speaker-1: Welcome everybody to another episode of Decline Invite, the only invite worth accepting. I'm your host, the extroverted IT leader, Ryan Drumheller. And a lot of companies are trying to implement AI right now, but most of them aren't getting the results they expected. Not because the technology isn't there, because something else is missing. Today I'm joined by Nikhil. them all. ⁓ DeKiel is the founding founder of the Accelerated Company, where he's working with leaders, founders, and organizations on how to actually build and scale AI. We're not just talking about it. He spent over a decade across AI, product, startups, enterprise environments, including building generative AI platforms for Fortune five hundred companies. So he's seen from both sides what works and where things fall apart. And what stands out from his perspective is he doesn't start the technology, he starts with the business, how it actually operates, how work really gets done, and where things break. Because that's usually where the problem is, not the AI. So Nikhil, welcome to the show. I'm glad that we were able to do this. We just met on a different podcast the other day and here we are again continuing where we left off.
speaker-0: Yeah, no, thanks for having me, Ryan. And I'm excited to sort of dig into the the topic we have at hand today a little bit more. So yeah, it's an honor. I'm I'm happy I accepted this invite, right?
speaker-1: That's a question for later. And we're not done yet. So first thing we like to do is get into some icebreakers. So what is one AI or product buzzword that you'd retire immediately?
speaker-0: One changes like every week. ⁓ I because it's tough, right? I mean, I gosh, like I've been pretty anti-generative as a word anymore because I think it I think it actually kind of degrades to like when we're just thinking about producing and not, you know, there's value in regurgitating and there's a value in enhancing and augmenting. There's other words to use, right? And generative almost seems a bit derivative where it's almost like you're just kind of making something out of nothing or out of some minor input and it's supposed to, you know. be this like end all be all silver bullet kind of thing. But I think it's just more nuanced. And it o it always has been. I've been in like prior to generative AI being like the thing that shifted it. It was conversational AI. Machine learning is what I used to work in. It wasn't really it was AI, but not like that wasn't the word we used every day, right? I think it's actually kind of starting to rotate back into that. And I think that's a good thing. I think it I think it sets an expectation. I think it's a little bit more scientific to look at things in their nuance and and maybe find better words than just generative, right? I don't know. Like that's that's just not as strong as it Maybe people had thought it was. So yeah. I don't like the vibe word either. I don't know why. That's not I that is the only way to describe it. But I think there is a better again, there's a better ways to I'm hoping the word augmentation comes into the poll. I think that's a good one. But it's not it hasn't taken off yet. So yeah. Anyways, that's my that's my answer.
speaker-1: It's a great perspective. It's the first one that I've heard and I think it's great from from your point of view because you're actually in it working every day, which we'll get to. What would be the invite you declined now that you didn't earlier in your career? Man.
speaker-0: And do you mean across like any opportunity or anything like that, or do you mean a specific type of in
speaker-1: Like no, no, just yeah. So something that somebody asks you to do and you're like, I no longer I no longer want to participate in that because I've learned from that approach.
speaker-0: Man. I mean, that's kind of the secret of my whole career is out is when I knew how to outgrow things, right? So there's there's like I I'll think of like one one good example I think was, you know, early on in yeah, like earlier in AI, I think I didn't believe in it as much as I maybe should have been an interesting way, because there are people who are like, we always talk about niching, right? And I think niching is good industry wise sometimes, right? Like don't like just I mean, you can't possibly build something for everyone. But I think at that time I had sort of like I'd I got really into conversational AI, which is very good. And So it's a very strong community. And it is it's everything is conversational. Yeah, I know, basically. But there's like a even further in the customer support world. But I think I wish that maybe I had expanded my AI sites beyond customer experience like 10 years ago. Cause I think I would have just been like there would have been a little bit more like speed in my career to to look at some of the other parts of it too. And that's why even though I was talking about the word generative, like I'm not a fan of of digging down like with that, there's nuance and there's interesting ways to do it. And we should be allowed to have room to experiment, right? If we if we have the privilege too. And so ⁓ yeah, I think I think that was the only thing is I almost like discounted how broad AI can be because maybe and at that time it could have even just been like imposter syndrome, right? I was just, you looking back, I was okay, well, conversational AI at that time, Alexa and chatbots, that was like the cool hype thing. But there was also there's always been a lot of really cool stuff in machine learning and I've had to catch up over it over the years, but I'm I'm happy because then I was ready for this AI boom better than than that last sort of like Sassin AI ⁓ boom that had been again, maybe a decade ago. So yeah.
speaker-1: Yeah, you I mean you bring up Alexa. I mean I remember playing around with Lambda ⁓ at some point way early and I am so much further away from the area of AI that that you're in. But that was like my first foray into playing around with it and I was so proud of myself for getting simple commands even though cool though there.
speaker-0: Yeah. I mean it was you you felt proud doing it though, right? 'Cause it was it was like the early stages of what we have now, which is just accessible AI, but it felt good that you could you could do something. You didn't have to be like, you know, a a a rocket scientist to to figure it out.
speaker-1: Yeah, like whatever room name you pick, right? I wanna program it with this, turn this on or off, and it's like awesome, it works.
speaker-0: Internet of things, right? Yeah. Good old days. Good old days.
speaker-1: So what's something companies think AI is solving that it isn't? ⁓
speaker-0: Yeah. Another ⁓ like I I always kind of crack up because it's like these are these are just great questions to start off with and and do set the tone, right? So ⁓ it's it is I I'm very much against the let's like for everybody and then rehire after AI has been solving whatever it solves and whatever it doesn't. I do understand 'cause I mean again, as someone who's worked for five hundreds, I do understand the notion and I understand why they kind of do it for my, you know. like traditional MBA business standpoint, like there is some, you know, I it's it's quick win, right? It's quick win and it shows market signals and all these kind of things. But I think a lot of companies and not just Fortune 500s are in for a rude awakening when they realize that like they could open claw and they could automate and have a thousand agents who still can't solve the value problem. Right. Like how do you how do you actually make buyers click purchase? How do you make customers sign contracts? Like you can't have AI for somebody's hand. We're not there yet, at least, right? I mean, there's there's a world where AI is talking to AI, but the biggest, and this is like the biggest shock. And it's still this day, I've been in conversation with AI now for 10 years, right? And ⁓ generative AI since it was, you know, whatever introduced. And I still see an overwhelming amount in any big company that I work with or I work with like e-commerce, whatever, the whole thing where it's like, I want to talk to a human, even if AI solves the problem. Presents the right FAQ, enriches the right answer, schedules the refund. I still frequently, frequently see people in different age groups, probably a little bit on the older side, but that's still a pretty big buying power audience, saying, I want to speak to a human. I want to really make sure my refund is coming. I myself, like I've used automated services and the, you refund never came or the product never got returned properly until I talked to a person who confirmed it, right? The AI didn't come back to me, wasn't proactive, it didn't email me a confirmation. Like these little things are missed way too often for comfort. And that's a small example, but when you look at it on a grand scheme of things, I mean we're we do see efficiency gains. I totally agree. But the solution isn't just to like start, you know, firing left and right and introducing new departments for the say I I say a lot of people recklessly introducing like R and D departments and letting go of all the IT team, right? Or or replacing a C D T O with a C A I O. But why like that's not it's not just a signal thing. It's noise. It's becoming like they think it's a signal for the industry, they think it's a a competitive advantage, but they're really just making the noise more noisy in the in the industry. And usually ⁓ I think the number is still, I think it's gone down a little bit, but like a year or two ago it was 90% of AI pilots are failing. I think it's still up there. I think it's still 82% or something. So there are more pilots that maybe going to phase two, but there's still a lot of enterprise opportunities which are are being completely ⁓ messed up because you can't have like AI build its own good pilot if it doesn't know how to like v establish that value. Right. So I could go on a tangent all day, but I will say that one is That one is very tough and usually with the companies I work with, it's a lot more holistic than just let's clear house and start over, right? Like that's that's not usually the way to go here. It never has been in business, to be honest.
speaker-1: And like I I just read an article this morning that was just like CEOs throwing in AI projects just to say they're trying to do AI projects and don't even align to the business goals or or anything that they want to do. It's like what are we doing? Where how did we get to here? It's like AI got marketably big, we'll call it, you know, marketable. ⁓ yeah. And ⁓ yeah. Then we lost our minds of how to do business. So
speaker-0: Right. Which kinda like defeats the point, right? Because like the ⁓ per purpose of business to buy and sell product and then like you know, create supply and demand. Like there's all these fundamental principles which AI like definitely doesn't understand. Otherwise, we would all be saying, Make me a billion dollars and I'll have a billion dollars, which is not the case, right? So but like funny how you like when you really dig down to the philosophy and holistics of it. Yeah.
