AI That Works: Turning Hype Into Real Business Solution with Daniel Levin

Speaker A: Hi, welcome to Live By. My name is Danielle. Um, welcome to the podcast. Today we have an exciting episode. We have Dan Levin, our La Vinta's president on, and I'm excited because we're here to talk about AI. I know there's a ton of information about AI every single day, every newsletter, every social post, and it's drinking from a fire hose. And I'm trying to absorb as much as I can. What Dan has been doing for me and the rest of Laventus team is taking all that information, distilling it down, and translating it for us to its most salient points. So I don't know, he's been doing that for five, six months. It's been really informative and critical and really a time saver. So I think this will be a really interesting podcast as we can pick his brain. He has, as most of us, but I think even more so diving into all the AI news and trying to translate and understand. And so let's like, open that brain up and see what's in there and uh, uh, get some of the stuff you've been reading and learning. I mean, that's, I'm excited for it. So thanks for joining us today.

Speaker B: It's quite an intro. I don't know. I don't know what that's going to look like when you open up this brain.

Speaker A: Okay, let's dive in. I want to talk about adoption. You have been implementing enterprise technology solutions for over 20 years. Um, I'm sure you've seen a lot of changes in that time. What's different about this AI wave compared to previous tech adoption cycles that you've seen?

Speaker B: I think this wave is different because of speed, accessibility and uh, how an adaptability, uh, you know, the value shift isn't just about replacing processes. It's about, uh, transforming how people and systems kind of work together. Right? So like in past cycles, take the Internet boom. Or mobile, or even close cloud. In the early days, you had time to plan, you had time, uh, to adopt and integrate. Even though it seemed like it was moving fast, it did move over a longer period of time. This AI wave is moving exponentially faster. You know, we're seeing major leaps in months, not years. And it really compresses the adoption timelines, increases, the hype, the fear tactics. I mean, it's just, you know, a crazy cycle, I think, that we're, that we're experiencing. And the competitive gap, you know, between the early adopters and the people who are more late to adopt can widen very quickly, almost instantly. It would seem. And all these AI models, you know, one is created two weeks later, you know, one's created. And it's the greatest thing. It's beat all the benchmarks and then all of a sudden a week later, one surpasses it. So it's just, it's moving incredibly fast. So, you know, I'd also say, you know, with previous waves, I think that the bottleneck was probably different. Uh, you know, bottlenecks previously were maybe an infrastructure or budget, which they, those still exist, no doubt. Now I think it's just the ability for people to adapt their processes and their governance and their learning so they, the tech can do the work, but your team has to adapt to the workflows and the compliance and the culture fast enough to leverage it properly and responsibly.

Speaker A: Uh, yeah, I think the pace of change is just beyond a lot of people's ability to. Whenever new technology comes out, you have to understand it, kind of get a sense of what it's doing and then figure out how you can apply it to your business. Those take a minute. And to your point, about one thing comes out one week and the next week it's something new. And now do I have to learn that new thing or is the thing I learned this week gonna be sufficient? So it's a lot coming at people and, you know, you don't wanna be wasting money, you wanna be affecting your business. So it is, um, I was describing to some, it's like being in the washing machine. I bet for some people, just so much going on.

Speaker B: It's a good analogy.

Speaker A: Um, so you've had, uh, we're talking about approach now. You mentioned some companies are approaching AI backwards. Uh, what do you mean by that?

Speaker B: What I probably meant. Remember when I said it, you know, certainly my opinion. I just think, you know, kind of going off of what you were saying. You were kind of talking about hype a second ago. Word. But it's very easy to get caught up in a shiny tool instead of a real problem. Right? So the companies that start with the shiny tool instead of the real problem creates a science project. If you start with the problem, you know, you're, you're creating an roi. So, you know, don't chase the trend, chase the problem. And um, and it's very easy to get caught up in the hype because, you know, in the, in the, the clickbait and the fear headlines are crazy. You know, it's. And it's, and it's happening so fast and it's in front of you Every single day, at least for me. I don't know what other people are, are searching and what other, what they see but navigating that. It's very easy to just go down rabbit holes and get caught in hype cycles and, and chase uh, chase the, the shiny new object instead of, instead of business problem. So that's probably what I meant by that. I have to remember when I said it or the context. But yeah, I'll go with that.

