: That's right. speaker-2: A lot of people before me have made it, a lot of people after me have made it. It's not impossible. ⁓ I don't let that define me, who I am. You know, that was an identity at a point in time. ⁓ something that I'm very, very grateful for. the experience is really unmatched. You know, I I think sometimes to myself, like, you know, I'd go back today if I could. ⁓ you know, I definitely miss that. that side of things. ⁓ but yeah, all in all, yes, strenuous, difficult. Everything that you can imagine, every emotion that you can experience, ⁓ definitely felt for sure. speaker-1: Well, appreciate that you did that for your country because if people didn't, I don't know where we would be. speaker-2: I I appreciate the thoughts and the sentiment. ⁓ you know, I I that's a I hear that a lot from people where people will say things like, Hey, I'm not gonna, you know, not my child, not my kid. And you know, I say, Well, if if if not then, then who? Like who's gonna do it? Like, you know, that's that's that's the question I think people need to ask themselves. It's not for And by no means am I saying it's not for everybody either, right? There ⁓ but there's this'll sound like I'm talking out both sides of my mouth. It's not for everybody. But I don't I don't really know anybody that is like purpose built to go do the what I did. Like you have a desire and a passion, but nobody's you're trained to that level. speaker-1: Relentless work that goes along with being in the military. Yeah. speaker-2: Yes, like there's no like one size fits all. No. And you'll hear a lot of stories from folks in special operations community and you know, talk about the selection process. I was an instructor for a handful of years. You'll talk about the selection process and you'll look at people, and it would be really easy to look at guys that are just look like physical specimens. And there are a lot of guys that end up you know, becoming that, but you would look at those guys in the assessment selection process and say, ⁓ yeah, that'd be the person that's gonna make it. Almost never. Almost never. And what you really realize is is that the whole thing where there's a ton of physical strenuous tests, the whole thing is really a mental evaluation. speaker-1: Yeah. How strong are you mentally to go through something and keep going, even when it feels horrific. speaker-2: Well, and how you're navigating like some very complex challenges that are put in front of you. And y you're constantly being evaluated psychologically, constantly. speaker-1: ⁓ I wouldn't last five minutes. I think you might. No way. Number one, I don't crawl. Number two, I haven't had a good bill of mental health since nineteen eighty six. The physical part would be hard for me. I'm too old at this point, I think. A lot of gym days to prepare for. Not even on my best day could I do that. speaker-2: Well, it's like my ⁓ one of my son's teammates, the father, he he asked me, last year, last summer, when th the tensions with Iran had first kicked off. He's like, Hey, could you go back and you know, get be on a team right now? Like, are you crazy? Have you lost your mind? I'm like, they would leave me for dead. I would be a liability to them immediately. speaker-0: you speaker-2: You know, he was shocked. He's like, But you look really fit. I'm like, I just it's not it's different level that you just don't you can't really appreciate unless you've been there. Yeah. Yeah. And it's you know, and I have a tremendous amount of respect and like that's the standard that I expect that they continue to uphold in in those organizations. But it's yeah, there's just walking, you know, yeah. You're not doing that in your mid forties, it's not happening. speaker-1: That's why you gotta promote. If you're gonna stay in, you gotta promote. That behind a desk, not out there doing the crazy shenanigans. speaker-2: ⁓ yeah. Well, that was I mean, ultimately that's why I decided to to retire when I did. so I got I you know, I was injured pretty bad in two thousand thirteen and know, I spent a number of ⁓ months in and out of the hospital and you know, it was really like, Okay, well, you're not gonna be operational anymore. And for me it was like I've been promoted already to there was only really one more level I could go up and I didn't wanna sit behind a desk the rest of my career. it just it felt like ⁓ it didn't feel like I was doing what I was supposed to do, you know, so it was time for me to kinda like pack it up and and make a journey into the civilian sector, so speaker-1: Well, I'm glad that you did. Then I got to meet you. My world since twenty seventeen. It has, yeah. It has. Well, good to chat with you, Kirk. Thanks so much for your time. speaker-2: Standard. Same here. Thank you. Take care.