speaker-1: Yeah, absolutely, absolutely. So transitioning a little bit, let's talk a little bit about your background. So what pulled you into AI and
speaker-0: A fun a fun ⁓ opportunity actually because I used to be a self-taught like software engineer and and full stack engineer back in the day. I got into the tech industry around web two point ⁓ so AI definitely wasn't the top of mind. But like at that time, Whites web apps and apps were were just hit blowing up for the first time, right? And ⁓ it wasn't really about the to be honest, it's funny because now it's all about the libraries and tools that we're using, but back then we had very little to begin with. G GitHub was a new site when I got started, right? Which is crazy thing to think about. And all these like Code Academy and all these ways YouTube university, like all that was a newer thing. And so one of the philosophies, I still champion it today, but one thing that I really got into that helped me learn was the world of open source. Like I've actually personally had the opportunity to contribute to programming languages and actually like Firefox, for example, and other places. And I learned how to code better through mentors and and other opportunities found in that community. So I'm just giving that a nice shout out there. But through open source as well, there's a little company, which is not so little anymore, but Hugging face used to be a side project essentially before it became I I think they're a unicorn now or they're or nearing it at least. But they used to be a website where people just shared code. Like it was just like you were sharing like some machine learning stuff, some internet of things. It was a very, very basic site and it was like an it was a layer on top of GitHub. And anyways, I like randomly ran into somebody's like self made ⁓ I think it was like an Alexa machine learning model or something. So to help like Alexa get even smarter or something, you could like hack the voice AI side. It was really cool for twenty sixteen. Right. That was like very, very, very cool for that time. And I just started tinkering around. Right. And suddenly I discovered ⁓ their companies that got rolled up into like Microsoft and NVIDIA now, but they were literally just startups building libraries, right? And open sourcing their code. And the the value was in customer support, or you could buy the enterprise version, but they were building the early stages of voice AI, what we call generative AI, research, machine learning. Principles, voice recognition and voice generation. I've been doing that like a lot longer than than it's been around, than 11 labs have been around for, right? And it wasn't really like they they're the incumbent now, and then it's great to see their growth, but there were other companies trying similar things. And and yeah, it just like it was such a great, it was a great rabbit hole to go down because obviously it was like I was able to look at the foundational code even before using a cloud system, right? I was able to like integrate it on my own websites and play with it and and mess around with it as an engineer. And ⁓ you know, kind of pivoted and then the business mind game, right? It's like, okay, we have all these tools that are for free. How can we, you know, in open source you can quote unquote fork a library and ⁓ you know bring it to an enterprise without having to pay for like a license, right? Like there's or you have to follow some license, but you can you can do things very differently if you if you know how to utilize open source well. And then it's not so much about the code, right? Again, it's it's about how you utilize it, how you modify it and how you turn it into intellectual property, which I was I was doing back then, which was really cool. So yeah, I like to I like to champion open source still and I I give it a lot of credit in ⁓ A its role in AI's growth, but also of course my growth specifically.
speaker-1: Yeah. That that is one hundred percent amazing. ⁓ I've always been on more of the open source side, ⁓ than any kind of closed resource. And maybe I'm showing my age as well with that. But yeah, absolutely there. Android fan through and through for a while, ⁓ on my side. So you've worked across startups and enterprises. What's different between
speaker-0: Startups want to be enterprises, enterprise want to be startups. And it's the funniest thing because again, if we some of the best advice I ever got in the startup world, which has led to sort of where I am now and led to a lot of, you know, what I would say like my my more successes in my career is to get out of the startup mindset once you're not a startup anymore, which means like, you know, the first year, it's fine. Flail a little bit, take your time, you know, f pivot, like all these kinds of things. But if you're doing it three years in, there's something's gotta change, right? You gotta start outgrowing your, your, ⁓ you know, your mindset there, right? And it's interesting because I've also worked with a lot of enterprises who bring me in as like someone who wants to build a startup division, you know, like a an R and D department that runs like a startup, but it doesn't work that way. Right. You're not supposed to with all those resources in place, you're not supposed to, you know, starve a a part of your company that could have all this opportunity and could speed ahead of the competitors who're supposed to properly and again that's that goes into my the crux of my work, right? Just helping people see that that's not really what they're looking for, even if they kind of have like a preconceived notion. Cause then the startup world is not as it's like that's a 90% failure rate, right? Why are we trying to emulate something that doesn't work most of the time due to infighting ownership? Like there's all these other things that happen. So I I find myself at the intersection of this a lot, which is kind of funny because I end up working with a lot of like really interesting like mid market size companies that are still doing like a lot of money, but they have they just don't have the same responsibilities and pressure as enterprise does. And then I've also had the chance to work with like series C, Series D startups where like, wow, we really wish like we could have gotten here faster. If we weren't treating ourselves like a startup at Series A, even right. If we were already not we could call ourselves a startup all day, but like in terms of the actual internal culture and the way that we're going, like we needed to grow up. And and it it helps to grow up, right? For for the people that you're selling to. So it's been interesting to see that AI sort of leveled that playing field in a few ways too, right? It's it's humbled people. Whereas like, okay, if you're using if you're using AI like a startup does at a large corporation, it actually will get you farther. But again, it it doesn't matter. How how quickly you build internal business tools or whatever if there'sn't the value, right? And then on the startup side, it's actually diluted the whole space because now everyone can start a startup. And the same problem is b we go by right back to the crux of it all is like, how do I actually make money? Right? How do I re how do I make a profit and not spend all my money on tokens and hosting, which is very possible if you have a really complex product? Like there's still margins. We still are doing the same business thing and Unfortunately, I can't say AI is necessarily getting cheaper every day because there's a compute problem. So it's fascinating though, right? It's like it it's been, you know, be being in this space for more than a decade now and being very, very in it. Like not just, you know, I didn't come from corporate. I was I started in startups and tech specifically. So seeing it evolve like this and seeing how enterprise, like as I work with them later in my career and and got to know them more, it's interesting to see how everything's kind of flip-flop. But it all kind of comes down to the same fundamental things that are missing. That are just being compounded if they're not solved, essentially, right? And enterprise and startups almost both have that in common beyond their differences.
speaker-1: Which I think is a actually a great segue that I'll go right into. What patterns have you seen when companies try to adopt AI?
speaker-0: Yeah, yeah. I mean, it's always speed. The the top pattern is like, let's be the fastest possible, which like look, nothing wrong with being fast. I move quickly in my projects and consultancy and I advise it. But you have to run at a pace, right? It's not like ⁓ it's a marathon, not a sprint. I I do genuinely believe that. And I think ⁓ I think we're actually seeing that happen with ChatGPT, who just closed down Sora and has had soaringly negative reviews, right? Where they were the fastest and they still are releasing stuff all the time, like no doubt. But they're getting humbled by a company that took its time, ⁓ built a company culture around the kind of AI that they're building, you know, like Chat GBT's kind being left in the dust. I hate to say it even in Grok, right? Like, which is, you know, sort of whatever your alignment might be in in Twitter and it's its dealings, right? Like it is a it is impressive how quickly that's growing too. Magnus over at Meta, like all these things are starting to catch up. Gemini still needs a little bit of time in the hopper, but you like nano banana, right? And all these things. So there is something to be said about like, yes, you might get laughed at a little bit. And that sucks. Like some of these people, I think, went through that Atropic and Google primarily. But now they're showing that like they were actually just being strategic. And and Apple blew a bit of a lead, right? I love I I think AI, whatever Apple tells, just has a chance to still be something, but it's not going to be so model, essentially, right? They're they're even letting people get into model switching. So, anyways, big pattern there. Like that's I'm just saying that even beyond those tech companies, that's a pattern you see everywhere, right? Like the people who are treating it like a marathon and not a sprint are the ones that are winning. I think back to that whole thing of like I think a lot of people are trying to act like they're trying to. change their engineers or their RD departments to turn it to startups, which I'll put it in like two different buckets, right? Positive side, you're being scrappy or using ⁓ company resources well, there's a good opportunity. Bad side is that your expectations are completely misaligned. And that's another pattern. Actually that segues into the, you know, sort of third pattern here, which is across everything, there's always very, very different expectations person to person, because we've all been given AI to use on our personal devices. We all have, you know, we have been using Alexis for a while. Like there's all these different things that have Diluted our perspective on how to actually use AI. Some people believe one-shotting prompts are good. Some people picking up prompts are good. And there's a again, it's not even an easy answer. There's like different AIs that require different things. Image generation requires different than text generation, document versus a quick response and research. Like you still have to have a bit of a sense of, you know, again, like what are you actually looking for? And if your expectations are just that you're gonna make more money with it and not that it's actually gonna drive value and and lasting value, you're not gonna get very far. Right. So it's like Kind of to yeah, to to bring that back down, it's like the it's the expectations, it's treating it like a startup or treating it like a little bit like over correcting on scrappiness and then like speed, misalignment of speed, good speed and bad speed, kind of as such, like good pace and and and sort of like rush jobs. Those are sort of the the main three pillars I would say. There's more nuanced stuff, but they all kind of fall under the same thing. But those are the three that kind of link together one major pattern, but breaking up into three easy ways to see. And this is across mid market enterprise startups, any non profits that I work with, civic I work with some cities on their AI adoption. Everyone is everyone is is having the same sort of issues, but with the right intentionality and clarity, ⁓ very different results, like in a positive manner.