Speaker A: Yeah, I couldn't agree more. I think, I think it's, it's, you know, we'll get the question sometimes, how should I be using AI? Right? And the real question is that maybe the better question, it's okay to ask that, but you really don't want to just implement something that doesn't have an roi. So maybe the next question should be what are manual processes? Where's friction with my customers? Stuff like that. And what's fun. Even though it's so much change and so much um, I guess volatility, you can kind of just blue sky it and say I bet something can do this for me. Versus in the past you may have been more limited. Now that's not exactly true, but kind of, you know, we used to say king of France type of approach. Like I can get anything I want and then see do the start there with your total blue sky optimism. Um, what do I need for my business and see if AI can deliver it right.

Speaker B: And it's no different than addressing any other problem by the way. I mean.

Speaker A: Mhm.

Speaker B: Forget about AI. Just pick any technology or any shiny new object. Don't chase that. Right. You gotta rebrand yourself and figure out what, what, how are you gonna use it to your, to your benefit? What problems are you gonna try to solve and, and go from there. Right?

Speaker A: Yeah, absolutely.

Speaker B: Unless you're the king of France.

Speaker A: Unless you're the king of France. As I mentioned in your intro, you're definitely someone, I think more so than most people I know just trying to keep up with the latest AI tools. Everyone's I'm sure to some degree. But you know, to me you're you know, on the leading edge of it. As I said, use you uh, and appreciate you for helping me stay up to speed and your communication to the team is invaluable. So giving that kind of context, you know, what AI tools do you use personally every day and what AI trend is you think is overrated? Like you said, there's some hype in there. Not all of this news and all of these headlines are going to turn out to anything meaningful. So you have to distill between the noise and the, uh, kind of reality. So anyway, tell me what tools you use every day.

Speaker B: I'm, um, curious about this. Uh, well, I'm flattered, so thank you again. I think of practical ways of integrating it into my workflow and not chase a, uh, shiny new hype object. And for me, navigating that clickbait is also a real thing. Right. I have to, uh, you know, I gotta go through the same hype cycles that everyone else is seeing and, and you know, when you see GPT5, just release and it's going to change everything everywhere all at once. And you're behind if you're not using these 10 things and yada, yada, yada, you know, just be grounded enough to read through that BS for a second and, and think practically. And that's what I try to do. And so tools that I use are, you know, I think I use AI in, in every part of my life. I'm not using it probably nearly as much as a lot of other people. So there's, it's a spectrum. But I use it in a lot of things. And so Chat GPT and Grok, you know, just off the bat are probably my main co pilot and brainstorming and drafting and problem solving, you know, Chat GPT is, you know, knows me definitely better than I know myself. And it is a, uh, you know, I use it very Socratically and I'm constantly challenging it. You got to read through the bs. I mean, these things are designed to love you. These things are designed to compliment you. These things are designed to, you know, kiss your butt and, and you gotta, at first, you know, when these, when those came out a couple years ago, you know, they're definitely confidence boosters. And you're like, whoa, see, I'm on the, I'm thinking the way I should be thinking. And then you have to be grounded enough to say, wait a second, why, you know, why does it always compliment me and everything I'm doing and fight back and, and, and use it to bounce ideas but you know, continue prompting it in ways where you're playing devil's advocate, where you're calling it out with things and just like you would maybe in any other kind of conversation, don't take its first response. So I, I use that a lot in brainstorming. I'm constantly using other tools like Gemini and Claude and Perplexity and a lot of these GPTs as well, because they all offer different strengths and and they're also getting surpassed by one another every couple weeks. So just in someone that is in the AI space, I have to set aside time to just understand as much as I can play with them. We use AI coding assistance like GitHub, uh, Copilot and Cursor and many other tools that will accelerate our development work for data. There's a ton of AI features being introduced in Power BI, and we're developing our own in our SparkFlow product that, uh, can surface data insights faster and really turn data into a, uh, proactive versus reactive tool. Other tool, you know, Fireflies, uh, is a great meeting summation, summarization tool, uh, that I use daily. And it's kind of like a lifesaver learning workflow automation tools. You know, I've talked to you about N8N Power Automate. Obviously, you know, these are things that aren't so, you know, they're not secret. They're just very powerful everyday tools that a lot of people should be using. Let's see, what else. You know, I, uh, probably am using others. There's probably more complicated things. You know, we're doing, we're building our own MCP servers and we're building our own, you know, models and you know, for some of our products that we're developing. But those, uh, you know, I'm not personally using those every day, um, because they're not ready 100% yet. But, um, yeah, I mean, those are the ones I listed are probably great everyday tools that a lot of people can get. M. You asked me about the trend.

