The Cost Management Center Podcast

How AI Will Transform — and Shrink — Your Technology Stack with David Habib, Author of Latent Vector

• Duane Deason • Episode 7

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 44:30

What if the cloud applications your company depends on today become obsolete? In this episode, I sit down with David Habib, a Chief Information Officer and author of a book on artificial intelligence, who has spent considerable time thinking about where AI is taking us — and the answer has major implications for your technology budget.

David's central thesis is as provocative as it is compelling. AI has the potential to reshape a company's entire cost structure. We focused particularly on his premise that AI will eventually eliminate the need for many of today's most popular cloud applications — not just the licensing costs of those platforms, but also the expensive, time-consuming work of integrating them. Beyond the obvious cost savings, companies will be less constrained by applications that never seem to offer exactly what they need.

For cost-focused leaders, this warrants close attention. Listeners will walk away with a forward-looking perspective on how AI could reshape technology costs and unlock capabilities that current platforms simply don't allow.


Connect with David Online:

LinkedIn: https://www.linkedin.com/in/davidphabib/

Website: https://davidhabib.com/

Latent Vector book: https://www.amazon.com/Latent-Vector-David-Habib/dp/B0FPTFD861

Find more from Duane:

Email: ddeason@efficacygroup.com

Website: https://efficacygroup.com/

LinkedIn: https://www.linkedin.com/in/duanedeason/

Buy the Operationally Svelte book: https://efficacygroup.com/company-efficiency-book/

SPEAKER_01

My name is Dwayne Deeson, and I am the host of the Cost Management Center podcast. I am also the author of Operationally Svelte: Manage Costs to Increase Profit and Enhance Performance. In this podcast, I interview experts and leaders in order to give my viewers and listeners the opportunity to learn about how to manage costs in their companies and organizations. Welcome back to the Cost Management Center podcast. I am Dwayne Deeson, your host. I have a great guest with me, David Habib. David and I actually worked together at a company many, many years ago. David is also an author and is a specialist in AI, and that's what we're going to focus on today is that how AI is going to impact the cost structure and cost platforms of companies. So welcome, David. Thank you, Dwayne. It's great to be here. David, I always like to start off with a little bit about the background. Tell the audience a little bit about yourself and particularly how your focus became more on AI because you've written a book on the subject, and we're going to get into that in a little bit, but I want to, I think that's an important aspect for the audience to hear about.

SPEAKER_00

Yeah, for sure. I I my my career-wise, I got into the internet uh very, very early on. I worked at an internet service provider in the call center. Um so I kind of fell into technology by accident. But I've been working in on the internet side of technology since the late 90s. And uh I've been focused on like trying to play in every dimension of technology. So software development, quality assurance, uh, you know, what became cloud operations that we used to call SysOps, uh, back office IT, uh, InfoSec. So I'm sort of voracious for all different aspects of it. And that's helped me sort of get to the executive level. So you know that that kind of that kind of breadth and that kind of understanding. AI is sort of the logical next step in that, in that conversation, right? It's an emerging uh an emerging thing and a disruptive thing. And because of the sort of time compression that that folks like us have gone through, I was able to, or I thought I was able to start seeing patterns that uh you know we saw at the early stages of the internet repeating themselves in the early stages of AI. And I don't mean the technology, I mean the what's going on in the marketplace, what's going on in in businesses, what's going on uh in the industry. With Bright Spot, I had the opportunity to sort of be at the forefront of of helping shape the adoption of AI organizationally, right? So as as CIO, like what are the policies, what is what's the governance? How are we going to educate people? How are we gonna facilitate them using it? What are the risks, right? Security-wise, business-wise, what are we building into our product? And so um I got to sort of help watch and help steer that. Uh and and the nice thing about where AI is now is that you don't have to be an expert to get your feet wet, right? So I also had the opportunity to like really play with it because I needed to teach people, other people how to use it. I needed to teach all of us how to how to govern it. So um AI was sort of like a logical next step. Uh, and I just I felt like I was seeing a movie that no one else was seeing when we're talking about like what the implications are, right? And I I know we'll get into that, but the book was really uh a like I wanted people to like see what I was seeing because it's all it's you know, if if if you see it, you can prepare for it. And if you don't see it, like you you have a store in the mall and and the you know, e-commerce is gonna come eat you and you're not gonna know, right? So right.