speaker-1: Yeah, I'd say the the wide margin of result and just the same kind of ideas of like they wanna do good with it, but then when we get to the point of actually implementing and then deploying it's like what happened and why are we in such wildly different, you know, situations. It's it's interesting. It's it's a very interesting time. Yeah. Which is gonna lead to a great question based off what you just said, which is why do most AI projects actually fail?
speaker-0: Yeah. Yeah. The the title of the episode or whatever, right? Or or working title. I do think it comes down to it. I actually don't think like the speed is a problem, but like you can still you can still correct, right? I think it's it it goes way more and even like this team thing where it's like you're firing everybody and rapidly rehiring or trying to make turn a department to something else. Like I even see people correcting on that and and being able to stabilize, right? I do think it's an expectation thing, right? I think it's a lot of people, and I'm not just talking about people at companies, but about vendors and partners and even larger companies making some pretty extreme claims of you can do all this with little clarity, with little work and like, and sure, that's like very attractive, right? And it's a great sales technique to just say AI will do this for you. And I and it can do a lot for you. And I and I'm I'm totally, you know, like not objecting to that. But I will say, and I'll kind of I'll actually take it a step higher. to say that one of my favorite phrases, and I use this very frequently. I don't remember where I first heard it, but I use it all the time. It's good data in, good data out, bad data in, bad data out. And these days with with how fast AI is going, it's like good data in, better data out, bad data in worse data out, right? It's starting to, you know, ChatGPD's results results are definitely different than they used to be. And even Claude, right, if you're using a different version of it, it's not the same quality as it used to be because it's just, you know, the the way of the sort of movement at the the space, right? So if you have a data problem, If you already have things that are siloed, roles in organizations, you know, you have your policy, like you don't have like a cybersecurity's week over here because you've been prioritizing tech way too quickly and your website loads in like 10 minutes, but you're trying to adopt AI to replace all your customer service things. Like if you have all these mismanagements within organizations, which we do, there's no, there's no way organizations have figured out the data problem. Otherwise, it would all be as big as NVIDIA, is basically, right? They'd all already like, yeah, I'm talking about even Fortune 500, but. figured out things that we haven't figured out yet. But like even even the likes of Disney are not perfect companies, right? And and there's nothing wrong with what they do, but they have to go through their restructuring process. They have to frequently ⁓ course threck, right? So let's just compound that, right? So let's say that they're you're trying to use AI. I'm not saying I I know what these companies are using it for, but I've seen it in some instances where they are they kind of think they have all the data right. But that's more about like and they put all these band aids and then give it to AI and hope that it just again Filters it all out, figures it all out, comes back with something clean and sanitized. But again, it's like it doesn't know the value of your business, it doesn't know the history, it doesn't know every client intimately. It's like it's a very, very nuanced process that if you just don't have good data, you're giving AI data that it it will not only learn from the first time, it will retrain on that initial bad data. So even if you start giving it good stuff, it's muddy it's like mud in the water, right? It's it's like an oil spill, right? Where it's like it's it's not automatically taken care of. So I think it's like the biggest thing about AI like fundamentally why AI projects fail is data. Like I'm just gonna like go out and say like that's the new oil we call that, right? I mean, I'm trying to figure out the analogy here, but we but that is what they call it. And let's just say it's like there's a there's a very good opportunity for that data to be used properly as fuel to to find out some of our problems and and have AI really like, you know, become more autonomous, right? But if we're giving it a very, very bad foundation, it's like building a building, right? You don't want bad foundation. Like that's that's the kind of the whole thing, or like a car. You don't want it to have like a weak foundation. All these cool features don't matter if it's gonna like fall apart in the middle of the highway or something crazy like that, right? Like we have to make sure that these things are are tried and true. And I really think it all like it just it always goes back down to the kind of data that people are giving it, the training data, the data that, you know, ⁓ the everyday changes that a massive corporation goes through. It's it is I mean, it's not just that it's the data, but it's like the fact that that's what's helping decision making and what's driving actual value creation. That's why I'm going all back to I those are also reasons that the AI projects fail. But if you really, really drill down, and I don't know if your next questions or whatever, I can always go into it more, but like it data's like the word, like bad data basically is like the the keys to the kingdom to consider. And I think we kinda skipped a few steps industry wide. So I'll stop there, but that's my like big answer.
speaker-1: Yeah, and I and I was gonna say two things come to my mind. Find me a company that doesn't have bad data because how long how long have you been dealing with data that's been thoroughly cleaned? Like yeah. Yeah. And and two, going back to your to your point when you started there, all these people claiming all these different things, okay, show us then. Because if you're able to do that, you should have such a massive amount of money sitting here from a consultancy firm. That you can spin right up and start teaching everybody all those things and that's not happening. So it kind of tells us what, you know, what is really going on, which is all the other things you pointed out about how the models are kind of getting worse. Yeah. Slipping into bad data. So Yep. From that, you've talked about understanding how a business actually works first. Why would you say maybe clarity is more important than the technology itself when you're going into those situations?
speaker-0: I always look at it like, I mean, at the end of the day, technology is a tool, right? Like to to some and I think there might be certain exceptions to the rule, like the internet and the smartphone, which are like really like they they started as a tool, but they really became kind of like a major like phenomena, right? And and AI has like a weird it's it's a hybrid thing 'cause it is gonna change how we consume things. It already is. And it is it is a way to it it changes the way we it's consume media, information, knowledge, understanding of the world. Like there is a lot of a lot of important things there, but Clarity, at the end of the day, like I'm talking from perspective of someone who's like a business person, right? I've I've been running startups and I've I've I've helped companies raise money. I've sold a company before. Like I've done sort of what I would call like a good a good amount of work. I've been a serial entrepreneur. I've I've had a few times to try, I've had my failures and successes, right? And so I'm looking at it that perspective. But people everyone has their own definition of success individually, right? If you're using AI to help you lose weight, help you travel more, whatever, save more, like That's a different sort of thing. And I'm not going to get into all that because it's a whole different kind of conversation about how much it can actually help you, although bad data in from the internet, bad data out, right? Like story there too. But looking from business, specifically from business and sort of where I'm kind of going to come from here, is that the point of a business hasn't changed, right? You're not necessarily a charity, right? It's it's not necessarily supposed to be easy. Otherwise, everyone would do it, right? It it there is seasonality. There are things that go up and down with the markets. It is good to niche yourself, but niches die and And evolve, right? Like real estate isn't the same as it used to be, even though it seems like a sure thing. So all of these things are like they require the human element to adapt to, right? You to understand where people are going, consumer trends, industry trends, the the incumbents who are your partners back then might not exist anymore, right? Now you have to, you know, align with a whole new group of people, right? And find new ways to market it, right? Like we had media before for marketing, now we have social media. Advertising used to be more Print. Now it's a, you know, actually it's funny it's coming back to TV with like YouTube TV, but that's the point. Like even when it's come back, it's evolved into new distribution channels and other things too. And so the way to look at it is again, like technology is at a fundamental ⁓ it's a tool. And you still have to find like that clarity in you have to go back to what a business is, right? Driving, you know, the right demand and and coming off the right supply. And like there's like that's not, I don't think that's going anywhere, right? We still that's still very much the way that. That the world's talking about business and that's still the way that it's known and and that's just the way that it's gonna be. And I think once you start looking at kind of like it's almost like you have to be if if I'm a founder of a large company right now, right? And I'm still the CEO or I'm on the board or whatever, it's not so much about just adopting the next tool. It's about re-imagining with these tools in in tow and in in the loop, right? It's not just that the tool is gonna solve that problem, which I think is the expectation that people are frequently and and what people are getting sold. It's that I actually have to really come up with, okay, how am I Like how has the industry changed? What is like the signal versus noise? What are where are the consumer trends at? Where's the value at? That's that clarity. And then yeah, the tool is supposed to help you execute on that faster, drive, you know, repeat customers. There's always a goal, right? Reduce churn and it's cheaper to keep your current customers than buy new ones, right? Like I totally get it. And that's another that's another like business fundamental. Like you don't have to be an MVA to understand that's like that is kind of the the point of all this. And so once you kind of look at those fundamental things, once you look at like the reason you even have this business running in the first place. The reason it's in the niche that it's in, like then you use that tool to augment. And I think ⁓ a lot of people are just using it backwards, right? Where it's like, Claude, I work in travel, right? And I want to become the biggest travel company in the world. Of course it's gonna give you an answer. Of course. And the answer might sound inter incredibly intelligent, might sound like talking to a Harvard professor, like no doubt, no doubt. But like that's that's you over relying on like it there's just a there's almost like a psychosis, I is like the word that is in psychology, right? Where you're just like tricking yourself into thinking there's value. But I think most of the other people around you are gonna be like, what are you doing, man? This is like saying buzzwords that you're just it sounds like you're literally repeating what AI is telling you to do, right? Which is not, you know, that it gives great advice, it gives bad advice, but only we know as sort of business leaders and owners and other things, how to really like we need to understand the humanity and understand how people purchase, not just what AI thinks is the way people purchase and how markets move. We can see this with our own eyes, right? AI can help us like bring that information to us, but that actual action is very, very abstract. So yeah, that's sort of my fun, little bit philosophical answer. But I guess that's kind of the difference these days. Like I think if I if I came up with an answer that sounded too technical, it might even sound like I came up with a chat GPT. So I'm trying to I'm trying to be holistic, you know.