Speaker A: I asked you about.

Speaker B: Did you ask that overrated trend question?

Speaker A: Yes, I did. Trend. That's overrated. Yep. That's the next one.

Speaker B: AI generated everything is not a winning strategy. Right. Or everything could be done in seconds. I don't, I don't know if it's an overrated or overrated trend, maybe a misunderstanding. You know, just be. Just because AI can crank out a bunch of content or code or images doesn't mean it should necessarily. You know, quality and security and alignment with business goals is still, you know, number one. So security holes are very, very real. AI tech debt is very real. Flooding your systems and just the world, your zone, whatever you want to call it, with mediocre AI output, um, is going to hurt you more than it's going to help you. Having an unlimited amount of employees doing an unlimited amount of work without proper instruction and direction and training is going to do more harm than good. Right? So AI is really no different. You have to use in A structured manner. So I, you know, I'd say that's probably an overrated trend that, you know, that's my favorite Jurassic park quote. Uh, you know, now I'm forgetting what it is exactly. But just, you know, people get so consumed with whether or not they can't. Whether or not they could doesn't necessarily mean they should. So. And probably a very common misunderstanding which is along the lines is just that AI can just magically do everything all at once, right? So magically just recreate a system for me. And you know, a lot of the times, especially with legacy systems, it can do certain things and it can definitely speed it up, speed up the development quite a bit. But it's not going to get you a hundred percent of what you want overnight. It's, you know, you need direction, you need business rules, you need logic. It's. AI can't do that. This is still just uh, generative output based on statistically what it thinks is the next uh, part in the pattern. Right? So the, so it's getting very good and it's getting some self correctiveness. It's not um, these, these tools can't think for themselves so they're improving memory which can give it more, you know, more space to, to think back in its, in its context. But it's, but it's still generating output based on statistics. So it's not thinking for itself. It can't create your business rules for you. It's, it's gonna, even though hallucinations are down in GPT5, they're still going to exist. Um, you know, so it's a, it's, it's an efficiency multiplier but it's not a replacement. And so really understanding that difference is huge. And, and there's, you know, a AGI is the next, you know, the uh, you know that, the next model gener, you know, general intelligence. That's, that's kind of the next huge leap, right? That's when AI can produce things um, that it's not trained on. Right? So today you've got the narrow intelligence which means that AI can only produce, based on statistics on what it's trained on. Um, right. But what does it think the next word in the pattern is? You still gotta have the governance in place. You still have to, you still have to have the understanding that AI is not gonna get you from zero to a hundred percent. You, you still need all the things you needed before you can just get there faster, right? You still need people, you still need governance, you still need requirement Building, you still need security layers, you still need an iter, an iterative approach to get to where you want to be. So hopefully that makes sense.