SPEAKER_01

So that's very helpful, and I I I appreciate that background. In fact, um I did want to start with your book just because I actually read it. So and I have it here for you. Um so so it is uh it's called Lane Vector, and I thought it was uh it was actually eye-opening for me because uh I've I've heard about AI and we've I've worked with it in some and I certainly understand how it's going to impact people's jobs and make them easier, but I think your book opened my eyes to the to the next level, which was uh much more disruptive, uh whereas opposed to slowly um companies slowly replacing employees with headcount was is really what it was uh sold as. And now I think people are uh talking about much not only just headcount, it's gonna change everything. And your book was the first to kind of open my eyes to that. And what's interesting about it is you uh um just two or three months ago when I realized that it's almost like the market just started making that acknowledgement four to six weeks ago, where they're starting to realize that all software is at risk of being replaced by AI. And so I felt like, oh, I knew that. And I really knew that because I had read your book.

SPEAKER_00

It's very gratifying when people send me articles and they're like, this is what you were talking about. Exactly, exactly. So hope hopefully you get some sort of recognition, but probably nothing.

SPEAKER_01

It's unfortunately not the way the world. I get this thing I touched. But so let's let's um uh go ahead and frame this discussion in a what in what your book um presented. And uh so I'll let you kind of explain that because I probably won't do it justice. Um but what uh the premise of your book goes as I was explaining, goes well beyond the idea of slowly replacing employees. What uh go go can you go into that just a little bit?

SPEAKER_00

Absolutely. The premise of the book is based on like one notion, which is that the way AI stores and interacts with data is is the disruptive part, right? Like the the interactive technology, like the Chat GPT is is fascinating. But the really disruptive part is actually what's going on behind the scenes. And so uh I spend a fair amount of time in the book talking about the history of information management, like going all the way back to clay tablets. But um, if you take a look at information management in the enterprise, it is a self-reinforcing sort of silo, series of silos. Uh uh uh and that's just the way it's always been. So the the classic example is you're a B2B company, you've got 2,000 customers, you're doing fine, um, but you're you've got information in your AR system, you've got information in your CRM, you've got information in your like support system, information in, you know, all these different systems that are have information about your customer, right? Um, and those different systems are you have to upgrade them every year, and that when you upgrade them, you're actually just sort of reinforcing the silos. And so the problem for companies has always been like, but my answer to my question is actually a superset of each system, right? HR has a system, HR system is meets their needs, they hire more people who know that system and it's self-reinforcing. Um, but the the information, the decision important information is actually scattered. And so the the story I tell is like if you have customers, you know, they're opening support tickets, that's in one system, they're talking to their uh account manager, and Google's capturing the the conversation, that's in a different system. The like how often are they paying their bills? When's their renewal? What's happening to their organization? Did they hire a new CEO? That's scattered. And there's actually no way for that information to come together. And so you layer on top of that a very large consulting industry whose job is to help you with that problem, right? So, and like a lot of of expenses that businesses will see uh, you know, like as your IT investment increases, your spend on consultancies increases as well. And that is what I call in the book the information complex, like the information commerce complex, where it's a self-reinforcing thing. And then the the regulations come in over the top. So the point is that AI is not from that perspective just a step function, like the latest technology. It is, it's a it is a sea change because when what AI is good at is a diversity of data arranged in a non-deterministic way that's available for interaction, right? And so, like factor databases, and I don't I'm not an expert in the technology, but if you take that point, so we're no longer doing information silos that are glorified file cabinets. We now just have what I think we tried to do 10 years ago with data lakes, as you may remember, right? But a data lake was just a pantry, like everything was still in its own bottles, it was just in one point. If you talk about how an AI deals with information, it's it is agnostic in like a very real way. And so the data stops having a deterministic nature to it. And so what that means is that if you are in the business of maintaining one of those silos or selling one of those silos or integrating one of those silos, if the silos no longer necessary, you're in a situation where the bottom like falls out of the market, right? It's not a step change, not a we have to go train consultants to know this new technology because everybody knows how to do that. It's entire like multi-trillion dollar industry is under threat, and so it's gonna go into defensive mode. And we're already seeing, right? We saw the we saw IBM stock drop because somebody made a statement that you know AI can do COBOL. I think we all intuitively knew that that was gonna be true, right? But when it came out, like the stock drops and people are gritting their teeth, and we're gonna see that story more and more. And so that that's the premise of the book, is that there is a very, very big part of uh what companies depend on from an IT perspective and what our market, like what our economy depends on, that is the only reason it's not going to be taken over by by AI is because people are pretending that that's not the inevitable outcome. Like that's that's the premise. That wasn't very brief, but that's the premise.