speaker-1: Yeah. Yeah. And the one part of the quote that I want to call out there is that yeah, businesses still want us to make money. And I go, Hmm, are they? With how many, you know, billions of dollars they're throwing in the trash? It's like they forgot how to be a business once AI entered the room and they walked out of the room with their new best friend and left all their other friends in the other room. So you kinda gotta get back to that and then, you know, like you said, it it's a tool at the end of the day. If you can justify the tool and the return on value equals plus, you're still learning. If it equals minus, what are you doing?
speaker-0: Yeah. And a lot of it is equaling minus right now because it's like, ⁓ I'll take the risk before the reward. But which is fine. Like I actually don't I'm I'm a startup guy. So I'm not gonna say no, but how long are you gonna be in the red for? Right? Like what are we doing if you're if you're being if some company's telling you have to be in the red for four years, that's ridiculous. That's like that's literally against there's why are you why are you d investing in a lost cause, right? So that's constantly bringing you down just because it has the word AI and it's because this tool has AI in it, like You gotta move quickly. You gotta learn how to outgrow these kind things. And and that's kind of the the nature of the industry right now. But people are, I mean, sucker born every minute, right? Sales salespeople are ⁓ it's funny because we've always talked about like replacing salespeople, but some of the top sales leaders are still people selling to people. Like that is still that's another fundamental business. There's still very much a lot of that. And AI is not changing that overnight. It might maybe in the future there's might be different things, but we're not there yet. Read the room, right? Much less you will actually earn more. Like I I tell the stars all the time, like you're actually hurting yourself by like having an AI chat bot when you actually need to be a startup that goes to the office of an enterprise and begs on your hands and knees to like have them do a pilot with you, right? Like it's that you don't just send out your little army. Like you can send out millions of cold emails. I'm not impressed if you don't have one customer, right? So I don't know. Maybe I'm too harsh, but I I mentor startups and I I become a bit more of like a feels like Rocky sometimes, you know. Like I have to like it used to be a bit of a industry where like I get it, like there were all these MA. We had a happy golden era of SaaS and MA and all those other things, but things are tougher now. And if you're not gonna admit that, then you're in denial and you're hurting your own business and your and your team, right? So that's another tangent, but just selling like that's another startup versus enterprise thing too to
speaker-1: Yeah. Yeah, yeah, I have a feeling we'll have many different conversations because my my area of ⁓ and A has already came and went, so and I think with that is it's a great time to go ahead and hear from our sponsor. Hello everyone. Ryan here from Stellarhorn. Let me take a quick second to talk to you about something I see all the time. Most businesses don't actually have a technology problem. They have a leadership problem. They've got systems, tools, even teams in place.
speaker-0: Yeah, yeah, of course.
speaker-1: But no clear direction, no strategy, and no one connecting technology to the business. That's exactly why I built Stellarhorn Group. I help organizations through fractional CIO leadership, advisory, and project based work, bringing executive level technology leadership without the full time CIO cost. Whether it's strategy, cybersecurity, AI governance, or leading critical initiatives, it's about helping you make better decisions, reduce risk, and actually move forward with confidence. If that sounds like something your organization is dealing with, please reach out. Let's have a conversation. Stellarhorn Group, rare leadership, real results, technology leadership without the full time CIO cost. And with that, let's go ahead and take a look into how does product thinking change AI adoption?
speaker-0: Good one. Yeah. Yeah. I love my my product. It's funny because I I've been like my product background is always a an ever existing current, but it's like it's evolved so much, right? Because of the way it's gone. Product thinking, what I love about AI in particular, and and why I didn't like I didn't I didn't, you know, fault vibe coding too much is because there is a world now where you can, as a product leader, build what you want to build in a in a very nice MVP or even like the design state, right? Using AI and and If it's an internal tool you're trying to build or something like I don't recommend using, you know, whatever off-the-shelf lovable templates and and and quick one-shot generated things as live products, but like you have a way to work with an engineering team and with a design team and say, okay, here is my framework. And here's like it's not just a Figma prototype, right? It's not just one of those things where you click and it's like a fake. You can do a whole experience, but then you're holding your team accountable to help you really come up with the value of it, right? Okay, how do we, how do we, you know, incorporate the UX and the onboarding and What we have here? Right. Because that's that's what AI I think is still very much missing. Right. And so I think the product thinking there is basically I can actually build a version of what I want to get at as a product owner, what I want to add. And my team, it's not that, ⁓ I'm gonna replace my team overnight. I'm done. They're gone. Right. It's that my team's gonna help me. They're gonna hold me and the company accountable to make sure the security practices are there, right? The compliance for whatever industry I'm in and the and the interface and the way the data's moving, right? It's it's being able to basically create these products and prototypes on the fly. To be able to unify ⁓ a w like an experience that everyone can test and use and then come back with like real feedback. Versus again, like engineers, I've I've heard like they weren't always fans of sifting through Figma and designers weren't always a fans, fans of like having to set up a local machine and set up a website or a web app or something, right? I think this idea that we can have, you know, these bive coded environments we're all working together collaboratively is a very awesome opportunity and can allow people to perceive things in different ways, but with again with a unified language of like this is actually a usable. product, even if it doesn't have like a back end or something, it is something you can go down a few steps in and we can we can leverage that and use that together. So it puts product people in a position of execution, right? Not just not just product thinking, right? It's product building now again. And and I've always been a technical product manager. So I've naturally gravitated. I've already been, you know, it was maybe not the best use of my time back then, admittedly a few times to build a website on my own, but it also was because I was an engineer. So maybe it wasn't all that bad, right? I could actually speak the language. But non non technical product managers were at a bit of a disability sometimes, right? In in more technical companies or even being able to get into it. And I think now it's not even about being a product manager as much as it is being like a product executor. Right. It's like, okay, now I'm going to connect, you know, the the engineering team and design teams to the sales team. But again, the sales team can come in and use ⁓ this environment, these vibe coded environments, these vibe coded templates and things like we can just come up with this faster and I can adopt the language to different like, you adopt the product rather in different languages to different departments. So there's the marketing, what marketing needs to know, what sales needs to know. And I think it's good, you know, again, I think engineers and designers won't always have that eye for business the same way some business like product owners do, right? In the way that they see the business and the intersection. And so they can be the translator in a way that I think AI still can't quite can't for the same reason of that clarity and that ⁓ you know, abstraction layer, right? That that is the humanity part of it all.
speaker-1: Yeah, no, very Barry right. ⁓ you know, one one of my closest allies and a friend is a is a product manager. So I learn things from him all the time. I'm fascinated with what draws people in and why you would change a button to this color or this, you know, shape or or whatever. And it's always fascinating to me the the psychology behind it really.
speaker-0: Exactly.