Speaker A: Yeah, I think it's really interesting. So instead of just one tool being overrated, um, I like how you put it, how it's not like, so what I was thinking, you know, I've watched videos, you've sent me videos of, hey, you can get this going up in three minutes. You can build this workflow in three minutes. And they do it. And the video is not inaccurate. But if you want to apply it to your business, it may take longer than three minutes because you've got to, you know, troubleshoot a few other nuances. My point in that, saying that is there what the news headlines are not exactly inaccurate and those demos you see are not exactly inaccurate. But when you, if you expect to apply that to your business and have effect, you may not have this. It may take a few more steps for you, and you may need a few additional tools. You may need some additional learning. So you watch the ChatGPT5 release video and they're doing all this cool stuff and it takes minutes, doesn't always mean it's going to take you minutes. It will eventually save you time. But to apply it specifically for your business, you may have to, like you said, iterate a few times, be patient and play with it type of thing, Right? Yeah.

Speaker B: And, uh, you know, we're in this, we're in this, we're in the middle of this boom, this technology boom, if you will. And we really don't know what tomorrow will bring, but what we do know is that there's a huge, there's a lot of technology coming, right? And so again, if you just think about the fact that AI is only good on what it's trained on, right? So these huge amounts of, of, uh, data that AI is being trained on. So AI can produce code quickly, sure, it could produce books, it could produce a lot of things based on what it, what it's trained on, but it can't create things that it doesn't know about. And we're about to enter a world of new technology that AI doesn't even know about. You're going to have to, it's still going to take time to adapt and create all this new technology. Now that, you know, these new types of, whether they're ways AI, uh, can integrate in the systems or different, different interfaces and agents that are yet to exist, you know, you're going to have to create all that. So, you know, and and your, the time it's taken to create your interfaces. You can only focus on that new technology if your interfaces are getting spun up quickly and, and some other code is getting spun up quickly. So the, you know, so things that are taking a long time, today or yesterday, I should say, things that used to take a long time are now sped up. But things that are now going to take a long time are the things that have yet to exist that we have to now spend our time and our thought and shift our mindset and our mental energy to creating these new things that we didn't have time to do before. Hopefully that's kind of making a little bit of sense. Maybe you could articulate that better.

Speaker A: I think you nailed it. I think it's just, um, a little bit of seeing the forest for the trees. Sometimes when it's coming at us this fast, you know, gain a little perspective helps. We mentioned a couple times in, so far ChatGPT 5, we've talked about Chat GPT. So the 5 release just came out last week. Was it maybe something like that? Maybe two weeks ago? Yeah, I don't know. It's measuring these things in hours now. So, um, again, using you as. I've appreciated you for the past couple months of getting all this information, distilling it down. So for our listeners, what does this release mean for the latest advance, mean for business? Should they kind of take advantage now or kind of what's coming on the horizon? What's the next big thing? So I guess maybe characterize ChatGPT for us a little bit, how you think businesses can leverage it? Um, and I think this question's actually asking to look into the future, but just, you know, what do you think about adopting now versus waiting?

Speaker B: Well, I don't think you should wait. I think that, you know, I think there's a saying, perfection is the enemy to progress or something like that. So don't wait for perfect AI. You know, today's AI is definitely imperfect, but it's already outpacing a lot of, you know, things we were doing yesterday. GPT5 is just the next major step forward.

Speaker A: Um,

Speaker B: but waiting is a, is a bigger risk. You know, the advantage comes from starting now and evolving as the tech evolves. So GPT5 is, it is smarter, but it's not just smarter, it's more accurate. It's better at reasoning. It can, it can work on, like I was mentioning before, it's got a larger context window so it can remember and process a lot more information at once. And that makes it Better at, you know, enterprise level tasks and analyzing a lot more data at once, analyzing thousands of documents, you know, orchestrating multiple multi step workflows and you know, understanding larger code bases and integrating with, with systems that act as a, uh, you know, it can act as, as a better co pilot, you know, rather than just a chat assistant. So I think you should not wait for the next big thing because there's, you know, it's, it's already a pretty big thing and the capabilities are just compounding, you know. Uh, so today's GPT5, which came out a week ago, you know, they're already talking about, you know, there's so much hype behind now Google's next release and um, I'm sure Grok is following Meta. I mean don't, you know, the amount of money that Zuckerberg is throwing at AI is just absolutely insane. So uh, you know, if you wait, you'll have to build under a lot more competitive pressure, right? So if you, if you start now, build the infrastructure, the culture and the muscle memory, uh, you'll be in a much better place, you know, and there's so many systems that you know, are, you have to, you have to upgrade and you have to enhance to, to take advantage of some of this AI. So you know, that's a good place to start too, even before you can even really work and integrate with some of the AI. So I think the, I think the successful businesses will be the ones that just adapt quickly and, and have, you know, proper feedback loops to iterate through improvements. Right. And keep going. And that's probably true in anything. And it's no different with AI adoption other than the timelines are significantly increased. Right? You implement anything, build a good feedback loop, adjust and iterate and continue to improve. It's true in anything you're doing, right? Whether it's learn how to play soccer guitar or adapting AI.