SPEAKER_01

No, it it and it's it's very helpful. And and I'm gonna try to put it into a little bit non-techie language because I'm a non-techie and I want people to make sure they probably have listeners that that might fall in my category. And so essentially it goes beyond, and because you talked about a you know a company with 2,000 customers, it goes beyond the idea that okay, do we need a Salesforce for CRM type of support when everything that you do can go goes into AI, and I can now uh rather than be limited by by what Salesforce can provide and its functionality, I basically can ask AI to give me anything in any format, and it and it there's there's so much less friction in order to be able to get that information. And then even beyond that, it's that now it can go well beyond what what uh uh Salesforce really can do right now between other applications and feeding that so it can give me much better detail, much more. So not only is it a replacement for Salesforce, it is a superior offering. So I it's not that I can just save my because Salesforce is an extremely expensive product. Uh not only can I save that money, but I will actually have a better result uh because I chose to save that money.

SPEAKER_00

That's that not quite yet, but that's where I think we're right. And so like the there I have to because uh you know I can I can hear my colleagues snickering, like we're not there yet, right? The technology is still, you know, it could still hallucinate and it's still uh you're a victim of whatever the data is that goes in. But conceptually, you don't need to store your customer data in a CRM and your support data and a support system and your finance data and a finance system. Like conceptually, you don't need to do that anymore. What you do need is the appropriate framework, right? Because like, and you know this, like finance data isn't the spreadsheet, it's the reason that the data is there, right? Like what do you need it from a compliance perspective, from a forecasting perspective, from a risk management perspective? And so the frameworks need to be overlaid on the data when somebody has to know that, right? So your a your your AI mainframe is gonna have to have a framework for uh HR because that's a highly regulated part of the business. You don't want to like just hope that works. So I think what when we think about like what is the next generation of enterprise software, I think it is very much a mainframe, right, where all of your stuff is. And then people like Salesforce.com or ADP or or Workday, they're selling a framework, right? So they're the experts in the best way to manage HR data and what the relevant laws are in Tennessee, right? And so you can you imagine subscribing for that. So if you're trying to plan for how much is that thing's gonna cost, which I think is what your audience is is is thinking about, like it is necessarily dramatically less. Yeah, right, because providing a framework and providing expertise uh is cheaper. Yeah, right? It's cheaper. Uh and so their costs drop, our costs should drop, they should require fewer people, and the people who are in the business of doing the integration between silo A and silo B are gonna run out of, we're gonna run out of runway, right? Right. So from a cost perspective, it's very, very favorable. Right. The the that trend should be very, very favorable. Right. Um the question is how much disruption stands between us and that point, right? As you know, and I talk in the book about the resistance machine that's already spinning up, right? The like all naturally businesses, individuals, governments are gonna defend themselves against this sort of thing. And that's gonna drag out the trend, you know, like how long it takes for us to get there for several years, with the exception of startups. Right. And this is where I think the most interesting stuff is happening right now. I would uh uh I'm talking to a handful of AI start-based startups, and like when you're starting from zero, you can start with nothing. You can start with I have a database for all of my static files, and I have an AI that's doing cool stuff with my static files. And when I need to add new information, I just add new information. And AI have a new product, and you just add it, and it's just there versus oh, you have a new product, I need you to go enter it into this system, and then we have to add SKU, and then we have to write like it just becomes materially easier, and so the resistance machine is gonna be trying to hold the line, and the startups are gonna have that much more advantage uh coming out of the gate. So it may be for the next four or five years the purchasing decision is do you want to go with the incumbent and where they are in the you know, the the Gartner Magic Quadrant and the way of doing things, or do you want to just be like, you know what? I go if I go with the upstart, it'll take them a week to fix my buck. It'll wake take them a week to add this feature, and it'll be cheaper all the way along the line. So I think that's and I'm jumping ahead probably, Dwayne, but I think that's the way that we need to start thinking about it is not uh should it get cheaper, but should we keep feeding the old model? Right. Or should we, you know, because we got our shareholders money to deal with, like, should we be is it a gamble anymore? And I don't I personally don't think it's a gamble anymore because it's the the the iterative time is so fast.

SPEAKER_01

That's great. And that actually does lean well into what I was wanted to focus on a little bit at at this point is what is what is today and what is tomorrow, as far as um from that from this aspect of it. So if you were starting with a walking into a company, and it sounds like it would depend on the size of the company where they were, but but you know, they're a lot let's let's start with a larger company and they you walked in that door and they have all the usual players of you know everything from work day to Salesforce to uh they they used AWS and all those type of things. Where would you start with them? And they're let's say that they're in the infancy of AI. You know, I've I've heard a lot that almost everyone starts with co-pilot. What would you start trying to get them to focus on to be able to whether and and particularly to reduce costs, um uh their carrying costs of these things? What would where would be your focus?