speaker-1: what would you say is the fastest way to actually see value out of AI in a business?
speaker-0: Genuinely, like as as as funny as this answer might sound to some folks, and you might get it better, Ryan, because you you are also sort of business owner, right? But it's like automating my own tasks as a leader has been incredible. I mean, I use I use a tool called Motion, which is awesome for product project management. It's kind of replaced like a lot of different tools for me. And I kind of brought them into all in one for my company. And like we're we're not a huge team. We're kind of one of those like, what do you call them? Like the very tactical, like, you know, ten people, but we do the work of like thirty when we that's just the nature of smaller companies and And I'm very proud of my the way that my team is using AI to make that happen, right? But we, you know, we we don't like I hate the idea of having my team already siloed out at that small of a level. So we have tools that literally ⁓ we have Cloud Max, we have motion, we have all these other things as well, you know, Miro and and there's there's other like Slack that still use that too. But I've I've reduced the amount of tools but focused it even on like what I use it for day to day, right? Like what I the the kind of things I need to remember, the way that the tasks surface themselves, like. The tool that I use, like AI helps automate like the calendar for me and make sure it prioritize things for me. Like it's fine to have that. And I think the quickest way to see value is like to give it to leaders and engineers, frank. Like it's to give, it's to give for like the menial tasks, right? So you know, you don't have your you don't have to have your team vibe coding the whole product, but like they should definitely be writing unit tests with AI. Why are you handwriting unit tests and regression tests? Like it's so much and it's still up to you to be accountable and to check these things out. But yeah, I would say it's like it's the menial tasks. That are the absolute fastest. You definitely see an increase in productivity day one. And it's incredible, right? It is for I have things for my email, I have things for my daily reporting and analytics. Like I'm very happy to have those. And it's actually nice. Like they used to require more setup, but now all those tools have their own AI enhancements. And they I can have text messages. I can use Zapier. Like, I think those little elements and investments from NSV are the same thing. Startups, mid market, enterprise, the cities I'm working with, the universities I've worked with, everyone that I can think of is utilizing these, you know. sort of self-helping AI tools for the menial stuff. Like everyone is seeing value there, right? It's overtaking the bigger swings where that clarity becomes a lot more important. But on a day-to-day basis, why do I have to sit there and, you know, organize my calendar repeatedly, like 10 times a day, right? Now why not have AI? I'm a busy schedule. I'm out and about. I'm at events, meetings, podcasts. Like I'm happy you can help me do like that one little thing, right? ⁓ and that's instant ROI. Instant ROI. So yeah. Using as your copilot. Nothing wrong with that. Nothing wrong with that
speaker-1: That. Yeah, no, I'm like minded where that is the most return on value. The more me time that gets freed up to actually work on something or actually produce that value for a client or whatever it is I'm doing, fantastic. Get rid of all that that that time sink stuff that is still incredibly important, but you don't need to do it. That's useful. So what is one piece of advice For leaders trying to implement AI that you could provide.
speaker-0: That's gonna easy answer. Like look at the data within your company, right? Not not just and look at like the way that AI do your research on how AI is taking in that data, right? Especially if you're if you're using, you know, Claude at its most API level, right? And you there's differences in using it between an enterprise plan and not. But in general, like just look at how AI is evolving and look at your data and start to look at the ways that that how that could like what that outcome could look like, right? Try to for like work with your team, work with people around you to figure out like how what the value that you're trying to drive and look at how what's happening to other companies and what people are seeing and what kind of benchmarks. There's people using AI all the time who are coming up with great reports that that t that show you like here's how it's doing math, here's how it's doing data sanitation. And I think it's up to leaders right now to really understand the limitations. Don't just look for the positives, look for the because there are plenty of positives, but look for some of those negatives on purpose so that you can understand, okay, this is the data I need to give it. This is what we need to do with people and with other tools and other resources to make sure that we're giving it the data that actually produces something useful, versus just like here, take all my documents, scram that in there, and just ship me off the nicest slide deck you can, right? Or the nicest ⁓ product feature that you can and ideas. So that's like one part. It's like the data and also just like I was saying earlier too, go back to the beginning a little bit, right? What is this business's vision, values, principles, core value that makes customers say, I want this business and nothing else. Is it the relationships? Is it the trust? Is it the volume? Like the ⁓ kind of flip side of relationships, the commoditization of it. Look at all those things and understand the limitations of that and what you want to be too. But don't AI is not the it can be part of the evolution, no doubt, but it shouldn't be the answer to the evolution. It shouldn't just be like that. All right, ⁓ Claude evolved my company to stage, you know, five hundred of its lifetime, right? If it's if it's hundreds of years of existence or something crazy, depending on how old the company is, of course. ⁓ I think that clarity is just is way too important not to not to mention repeatedly. Same thing with the day with the data problem, right? There's all these little things, but they all go back to this problem fundamentally, these two sort of big items fundamentally.
speaker-1: Yeah. And I I think just ⁓ to leap onto that, you know, have honest conversations two way, honest conversations with your IT team to say, where are we from this standpoint? And if the answer is what I suspect it would be, which is we need time and we need to do a project before we can do this project in order to actually have better data go into that, then that's where you're at. And that's probably where a lot of people are at. So you're not really behind and if you jam like we've been talking about, bad data in. it's not really gonna get you anywhere and you have to be receptive to that as a CEO and I'm sure that's hard talking to boards and, you know, investment cycles, all of that good stuff. But that's the actual answer because if you do that cleanup, probably have a chance at more success, maybe even additional projects. Who knows?
speaker-0: Yeah. And as someone who's worked with executives, like there is a way to, there's a narrative for sure. I get I get how large companies work. And even startups too. Like I know there's hubris and ego and all these other things. Like I understand it's very, very hard to like seem like you're taking a step back to step forward, but it's also like you you could lose a lot of money and then no one is happy. Right. Like there is there is a way to find a middle ground. And I think each company is very unique. There's not and there's not I that's why I'm in business, right? There isn't an easy answer that I come in with, but there's frameworks to use. There's understanding of the human, like I do a lot of ⁓ you know, I I use this principle very frequently called most advanced yet accessible, which is I or yet acceptable rather, but it's also accessible too, where it's like, you know, the I like you could come up with a flying car tomorrow, right? You could literally present the flying car to the world, but we have no infrastructure and a lot of mistrust in the idea of just flying around and there's planes in the sky that are, you know, having issues and stuff, right? Like we we have way more to go before we have that become a normal thing. And so you have to start with like other ideas first and other sort of evolutions, same way you would building a startup, right? Your MVP. It shouldn't be that you're like necessarily building a car in pieces. It's like the scooter, then the bicycle, then the motorcycle, then the car, right? There's like a famous diagram. And that's what this Maya thinking is. And it's it's it's like the secret to a lot of brands that honestly seem unfazed by this AI, boom, not because they're not driving value or anything with AI, it's just because they are using it to always do what they do best, right? not trying to reinvent themselves or or or just do things just to be like, you know, ⁓ crazy overnight. I mean, it's in fact with everybody, but there are companies that are surprisingly like bulletproof in this, but I really think it's because they have they understand the movement. They're not trying to rush anything. They're not trying to just be in the press and the media all the time by doing these these crazy leaps, right? Like I said, it looks a lot worse if you do a leap and then you have to go back on it. Would not want to be Disney right now with the Sora and Epic Game situation, right? Like very hard to look at that and be like, ooh. That's like one of the biggest entertainment companies who made two major mistakes in AI, in believing in AI recently like a week, in a span of a week. I know they'll do fine. I know they'll do fine, but it's like it's not a great look. And so we have to kinda look at these things, right? Is that the reputation you wanna have? Yeah. Something something that I think resonates with CEOs for sure.
speaker-1: Yeah, and I'm just gonna put this on a limb. I I think it's pretty accurate. Not every company is Disney and has Disney money. So maybe maybe think about that investment a little a little closer. Maybe ⁓ actually pay attention to what your own teams are telling you and have that communication early. Have it now. Yeah. Like ha like listen to this, go have that combo. At the end of the day. Slowly.
speaker-0: Absolutely. It's a big secret to a lot of things. Good communication, right? Like definitely true.
speaker-1: It goes back to basics in a lot of ways that we seem to just run past for some reason.
speaker-0: Hey, you know, AI AI kinda had ⁓ it's like I said, there's a bit of a psychosis with it, right? Like you really get into it and down the rabbit hole and you kinda forget, We we got time. We got time. There's there's there's better better wins for all of us if we take our time correctly and and marathon not a sprint, right?