Speaker A: Uh, yeah, I think that's a good point on a good way to help people understand why they shouldn't wait that part of it. Even if some, some of what they do and learn may be outdated or there may be something better coming along, the advantage is still getting you and your organization accustomed to leveraging these AI tools, accustomed to finding ways to build them in the organization that adds value and it can be building blocks. It doesn't mean you start with something and it gets wiped out because the next big thing comes on top of it and replaces it. It could be foundational and you're adding these tools to your point about having an infrastructure there that can leverage these tools. And so starting now is important versus waiting for the big magical thing that's going to happen, which it's kind of the big magical things happening every day and then the new one comes. So I think it's a good point about the muscle memory you mentioned and the exercise of getting used to leveraging new tools and building them and using them in your everyday. I think that's you know, a case to be made to not wait kind of talking about not waiting. We were wanted to talk about kind of how companies get started and I think what I hear, I'm sure you hear it, there's executives, business runners, fear of missing out on opportunities combined with or you know, kind of juxtaposed against the risks of AI, maybe unchecked AI adoption or making mistakes. So we work with companies in highly regulated industries, Healthcare, financial services have a lot of regulations. You know, how do you help them navigate the wanting to be, take advantage of these tools but not fall out of compliance, not fall out, not have security issues, that type of thing.

Speaker B: I think it's about moving ahead smartly with guardrails. I think the fear is valid because unchecked AI and uh, you know, will introduce compliance risks, it'll introduce data exposure that creates reputational damage and all the bad things. Right. So in regulated spaces like finance and healthcare, insurance and whatever, you know, that fear is even more magnified. Uh, that doesn't mean ignore it and, and let your competitors figure it out first. Um, it just means move forward and progress. So as best you can just do it with governance and guardrails. Um, because your competitors are going to be doing it and other people are going to be doing it. So just figure out how to do it wisely. That's what I would say.

Speaker A: I have found it interesting in talking to friends at ah, unnamed enterprise level companies and how little AI they're able to leverage. And so I think people are struggling to find the balance of leveraging these tools, accelerating the workflow, workforce productivity and maintaining security and compliance and control. So I think it's still a pretty good struggle because again I was pretty surprised how limited and all these names would be ones you would really recognize and they can't, they can hardly use it. It's a, I think it's, it's. People are still trying to navigate their way through this. Um, so continuing on talking about getting started. So, and we get this question, so when a client says we want to implement AI but don't know where to start, how do you Handle that

Speaker B: same way I've been saying. I think you, you, the best technology projects, um, are no different than AI projects. They don't start with a question of how do I implement this technology or what can AI do. They start with what problem am I trying to solve for. Right. So if you start, if you start with the business value, not the technology value and work backwards, you'll get the right AI approach. Um, you know, are you trying to reduce costs? Are you trying to grow revenue? Are you trying to speed up decision making, you know, uh, improve customer experience? But you know, get the business goal and then the tech solution will kind of uncover itself and the AI solution will uncover itself. But without that clear business goal, you're gonna. What did I, I think I said you're gonna go. It becomes a science project, which are important too, just in terms of research and development. I mean, you know, it all kind of gotta start somewhere. But the, I'm sure the invention of AI itself wasn't just, hey, what can I do? Uh, let me do a science project. It was, it probably came from an idea. I'm speaking completely out of context, I have no idea. But I'm sure it came from how do we, you know, how do we automate certain things or how can we, you know, they're probably trying to solve a problem, um, that usually leads into a successful implementation.