SPEAKER_00

So I think that it's twofold. And I think everybody's doing the first one, which is like, let's figure out how AI fits with how our organization works. Okay. How do we make Claude make the folks in finance faster? Okay, that's useful, right? Like that's you're gonna have an incremental, you know, incremental value per hour for people. Like it's useful. Yeah, but that's not fundamental. I think what I would advise people to do is to take a step back and take a look at what we used to call their enterprise architecture, like your your full suite of IT investment and correlate it to the business, right? Like, because I think and you live with this probably more than I have, but I think you get to a certain point where it's just habit to have a CRM, right? Sorry. Um everyone in the comfortable with it. You're paying a lot of money for Salesforce, and all it is is it's a glorified database of contacts and subscriptions for marketing, right? Like but everybody in your organization is gonna have like they're gonna be religious about those things. That's how organizations work, right? This is where we store our tickets and we need our tickets because if we don't have our tickets, then we can't do our ticket things. And so I think the value is to come in and say, like, okay, but what's really what's really important to the business in this system, right? If I lost 90% of the historical data right now for my ticketing system or for my support system, would it matter? Yeah, right. And that's a hard question. And the answer is, by the way, almost always no, because like support systems like your time horizon's this long, and everybody wants to have the history. But if you're paying so that you have the history and the history is already in a sort of deterministic silo, then you can be like, you know what? I'm replacing that. Like that's a target for replacement. Not I'm gonna go back and negotiate with uh Atlassian. But I this is a target for no, I'm not doing that anymore, right? Like we're gonna reinvent how we think about this. We're gonna build a system that is exactly what we need the system to be. And either we have people on board who can help us build that system or we're gonna find somebody to build that system. But I think in that sort of situation, you'll find that the tool uh in year two is gonna be like literally 10% of the expense. Not half the expense, but like 10% of the expense. And it also is gonna free that business to be a lot more innovative. Like a lot more innovative because right now we're all tethered to the systems, right? Can you throw up a support portal for me like that shows uptime? No. Because my support system doesn't know about uptime. You'll be able to be like, sure, I can totally do that. Just give me the uptime data and then I'll just so to go back to actually answering your question. I think I would look at like the big cost pillars in your IT investment. Like, this is an IT solution, AI is an IT solution. Where's your IT investment? Like, you'll know in your gut, like, I'm overpaying for the value in this thing and I'm addicted to it. Because nobody wants to migrate off of Salesforce because it's a nightmare and it's gonna take 12 months. And so you're you become addicted to these things. Where are the places where it's disproportionate what I'm paying to what I'm getting? And like think about like if I could make a spreadsheet that contained all the data that I need that's in Salesforce.com, then why am I paying for Salesforce.com? Right. Right. And I think because you can do it another way. That's valuable.

SPEAKER_01

And I think it's good that we're talking about CRMs because one of the more sensitive areas in companies is always the of course the revenue and their sales. And I'm sure there's a uh progressive um chief revenue officer, but I'm not sure I've met them. So they because they they they're they really want to focus, they're they're 100% focused, their targets and their their incentives are all focused on um let's get let's we have to continue sales and revenue. Um and so you you start telling them you're gonna muck with their core system and and they're gonna have a heart attack. And I guess um what I had thought that might um, because you got me thinking and what what might start to evolve is that you can't go to them and say, look, I think we can do without Salesforce. I think they they they panic, they got too many areas other to focus on. But what you could do is for a period of time, if it's cost effective to do this, is prove it to them. You where you run parallel, where you show them that, okay, look, we're going to record all your information and everything you're doing, or that type of thing. I'm going to show you the power of it, and that you and therefore you alleviate their fear and their concern because you're able to prove it. And yes, there's a cost almost incremental for a very short period of time while you prove that out to them. And then that's when you they start realizing that that their their reliance on traditional applications is is not necessar actually not founded anymore into the new under the new world.