speaker-1: Absolutely. So we're at the end of our meeting here. So where can people actually connect with you if they want to get in touch, hire you, meet you, have you on future podcasts?
speaker-0: Sure. Yeah. LinkedIn is the best place to reach me at the moment. I I try to look at like I I see all the connections come through. And then I also ⁓ I I'm at nvall dot com is my website and I have like ways to contact me and sign up for my newsletter, which I talk about ⁓ more leadership stuff than but it also kind of links to some of the AI stuff. ⁓ but yeah, those two places are definitely the best. I have other plenty of other social media channels, but the spam that seems very AI driven is so bad on some of them I can't I can't r ⁓ promise I'll be responding. But luckily on LinkedIn I have a I have a good filter system now. So yeah. Yeah, that'll be the best way.
speaker-1: Wonderful. So with that, the closing thoughts are, you know, AI doesn't fix broken businesses, it exposes them. And the companies that understand how they operate are the ones that actually get results. ⁓ and with that, our meeting's been adjourned. And I so appreciate you joining the podcast today, Nikhil. And I imagine we're gonna be in a bit more of a communication going forward now that we've met one another. And ⁓ I hope that the viewers out there have taken some value because you said a lot of valuable things today. They should take car.
speaker-0: No, it's it's been a it's been an absolute honor. And I'm I'm glad like I said, like I I think before this ⁓ conversation even started, like we had a another podcast together. It was a good like taste of what both of us are good at and how we think. And now we've been able to expand on that. And and and you brought up some really great points too that I think like even if we had two, three hours, we could have just kept going, right? Could have kept the could have kept the tape rolling and just talked all day 'cause I've I have a lot of perspectives on on where like the CIOs might sit in IT people, a lot of people that I tend to work with. ⁓ there's a whole new world opening up there too, which AI is helping and also Diluting as well, right? So see, I almost almost went off on a different topic there alone. So, anyways, yeah. No, it's it's been a pleasure today, Brian. I appreciate it.
speaker-1: Absolutely. With that, ⁓ that so I screwed up about that at that. So with that, that takes us to the end. So take care everyone and see you in the next one.
speaker-0: Yeah, no, thanks for having me, Ryan. And I'm excited to sort of dig into the the topic we have at hand today a little bit more. So yeah, it's an honor. I'm I'm happy I accepted this invite, right?
speaker-1: That's a question for later. And we're not done yet. So first thing we like to do is get into some icebreakers. So what is one AI or product buzzword that you'd retire immediately?
speaker-0: One changes like every week. ⁓ I because it's tough, right? I mean, I gosh, like I've been pretty anti-generative as a word anymore because I think it I think it actually kind of degrades to like when we're just thinking about producing and not, you know, there's value in regurgitating and there's a value in enhancing and augmenting. There's other words to use, right? And generative almost seems a bit derivative where it's almost like you're just kind of making something out of nothing or out of some minor input and it's supposed to, you know. be this like end all be all silver bullet kind of thing. But I think it's just more nuanced. And it o it always has been. I've been in like prior to generative AI being like the thing that shifted it. It was conversational AI. Machine learning is what I used to work in. It wasn't really it was AI, but not like that wasn't the word we used every day, right? I think it's actually kind of starting to rotate back into that. And I think that's a good thing. I think it I think it sets an expectation. I think it's a little bit more scientific to look at things in their nuance and and maybe find better words than just generative, right? I don't know. Like that's that's just not as strong as it Maybe people had thought it was. So yeah. I don't like the vibe word either. I don't know why. That's not I that is the only way to describe it. But I think there is a better again, there's a better ways to I'm hoping the word augmentation comes into the poll. I think that's a good one. But it's not it hasn't taken off yet. So yeah. Anyways, that's my that's my answer.
speaker-1: It's a great perspective. It's the first one that I've heard and I think it's great from from your point of view because you're actually in it working every day, which we'll get to. What would be the invite you declined now that you didn't earlier in your career? Man.
speaker-0: And do you mean across like any opportunity or anything like that, or do you mean a specific type of in
speaker-1: Like no, no, just yeah. So something that somebody asks you to do and you're like, I no longer I no longer want to participate in that because I've learned from that approach.
speaker-0: Man. I mean, that's kind of the secret of my whole career is out is when I knew how to outgrow things, right? So there's there's like I I'll think of like one one good example I think was, you know, early on in yeah, like earlier in AI, I think I didn't believe in it as much as I maybe should have been an interesting way, because there are people who are like, we always talk about niching, right? And I think niching is good industry wise sometimes, right? Like don't like just I mean, you can't possibly build something for everyone. But I think at that time I had sort of like I'd I got really into conversational AI, which is very good. And So it's a very strong community. And it is it's everything is conversational. Yeah, I know, basically. But there's like a even further in the customer support world. But I think I wish that maybe I had expanded my AI sites beyond customer experience like 10 years ago. Cause I think I would have just been like there would have been a little bit more like speed in my career to to look at some of the other parts of it too. And that's why even though I was talking about the word generative, like I'm not a fan of of digging down like with that, there's nuance and there's interesting ways to do it. And we should be allowed to have room to experiment, right? If we if we have the privilege too. And so ⁓ yeah, I think I think that was the only thing is I almost like discounted how broad AI can be because maybe and at that time it could have even just been like imposter syndrome, right? I was just, you looking back, I was okay, well, conversational AI at that time, Alexa and chatbots, that was like the cool hype thing. But there was also there's always been a lot of really cool stuff in machine learning and I've had to catch up over it over the years, but I'm I'm happy because then I was ready for this AI boom better than than that last sort of like Sassin AI ⁓ boom that had been again, maybe a decade ago. So yeah.
speaker-1: Yeah, you I mean you bring up Alexa. I mean I remember playing around with Lambda ⁓ at some point way early and I am so much further away from the area of AI that that you're in. But that was like my first foray into playing around with it and I was so proud of myself for getting simple commands even though cool though there.
speaker-0: Yeah. I mean it was you you felt proud doing it though, right? 'Cause it was it was like the early stages of what we have now, which is just accessible AI, but it felt good that you could you could do something. You didn't have to be like, you know, a a a rocket scientist to to figure it out.
speaker-1: Yeah, like whatever room name you pick, right? I wanna program it with this, turn this on or off, and it's like awesome, it works.
speaker-0: Internet of things, right? Yeah. Good old days. Good old days.
speaker-1: So what's something companies think AI is solving that it isn't? ⁓
speaker-0: Yeah. Another ⁓ like I I always kind of crack up because it's like these are these are just great questions to start off with and and do set the tone, right? So ⁓ it's it is I I'm very much against the let's like for everybody and then rehire after AI has been solving whatever it solves and whatever it doesn't. I do understand 'cause I mean again, as someone who's worked for five hundreds, I do understand the notion and I understand why they kind of do it for my, you know. like traditional MBA business standpoint, like there is some, you know, I it's it's quick win, right? It's quick win and it shows market signals and all these kind of things. But I think a lot of companies and not just Fortune 500s are in for a rude awakening when they realize that like they could open claw and they could automate and have a thousand agents who still can't solve the value problem. Right. Like how do you how do you actually make buyers click purchase? How do you make customers sign contracts? Like you can't have AI for somebody's hand. We're not there yet, at least, right? I mean, there's there's a world where AI is talking to AI, but the biggest, and this is like the biggest shock. And it's still this day, I've been in conversation with AI now for 10 years, right? And ⁓ generative AI since it was, you know, whatever introduced. And I still see an overwhelming amount in any big company that I work with or I work with like e-commerce, whatever, the whole thing where it's like, I want to talk to a human, even if AI solves the problem. Presents the right FAQ, enriches the right answer, schedules the refund. I still frequently, frequently see people in different age groups, probably a little bit on the older side, but that's still a pretty big buying power audience, saying, I want to speak to a human. I want to really make sure my refund is coming. I myself, like I've used automated services and the, you refund never came or the product never got returned properly until I talked to a person who confirmed it, right? The AI didn't come back to me, wasn't proactive, it didn't email me a confirmation. Like these little things are missed way too often for comfort. And that's a small example, but when you look at it on a grand scheme of things, I mean we're we do see efficiency gains. I totally agree. But the solution isn't just to like start, you know, firing left and right and introducing new departments for the say I I say a lot of people recklessly introducing like R and D departments and letting go of all the IT team, right? Or or replacing a C D T O with a C A I O. But why like that's not it's not just a signal thing. It's noise. It's becoming like they think it's a signal for the industry, they think it's a a competitive advantage, but they're really just making the noise more noisy in the in the industry. And usually ⁓ I think the number is still, I think it's gone down a little bit, but like a year or two ago it was 90% of AI pilots are failing. I think it's still up there. I think it's still 82% or something. So there are more pilots that maybe going to phase two, but there's still a lot of enterprise opportunities which are are being completely ⁓ messed up because you can't have like AI build its own good pilot if it doesn't know how to like v establish that value. Right. So I could go on a tangent all day, but I will say that one is That one is very tough and usually with the companies I work with, it's a lot more holistic than just let's clear house and start over, right? Like that's that's not usually the way to go here. It never has been in business, to be honest.