Speaker A: Yeah. I think we're seeing companies with very interesting. I think we, we're getting this question less and less. We still get it, but I like how we're getting. People have ideas now and they're very distinct and specific. So just talked to somebody yesterday, big tile company. They, their salespeople have to deal with like 10,000 different types of tile and they'll be in front of a client. And he was trying to leverage AI to say, hey, I'm looking for modern farmhouse in teal with this type of look. And the really good salespeople can get to it in five, six minutes. And the newer ones might, could take them an hour to go through. Can you leverage AI to look at images and interpret things like that? So really specific and distinct. And we're seeing lots of cases like that. They aren't the Mark Zuckerberg level investments, but I don't think these companies want to do that. They want to start somewhere, find a place where AI can improve and reduce friction and improve efficiency. And we're seeing a lot of those really distinct cases or ideas or projects and I think that's a really good place for mid sized companies to start. They don't have to think pie in the sky. Just start asking your team, like, what, what makes you crazy because you're having to do it manually. Uh, it can be that simple.

Speaker B: I think it should be that simple. I mean it's, with anything, if you start too big, you're, you know, you're taking out too much, you're probably gonna fail. So start small, get through incremental changes, celebrate certainly, you know, celebrate the, the small victories and then you can grow, grow, grow. But don't take on a tremendously huge problem. Like I, um, you know, I mean there's definitely success stories of doing that, but I, I, I think it's more manageable when you, you know, take something that is a, a more manageable size and you get comfortable with solving that problem and then boom, win, go to the next one and you're going to continue to learn and, and continue to, to find better use cases.

Speaker A: Right, right. Um, um, um. All right, so getting into kind of the practical use of things and we face this question a lot because of all the tools out there. So it's a build versus buy question. Are we building everything from scratch or leveraging existing platforms? So what should a company consider building? When should a company consider building the LLM, GPTs, MCPs or agents instead of, or in addition to the ready made tools? So you get some of those questions of the project and it, you know, when should we, they leverage tools and when should they build them from scratch?

Speaker B: I think we're still talking about AI.

Speaker A: Yeah, yeah, I know this is a general question too that we always talk about, but uh, yeah, in terms of AI tools. Yeah.

Speaker B: Well I think, I think the companies that uh, nail down data quality and access, you know, integration, you know, existing workflows, you know, we, we almost never really build from scratch anymore.

Speaker A: True.

Speaker B: You know, if we're, if we're building, we're typically taking something and integrating and

Speaker A: um,

Speaker B: with GPT5 and so if we talk about like are we creating our own AI for a second? You know, GPT models like GPT5 and Claude and everything, they give us real world capability instantly. So we can fine tune those, wrap them into domain specific agents and um, and connect them to internal systems through MCPS using like APIs. So that gets you real fast time to market and you're leveraging real AI. And so when you talk about build versus buy in anything, it's no different. Right. Take something that exists and leverage it and you're going to have a faster go to market. When you do Custom. Again talking about AI and custom AI, whether you should build custom LLMs or agents, there's definitely going to be use cases, regulation and privacy. Those kind of demands might dictate if you have to build your own highly specialized language or formats that you know aren't handled well in general models. I mean there's a lot of that too. So you know, or uh, it's your own ip, right? Proprietary workflows that require unique reasoning similar to what we are building with our deal ends. And there's also cost efficiency at scale. So AI isn't cheap and the more that you use these models, the more expensive it's going to be. So at a certain time you might say, especially if you're, if you're using it at very high volumes, it might make sense to go private and take it to a cheaper model to, you know, to run privately.