SPEAKER_00

I think that's exactly right. I think if you're in a situation where like people are gonna be like, yes, that's adorable, but we don't have time for that, right? It is the bar is actually very low to implement a new system that is a better decision support system. Right. Because like ultimately, what are you paying for? Are you paying for you know decision support or you're paying for operational support? And so if you you could say in a hypothetical universe, like I'm gonna pull in with the data that's in Salesforce, Salesforce is a perfectly good API. I'm gonna pull that data in. I'm also gonna pull data in from these four or five other systems, and then I'm just gonna start I'm gonna start writing my status reports from this system instead of from the five other systems. I'm gonna check my work, I'm gonna make sure that I'm not blowing smoke. Like and start and introduce another voice into the conversation. Um, the the barrier to entry on that is very low. So, like, like anybody in your audience could do this because you just have to tell you just have to have a series of keys and you'd have to tell Claude what those keys are and you say go get the data, right? Right. Um, so I think your point of like, can we just introduce another voice into the conversation? Absolutely. And when you realize that you're answering 95% of the questions become based on five fields that you're pulling out of Salesforce, and then you can present back to the salespeople and the business development people more information than Salesforce ever gave them, then it becomes an obvious right, it becomes a layup to say, right, current, well, I I guess I don't read this anymore.

SPEAKER_01

Right. You've you've proved it to them, so to speak. They become more empowered and more effective because they are they are take using this new the new uh tool and new source of uh of solutions rather than the traditional one. Okay, that that makes a lot of sense there. A little bit more work, if you think, because uh essentially that means that you have to uh run parallel. Uh you you have to can of course keep Salesforce up and and uh and that cost up for a period of time while you are uh at um get of the the new solution up and running. So a little bit of overlap in time and cost and and heavier weight on uh kind of a department like yours, but I think the end result uh and once those things get momentum, people start realizing it. Uh yeah.

SPEAKER_00

So you make the investment in that center of excellence, and it doesn't have to buy substantial investment, and you should see the returns within, I would say within two years. You should absolutely see the returns.

SPEAKER_01

Yeah. Yeah.

SPEAKER_00

Um it depends on it depends on how voracious the the leadership is. Like if they like want to blow it up because like they would rather reduce exp you know IT expenses by 50% than staff expenses by 20%. Like, okay, we're gonna make some sacrifices. But if it's a we want to do it the right way, but we also don't want to like keep feeding the dragon, yeah. You just intro introduce it into the conversation. I think that's right.

SPEAKER_01

Well, initially it was focusing on AI displacing people. And um and I think the the we've already seen some of that, although I think a good a fair portion of of the layoffs of now are people had kind of overstaffed their departments a little bit and and and now you they're using AI is certainly helping a little bit, but it's only part of the reason that that they're reducing. I think that will actually accelerate because AI will get better and better doing the um they're doing operations and and things that people can do. Um then we've talked a little bit about how applications are are going to be not necessarily replaced, but certainly diminished or or their core costs will reduce drastically because they're um providing a structure and uh and information rather than just rather than a true application anymore. Are there other areas that we can't think of right now or that we should consider, I should say, um that AI will assist with? I mean, um uh innovation, advertising, other things that were that are direct areas that other than just technology and people?

SPEAKER_00

Aaron Powell I think so. I think and I'm not this is not an area I I I've spent as much time on as the as the enterprise side. But I think there are dimensions of our business that um are feel a little bit gambly, right? You said advertising. Advertising is a great example of that. Aaron Ross Powell The click-through rates and the SEO and the all the things that that those those teams work on. Like it feels like voodoo, right? It feels like magic, and sometimes it works great, and sometimes it doesn't. And when it doesn't, nobody can explain why. Uh supply chain works that way. Information security has a has a bit of uh like, well, I don't know, so like let's just spend more to protect us against the I don't. So I think these areas that are that are amorphous, um, that are unpredictable and tend to be expensive because they're unpredictable. I think we're gonna see AI bring some clarity into those spaces, right? I already mean we're already seeing it on information security because information security is a game of like there's a lot of variables and they're all changing a lot. So it's very difficult for a human, it's very difficult for a like a deterministic program. It's not actually difficult at all for an AI. Right? Like that's actually that's its bread and butter. Right. Um, the behavior of advertising, the behavior of SEO, the inf the influence of social media on traditional website media, on you know, more traditional magazine media, again, less less and less of those things are going to be c are are gonna be mysterious. And so I think we'll see more control on the less controlled costs, right? That that the businesses naturally have. This is the part that I don't think is getting enough attention. Uh what we've seen time and time again is when there's a there's a technology that disrupts an industry or a marketplace and we lose a lot of jobs because of it, every single time we've created more because of that new thing. And so I think the thing, I think the thing we can't see right now is like so we can see the bad news right now. We can see that, you know, the the demand for people who are sort of organizational connective tissue that are doing uh you know, doing doing that kind of thought work, knowledge work, like those roles are under threat. There's nothing nothing that you can say about that. But what we're not able to see is like what are the new roles that are gonna appear? What are the new jobs, new industries, new businesses that are gonna appear because the landscape is is evolving and is is changing. Yeah. And some of that I think is going to be the next generation of software vendors, right? I think uh and some of it's gonna be the next generation of lawyers, right? And and that sort of thing. So I think there's we need to prepare for there being a whole new, like a whole new set of of costs uh in in the ledger that we just can't predict yet. So we kind of want to you know clear room for that for sure. Okay. Um but I I do think in the short term the some of the variable costs will become more predictable than than they are now.