speaker-1: And like I I just read an article this morning that was just like CEOs throwing in AI projects just to say they're trying to do AI projects and don't even align to the business goals or or anything that they want to do. It's like what are we doing? Where how did we get to here? It's like AI got marketably big, we'll call it, you know, marketable. ⁓ yeah. And ⁓ yeah. Then we lost our minds of how to do business. So
speaker-0: Right. Which kinda like defeats the point, right? Because like the ⁓ per purpose of business to buy and sell product and then like you know, create supply and demand. Like there's all these fundamental principles which AI like definitely doesn't understand. Otherwise, we would all be saying, Make me a billion dollars and I'll have a billion dollars, which is not the case, right? So but like funny how you like when you really dig down to the philosophy and holistics of it. Yeah.
speaker-1: Yeah, absolutely, absolutely. So transitioning a little bit, let's talk a little bit about your background. So what pulled you into AI and
speaker-0: A fun a fun ⁓ opportunity actually because I used to be a self-taught like software engineer and and full stack engineer back in the day. I got into the tech industry around web two point ⁓ so AI definitely wasn't the top of mind. But like at that time, Whites web apps and apps were were just hit blowing up for the first time, right? And ⁓ it wasn't really about the to be honest, it's funny because now it's all about the libraries and tools that we're using, but back then we had very little to begin with. G GitHub was a new site when I got started, right? Which is crazy thing to think about. And all these like Code Academy and all these ways YouTube university, like all that was a newer thing. And so one of the philosophies, I still champion it today, but one thing that I really got into that helped me learn was the world of open source. Like I've actually personally had the opportunity to contribute to programming languages and actually like Firefox, for example, and other places. And I learned how to code better through mentors and and other opportunities found in that community. So I'm just giving that a nice shout out there. But through open source as well, there's a little company, which is not so little anymore, but Hugging face used to be a side project essentially before it became I I think they're a unicorn now or they're or nearing it at least. But they used to be a website where people just shared code. Like it was just like you were sharing like some machine learning stuff, some internet of things. It was a very, very basic site and it was like an it was a layer on top of GitHub. And anyways, I like randomly ran into somebody's like self made ⁓ I think it was like an Alexa machine learning model or something. So to help like Alexa get even smarter or something, you could like hack the voice AI side. It was really cool for twenty sixteen. Right. That was like very, very, very cool for that time. And I just started tinkering around. Right. And suddenly I discovered ⁓ their companies that got rolled up into like Microsoft and NVIDIA now, but they were literally just startups building libraries, right? And open sourcing their code. And the the value was in customer support, or you could buy the enterprise version, but they were building the early stages of voice AI, what we call generative AI, research, machine learning. Principles, voice recognition and voice generation. I've been doing that like a lot longer than than it's been around, than 11 labs have been around for, right? And it wasn't really like they they're the incumbent now, and then it's great to see their growth, but there were other companies trying similar things. And and yeah, it just like it was such a great, it was a great rabbit hole to go down because obviously it was like I was able to look at the foundational code even before using a cloud system, right? I was able to like integrate it on my own websites and play with it and and mess around with it as an engineer. And ⁓ you know, kind of pivoted and then the business mind game, right? It's like, okay, we have all these tools that are for free. How can we, you know, in open source you can quote unquote fork a library and ⁓ you know bring it to an enterprise without having to pay for like a license, right? Like there's or you have to follow some license, but you can you can do things very differently if you if you know how to utilize open source well. And then it's not so much about the code, right? Again, it's it's about how you utilize it, how you modify it and how you turn it into intellectual property, which I was I was doing back then, which was really cool. So yeah, I like to I like to champion open source still and I I give it a lot of credit in ⁓ A its role in AI's growth, but also of course my growth specifically.
speaker-1: Yeah. That that is one hundred percent amazing. ⁓ I've always been on more of the open source side, ⁓ than any kind of closed resource. And maybe I'm showing my age as well with that. But yeah, absolutely there. Android fan through and through for a while, ⁓ on my side. So you've worked across startups and enterprises. What's different between
speaker-0: Startups want to be enterprises, enterprise want to be startups. And it's the funniest thing because again, if we some of the best advice I ever got in the startup world, which has led to sort of where I am now and led to a lot of, you know, what I would say like my my more successes in my career is to get out of the startup mindset once you're not a startup anymore, which means like, you know, the first year, it's fine. Flail a little bit, take your time, you know, f pivot, like all these kinds of things. But if you're doing it three years in, there's something's gotta change, right? You gotta start outgrowing your, your, ⁓ you know, your mindset there, right? And it's interesting because I've also worked with a lot of enterprises who bring me in as like someone who wants to build a startup division, you know, like a an R and D department that runs like a startup, but it doesn't work that way. Right. You're not supposed to with all those resources in place, you're not supposed to, you know, starve a a part of your company that could have all this opportunity and could speed ahead of the competitors who're supposed to properly and again that's that goes into my the crux of my work, right? Just helping people see that that's not really what they're looking for, even if they kind of have like a preconceived notion. Cause then the startup world is not as it's like that's a 90% failure rate, right? Why are we trying to emulate something that doesn't work most of the time due to infighting ownership? Like there's all these other things that happen. So I I find myself at the intersection of this a lot, which is kind of funny because I end up working with a lot of like really interesting like mid market size companies that are still doing like a lot of money, but they have they just don't have the same responsibilities and pressure as enterprise does. And then I've also had the chance to work with like series C, Series D startups where like, wow, we really wish like we could have gotten here faster. If we weren't treating ourselves like a startup at Series A, even right. If we were already not we could call ourselves a startup all day, but like in terms of the actual internal culture and the way that we're going, like we needed to grow up. And and it it helps to grow up, right? For for the people that you're selling to. So it's been interesting to see that AI sort of leveled that playing field in a few ways too, right? It's it's humbled people. Whereas like, okay, if you're using if you're using AI like a startup does at a large corporation, it actually will get you farther. But again, it it doesn't matter. How how quickly you build internal business tools or whatever if there'sn't the value, right? And then on the startup side, it's actually diluted the whole space because now everyone can start a startup. And the same problem is b we go by right back to the crux of it all is like, how do I actually make money? Right? How do I re how do I make a profit and not spend all my money on tokens and hosting, which is very possible if you have a really complex product? Like there's still margins. We still are doing the same business thing and Unfortunately, I can't say AI is necessarily getting cheaper every day because there's a compute problem. So it's fascinating though, right? It's like it it's been, you know, be being in this space for more than a decade now and being very, very in it. Like not just, you know, I didn't come from corporate. I was I started in startups and tech specifically. So seeing it evolve like this and seeing how enterprise, like as I work with them later in my career and and got to know them more, it's interesting to see how everything's kind of flip-flop. But it all kind of comes down to the same fundamental things that are missing. That are just being compounded if they're not solved, essentially, right? And enterprise and startups almost both have that in common beyond their differences.
speaker-1: Which I think is a actually a great segue that I'll go right into. What patterns have you seen when companies try to adopt AI?