Speaker A: Yeah. And again, this doesn't mean you don't use AI to build these tools. Right. So that's, you know, is AI the tool you end up using or you have to build a custom tool but you can build it using AI. So compared to five years ago, building that tool for yourself can be a lot faster and a lot cheaper because you can still use the AI to build your LLM, build your, build your whatever you're looking. And I've run across a couple people lately. I just talked to somebody this week who now he's a development firm like ours and they were unhappy with the meeting note taker and the about of you um, mentioned Fireflies and they just had a specific use case so they built one for themselves and it, it's doing a lot more than meeting note taker. So instead of just leveraging Fireflies, they, they took that idea and built one for themselves because they had additional wanted to look at documents. I can't remember the specific use case but it was pretty interesting in that for them and made sense and they did use AI tools to build that but they didn't end up using an off the shelf meeting note taker. So there can be these reasons and it can be interesting depending on what you're looking for. You um, just mentioned dln, so let's talk about that before we close out you're building or have built. We're almost done. Uh, AI tool specifically for equipment finance leasing. Why do you think this is the right use case and what makes this in the area, what makes this area a right fit for AI use?

Speaker B: Well, I think that equipment finance has three things that AI loves and that's A lot of data, repeatable decisions and huge payoff for speed. So uh, it's the perfect case. It's perfect use case for you know, data rich decision, um, heavy decisioning burdened by slow manual processes. Right. Every lease application touches a ton of data points, credit bureaus and bank statements and tax returns and background checks and, and you know, resale values and equipment lookups and even location lookups. All that, you know, and a lot more uh, all that data has to be um, pieced together by humans manually. And even a lot of the automation that's put around that there's still a lot of manual, um, there's still a lot of human elements to it. So their decisions are complex, um, but they're repeatable and um, you know balancing dozens of factors becomes difficult sometimes if you're, if you don't really have it in uh, an automated process. So it's perfect for AI because those decisions can be modeled, that logic can be modeled, that can be improved upon, it can be applied consistently. And speed is a competitive advantage right in equipment finance and all kind of finance like that. So whoever can provide deals faster, you know, without increasing risk wins. And you know I can go down a whole rabbit hole here on, on um, crypto and, and how just our, you know, we're short on time now but you know, our banking system today is too slow for what AI is going to, I mean AI is going to be creating such fast decisioning, you know, certainly can fast forward in a couple years. But if you're paying attention to what this administration is doing with wanting um, to advance crypto significantly and the current modern banking system just being slow and you know, waiting day sometimes to settle transactions. And you know there might be a big future with finance and crypto because crypto is instant and it can be um, uh, governed with AI, you know, with these AI tools. So a lot of this is about speed, it's about repeatable decisions and just making things faster. And so I think kind of going into what I was saying earlier and how AI can only get you so far based on what it was trained on the future that we're headed to, there's going to be all, there's, you know, we're gonna, we're gonna get to a point where we have to redo a lot of uh, and create new technology around everyday task. Like our banking is just you, you gotta, you're gonna speed up and make a lot of things more efficient but you're still gonna hit a bottleneck and then you gotta fix that bottleneck. And, and so there's going to be a lot of things down the road that are going to be improved upon and changed significantly. And I, uh, think a lot of it is going to be in the finance space. So the DLNs and the AI we're creating around that is going to be helping, uh, those faster decisioning and uh, ingesting and um, analyzing a lot of data and, and take things that are even quick today, taking 15, 20 minutes, maybe today for, for decisioning is just going to be seconds. Um, and that's, and that's part of what we're doing and working on.

Speaker A: I love that initial filter. You kind of said it has lots of data, repeatable processes, you know, repeatable events and an roi.

Speaker B: And I can think of it's ripe for the taking.

Speaker A: I can think of other industries that uh, have that as well, like retail. I came from retail and wholesale. Same thing, Tons of data, repeatable, same thing and huge ROI to kind of narrow down some inaccuracies and friction with customers. Well, we're gonna need to wrap it up now. Thank you, Dan. This is super interesting. I really appreciate your time and your insights. Thank you for joining the Live by podcast.

Speaker B: Thanks for having me. Thank you.

Speaker A: Thank.

Speaker B: Bye. Okay.