SPEAKER_01

Fair fair enough. And and as far as people getting replaced and what departments and functions are gonna are kind of it seems like it started off with a lot of developers uh were originally uh targeted, or at least that's what the applications could were most effective at doing their work. Um uh and it's interesting. I had a good friend who was a uh uh is a software developer for a major company. Uh he said it what they went from getting bagels every day and you know uh anything they could do to try to keep those happy to make sure they stay with the company, to he said it's like it became like hunger games. So the colleagues were were getting let go, and then you know it was all became about what can you produce, how quickly can you uh you know adjust and um to AI. So it it's it's been a very tough on that group. Yes. And uh and then there's and then call centers, but it seems like everything else is a little bit more spread out. There has been talk about um impact uh heavy on the legal field because the for particularly contracts and a lot of that type of work can be um done uh perhaps even better by AI. It can even be negotiated with another counter AI and then and then just approved by attorneys and that type of thing. What what what else do you think is it gonna hit everything other fields pretty evenly after that, or are there other segments that are gonna be um and functions that are gonna be uh much more uh more rapidly um uh converted versus versus later?

SPEAKER_00

I actually think it's gonna be horizontal. I don't think it's gonna be uh focused by industry or anything like that. I think it's gonna be horizontal. And this is you're getting a free preview into the book that I'm working on now. I didn't realize there was a volume two. Well, no, it's uh it's it's uh it's it's related. But take a look at the amount of judgment that is necessary for your role. Right. Right? So if your role is primarily to be judgment into the conversation, s negotiations, sales, talking to other humans, uh strategy, uh you know, making decisions that'll impact the company, those positions are less likely to be uh sort of eaten up by AI. Okay. They're more likely to be accelerated by AI, right? Like because AI is going to make more information available to you in an easier way. So the you know, sort of decision maker roles, judgment-based roles, AI can't replace that. Right. It can make you worse at it or better at it, but it can't, it can't replace that.

unknown

Right.

SPEAKER_00

If your role is very low on the judgment scale, I take the requirements and I bring them to the developers, right? Which is the joke from a movie. But if if you are what I said earlier, connective tissue in the organization, right? You pull the reports and do the analysis and give it to the manager so that she can do her status report. Yeah. That role has nothing to do with me in that role, but that role is completely replaceable. Yeah. And so I think we're gonna start seeing in our very knowledge-based economy roles like that just disappearing. Right, right, the conversation about developers right now is is blown out of proportion. You're absolutely right. There's other things, a lot going on uh besides AI, and also like we're all still trying to figure out what to do about AI. But I think once the once the dust settles and we're able to see the patterns, I think we're gonna start seeing that connective tissue, which is in every department and every organization, right? Because right now it's necessary because of the silos that we talked about. When the silos come down, those translators are not as necessary. So I think you look at the role, you look at where they are and the sort of how much judgment do you need and expertise do you need to do your job. And if it's high, there you're probably okay, right? Like from a role perspective, and you should plan to staff that in the future and and that sort of thing. But if it's purely connective, it's an it's you're not gonna be able to explain to your shareholders why you're still paying a person to do that. Right.

SPEAKER_01

You can't justify that. Yeah, yeah. No, I I think I think that makes a lot a lot of sense. Now, when you ask for additional funds for for resources for AI, I I'm hearing most uh you know, CIO types, CTO types are getting uh told that basically they want they their expectation is almost like immediate payback. And is that is that a fair request and are you are you able to accommodate that um or do or should is that a longer horizon to really feel the benefits of AI being implemented? When you talk about two, you mentioned two years, if you invest a lot, two years it'll make a huge difference. But I don't see a lot of management teams and and and uh shareholders being at that patient.

SPEAKER_00

No, this is uh and we did we saw this with uh with big data as well when that was where the the boards were expecting we're gonna make this investment, but I want to see like new product. Yeah, right? Because I read this one article because there were new products. So I think we all know there's no such thing as a free lunch, right? We're not there's not a I'm gonna give you a hundred bucks to invest in AI and I'm gonna see $300 in return tomorrow. Um I think the conversation isn't about purchase AI, see return. The conversation is like, what are we trying to accomplish? Right? What are we what are we going for? Are we going for resolving like reducing the number of defects in our software? Are we going for reducing the amount of time that our customers are spending on the phone with our agents? Are we going for and once you know that that becomes the conversation that we all know how to do with because we're business people, right? Like, can you use AI to reduce the number of calls? Probably. How long will that take? A couple months. Now you've got a conversation. Yeah, right. Right. Now you've got that conversation all day long. Right. And you reduce that cost by 10%. That's gonna be substantial for some companies. And so 10%'s material, and you're gonna make an investment to bring that down. If you're a tech executive and they come to you and they're like, Well, I want you to like build AI into everything, and then I expect the cost to be down by 20%. You've stepped off the lot. Like you're not having a business conversation anymore, you're having like a fantasy conversation. Or I think that's we just need to bring those conversations back to what we know how to do, which is like, what are you what's your expected outcome? What's your objective? And then sometimes AI is going to be the right answer, sometimes it's not.

SPEAKER_01

And is it and you you mentioned a couple of uh well, we talked about a couple of tools that came up, but not not we didn't drill down on it. You mentioned Claude, I mentioned Copilot. You know, are there certain um AI players that are doing it that you're impressed by um that you that you want to call out, and others that are more that more hype that you that you'd feel a little less strong about?

SPEAKER_00

I so I'm I'm a huge fan of Claude. Huge fan. Uh in all the different dimensions. I use it as a as a uh for code. I use it to you know uh do research. I do it to like find how long I'm supposed to cook brown rice, and in all of those dimensions, conversationally, technically, like very, very happy with it. The price point is reasonable for what it is. Okay. I haven't spent as much time on ChatGBT, it's fine. I think that you know there are different AIs that are better at different things. So like the Google Gemini is uh I think great from a consumer point of view. Like if I want you want to figure out how to make brown rice or you want to figure out like because it's Google and so it knows a lot of things, but it's I would I uh nobody uses it to write code yet. If somebody starting from scratch, I would absolutely go with Claude. I would absolutely go with Claude. Okay. Um because it's it's very, very accessible and and I have not I sort of have not found the limits of what it's able to do. And I've been doing some pretty crazy stuff with it. So Okay.

SPEAKER_01

Yeah, and it and and it's it seems like it has the the uh a stronger depth to where it can help uh and anything that you need it for, whether it's b uh business applications, productivity tools, and stuff like that, it has a lot of different dimensions that it does well, as opposed to other ones. I think probably have a few dimensions that may do better, but or it's equally, but it's evolving so fast.

SPEAKER_00

I will say, because I feel obligated to say this, the open claw thing I would advise people not to do. Like the download it and run it on your own machine thing. Like there's from a security perspective, like if you don't know what you're doing, you're introducing a substantial amount of risk into like interesting whatever. Um, so I wouldn't do that. I wouldn't run your own unless you really know what you're doing. But it's it's pretty easy to experiment with all of them and and see which one you're comfortable with. And I think for most people, it's just gonna come down to the the the UI and the voice. Because if all you're doing is asking you how to make rice, uh you're all gonna have the right they're all but gonna be good good pretty good information, right?

SPEAKER_01

Right, right, right. Okay. Uh fair enough. And then um as far as like tips to leadership in there, and other than you mentioned to have realistic expectations and and uh and make sure that you're that the the answer to every uh problem is in AI. You know, there's it depends on on the problem and and how you face that. What other aspects would you like to see uh from leadership? Are they are are they doing well on this or are they too caught up in the hype? Are they um part of a problem or or uh the solution?

SPEAKER_00

AI for a lot of people is a solution in search of a problem, right? Um, and so I think for senior leadership, you have to watch yourself and watch your team and make sure that's not what's happening. Because I think it's happening a lot. And so I have this thing and I I need to find a use for it, and then I need to come up with a KPI for it, and then I have to measure that, and then I have to justify it, right? Which is, as you know, like the wrong direction, right? So I think be aware when this is a, you know, I I'm doing it because I think you told me to problem. And the other thing, which is harder, is we need to start being aware of when we as an organization are showing those signs of resistance, showing those signs of uh uh addiction, right? Because those things that now when you talk to your head of sales and and and he's like, uh you know if I don't have Salesforce, I'm not gonna beat my numbers, right? That's what he's you know, and and come on, right? Like that's probably not true. It maybe it is true, but it's probably not true. A grain of salt in those conversations because nobody's nobody's doing it to be malicious, but the org organizations get addicted to technologies, and so I think the from a leadership perspective, we need to pay attention to is this resistance uh instinctual or is this is it rational? Like this is where all the information is, and if I lift it out, it's gonna unsort and then I'm not gonna be able to meet this compliance regulation. Okay, that's real. Let's figure out if that's worth solving or not. Okay, we're all resisting. I see it every day. I see it every day, every conversation. People much smarter than me, like still trying to figure out like, is what I described in latent vector actually possible, or am I smoking something? Right. Right. Um So the first thing is to make sure that people aren't clinging, right? With their heads in the sand. Like, are we having a rational conversation? Because three years from now, we're gonna sure as hell hope that we were having a rational conversation.

SPEAKER_01

Right, right. After all this investment, absolutely. Right. Yep. I I'll make this one of the final questions here is that is about kind of risk. And you touched on a little bit there to the extent like there's there's certainly, I know when I ask AI information, there's certainly the what they call hallucinations and just you know uh failures, but then what about overall? Just data risk in general, the security, because uh, you know, there's heavy fines now, particularly in Europe for pri any privacy data gets uh exposed and those type of things. And if uh if AI is is taking that information, is there is there some risk on that? So what do you think are the risks of AI of using it at this point? Um hopefully that'll get and will it get better, I guess I'll ask it that way.

SPEAKER_00

So I think the the way I think about AI from a risk perspective is the same way I think about other software from a risk perspective. Because it just makes it easier because we're all we are all used to thinking about it that way. So if it's free, then you're the product. Right, right. Right? Right. Um so I try very hard not to let anybody install anything free uh because that they're doing something with with our data. Um and you have a responsibility to the owner of the information. Right. And so I have my customers' data. I'm responsible. Um I have to be a responsible custodian of their data, and which means I can't give it to an AI without their permission. Right. The same thing as you know, putting it out on the web or or whatever. It's this so you I think if we use the same muscle memory, then we're in a much better place. Um the risk to data loss, I don't think is I mean, it's real, but I don't think it's substantial, right? I think you know you know it is possible for someone to manipulate Claude into giving information that you gave it. But it's an edge case, and for most of us, nobody's gonna spend that kind of time and money to try to find out if I asked about Brian Rice, right? Right. Um but again, if it's free, like uh uh the the Claude bot is is free and you're installing it on your computer and you're giving it a lot of a lot of rights. Like if I say it like that, you know it's not a good idea. So I think when you're talking to your employees, uh when you're talking to your family, just sort of reinforce the fact that it's from a like from your computer point of view, it's just software. And how would how do you think about software, right? And what do you what do you want to what do you want to protect?

SPEAKER_01

Yeah, it makes I think it makes great sense. Dave, is there anything that I'm weaving out that you think is an important aspect? I tried to kind of cover the bases of uh particularly as it affects the cost basis of of companies. Um do you uh anything that I you think I left out that you that you think we should be discussing?

SPEAKER_00

I think that one of the things that we need to do as as sort of people is to think about what the implications are at on a global scale for this this kind of trend that's happening. Because we're see what we saw with uh you know a news report in IBM stock, what we're seeing with you know job losses and in in various pockets, um, these are patterns that I think are going to reverberate across the planet. And so if you take a look at uh you know world economies where uh they are primarily outsourced connective tissue functions, right? Then the implications are much greater to them than than they are to maybe to you and me. And so I think there's there's a dimension to this that goes beyond business, it goes beyond like my paycheck. And I think it's important for us to at least keep that in mind. Um, because like anything else, we're doing this to ourselves, right? Like we inverted, we invented this technology, we're introducing this technology into our businesses, we're harming ourselves with this technology, this is the pattern. And so I think as people and as investors and as as human beings, like being aware of the fact that this is a going to be have worldwide implications is important. I don't not that we could do anything about it necessarily, but I think it's important to keep in mind.

SPEAKER_01

No, I think that's I think that's um uh you know key and uh not so much from uh how to manage the cost, but but just understanding the whole getting into the whole ethics and implications and those aspects of of AI that um that that I think is uh a area of study exclusive that you could you could exclusively dedicate yourself to because it's massive and and critical and critical. So I I appreciate uh David you coming on, sharing this with listeners that that uh some great insight. Um look forward to your next book. Uh so let me know when that's out. We'll send you a draft. I would I would very much appreciate that. I've really enjoyed the first one. It was eye-opening. To me. So thank you for your time here. Thank you for listening. I encourage you to check out other episodes of the Cost Management Center podcast. I believe we offer great advice to individuals that are looking to manage costs within their organizations or companies. If you would like to reach me with a question or comment, please do so on my website, the efficacygroup.com, and that's EFFICACY Group.com. I would love to hear from you.