speaker-0: Yeah, yeah. I mean, it's always speed. The the top pattern is like, let's be the fastest possible, which like look, nothing wrong with being fast. I move quickly in my projects and consultancy and I advise it. But you have to run at a pace, right? It's not like ⁓ it's a marathon, not a sprint. I I do genuinely believe that. And I think ⁓ I think we're actually seeing that happen with ChatGPT, who just closed down Sora and has had soaringly negative reviews, right? Where they were the fastest and they still are releasing stuff all the time, like no doubt. But they're getting humbled by a company that took its time, ⁓ built a company culture around the kind of AI that they're building, you know, like Chat GBT's kind being left in the dust. I hate to say it even in Grok, right? Like, which is, you know, sort of whatever your alignment might be in in Twitter and it's its dealings, right? Like it is a it is impressive how quickly that's growing too. Magnus over at Meta, like all these things are starting to catch up. Gemini still needs a little bit of time in the hopper, but you like nano banana, right? And all these things. So there is something to be said about like, yes, you might get laughed at a little bit. And that sucks. Like some of these people, I think, went through that Atropic and Google primarily. But now they're showing that like they were actually just being strategic. And and Apple blew a bit of a lead, right? I love I I think AI, whatever Apple tells, just has a chance to still be something, but it's not going to be so model, essentially, right? They're they're even letting people get into model switching. So, anyways, big pattern there. Like that's I'm just saying that even beyond those tech companies, that's a pattern you see everywhere, right? Like the people who are treating it like a marathon and not a sprint are the ones that are winning. I think back to that whole thing of like I think a lot of people are trying to act like they're trying to. change their engineers or their RD departments to turn it to startups, which I'll put it in like two different buckets, right? Positive side, you're being scrappy or using ⁓ company resources well, there's a good opportunity. Bad side is that your expectations are completely misaligned. And that's another pattern. Actually that segues into the, you know, sort of third pattern here, which is across everything, there's always very, very different expectations person to person, because we've all been given AI to use on our personal devices. We all have, you know, we have been using Alexis for a while. Like there's all these different things that have Diluted our perspective on how to actually use AI. Some people believe one-shotting prompts are good. Some people picking up prompts are good. And there's a again, it's not even an easy answer. There's like different AIs that require different things. Image generation requires different than text generation, document versus a quick response and research. Like you still have to have a bit of a sense of, you know, again, like what are you actually looking for? And if your expectations are just that you're gonna make more money with it and not that it's actually gonna drive value and and lasting value, you're not gonna get very far. Right. So it's like Kind of to yeah, to to bring that back down, it's like the it's the expectations, it's treating it like a startup or treating it like a little bit like over correcting on scrappiness and then like speed, misalignment of speed, good speed and bad speed, kind of as such, like good pace and and and sort of like rush jobs. Those are sort of the the main three pillars I would say. There's more nuanced stuff, but they all kind of fall under the same thing. But those are the three that kind of link together one major pattern, but breaking up into three easy ways to see. And this is across mid market enterprise startups, any non profits that I work with, civic I work with some cities on their AI adoption. Everyone is everyone is is having the same sort of issues, but with the right intentionality and clarity, ⁓ very different results, like in a positive manner.
speaker-1: Yeah, I'd say the the wide margin of result and just the same kind of ideas of like they wanna do good with it, but then when we get to the point of actually implementing and then deploying it's like what happened and why are we in such wildly different, you know, situations. It's it's interesting. It's it's a very interesting time. Yeah. Which is gonna lead to a great question based off what you just said, which is why do most AI projects actually fail?
speaker-0: Yeah. Yeah. The the title of the episode or whatever, right? Or or working title. I do think it comes down to it. I actually don't think like the speed is a problem, but like you can still you can still correct, right? I think it's it it goes way more and even like this team thing where it's like you're firing everybody and rapidly rehiring or trying to make turn a department to something else. Like I even see people correcting on that and and being able to stabilize, right? I do think it's an expectation thing, right? I think it's a lot of people, and I'm not just talking about people at companies, but about vendors and partners and even larger companies making some pretty extreme claims of you can do all this with little clarity, with little work and like, and sure, that's like very attractive, right? And it's a great sales technique to just say AI will do this for you. And I and it can do a lot for you. And I and I'm I'm totally, you know, like not objecting to that. But I will say, and I'll kind of I'll actually take it a step higher. to say that one of my favorite phrases, and I use this very frequently. I don't remember where I first heard it, but I use it all the time. It's good data in, good data out, bad data in, bad data out. And these days with with how fast AI is going, it's like good data in, better data out, bad data in worse data out, right? It's starting to, you know, ChatGPD's results results are definitely different than they used to be. And even Claude, right, if you're using a different version of it, it's not the same quality as it used to be because it's just, you know, the the way of the sort of movement at the the space, right? So if you have a data problem, If you already have things that are siloed, roles in organizations, you know, you have your policy, like you don't have like a cybersecurity's week over here because you've been prioritizing tech way too quickly and your website loads in like 10 minutes, but you're trying to adopt AI to replace all your customer service things. Like if you have all these mismanagements within organizations, which we do, there's no, there's no way organizations have figured out the data problem. Otherwise, it would all be as big as NVIDIA, is basically, right? They'd all already like, yeah, I'm talking about even Fortune 500, but. figured out things that we haven't figured out yet. But like even even the likes of Disney are not perfect companies, right? And and there's nothing wrong with what they do, but they have to go through their restructuring process. They have to frequently ⁓ course threck, right? So let's just compound that, right? So let's say that they're you're trying to use AI. I'm not saying I I know what these companies are using it for, but I've seen it in some instances where they are they kind of think they have all the data right. But that's more about like and they put all these band aids and then give it to AI and hope that it just again Filters it all out, figures it all out, comes back with something clean and sanitized. But again, it's like it doesn't know the value of your business, it doesn't know the history, it doesn't know every client intimately. It's like it's a very, very nuanced process that if you just don't have good data, you're giving AI data that it it will not only learn from the first time, it will retrain on that initial bad data. So even if you start giving it good stuff, it's muddy it's like mud in the water, right? It's it's like an oil spill, right? Where it's like it's it's not automatically taken care of. So I think it's like the biggest thing about AI like fundamentally why AI projects fail is data. Like I'm just gonna like go out and say like that's the new oil we call that, right? I mean, I'm trying to figure out the analogy here, but we but that is what they call it. And let's just say it's like there's a there's a very good opportunity for that data to be used properly as fuel to to find out some of our problems and and have AI really like, you know, become more autonomous, right? But if we're giving it a very, very bad foundation, it's like building a building, right? You don't want bad foundation. Like that's that's the kind of the whole thing, or like a car. You don't want it to have like a weak foundation. All these cool features don't matter if it's gonna like fall apart in the middle of the highway or something crazy like that, right? Like we have to make sure that these things are are tried and true. And I really think it all like it just it always goes back down to the kind of data that people are giving it, the training data, the data that, you know, ⁓ the everyday changes that a massive corporation goes through. It's it is I mean, it's not just that it's the data, but it's like the fact that that's what's helping decision making and what's driving actual value creation. That's why I'm going all back to I those are also reasons that the AI projects fail. But if you really, really drill down, and I don't know if your next questions or whatever, I can always go into it more, but like it data's like the word, like bad data basically is like the the keys to the kingdom to consider. And I think we kinda skipped a few steps industry wide. So I'll stop there, but that's my like big answer.
speaker-1: Yeah, and I and I was gonna say two things come to my mind. Find me a company that doesn't have bad data because how long how long have you been dealing with data that's been thoroughly cleaned? Like yeah. Yeah. And and two, going back to your to your point when you started there, all these people claiming all these different things, okay, show us then. Because if you're able to do that, you should have such a massive amount of money sitting here from a consultancy firm. That you can spin right up and start teaching everybody all those things and that's not happening. So it kind of tells us what, you know, what is really going on, which is all the other things you pointed out about how the models are kind of getting worse. Yeah. Slipping into bad data. So Yep. From that, you've talked about understanding how a business actually works first. Why would you say maybe clarity is more important than the technology itself when you're going into those situations?
speaker-0: I always look at it like, I mean, at the end of the day, technology is a tool, right? Like to to some and I think there might be certain exceptions to the rule, like the internet and the smartphone, which are like really like they they started as a tool, but they really became kind of like a major like phenomena, right? And and AI has like a weird it's it's a hybrid thing 'cause it is gonna change how we consume things. It already is. And it is it is a way to it it changes the way we it's consume media, information, knowledge, understanding of the world. Like there is a lot of a lot of important things there, but Clarity, at the end of the day, like I'm talking from perspective of someone who's like a business person, right? I've I've been running startups and I've I've I've helped companies raise money. I've sold a company before. Like I've done sort of what I would call like a good a good amount of work. I've been a serial entrepreneur. I've I've had a few times to try, I've had my failures and successes, right? And so I'm looking at it that perspective. But people everyone has their own definition of success individually, right? If you're using AI to help you lose weight, help you travel more, whatever, save more, like That's a different sort of thing. And I'm not going to get into all that because it's a whole different kind of conversation about how much it can actually help you, although bad data in from the internet, bad data out, right? Like story there too. But looking from business, specifically from business and sort of where I'm kind of going to come from here, is that the point of a business hasn't changed, right? You're not necessarily a charity, right? It's it's not necessarily supposed to be easy. Otherwise, everyone would do it, right? It it there is seasonality. There are things that go up and down with the markets. It is good to niche yourself, but niches die and And evolve, right? Like real estate isn't the same as it used to be, even though it seems like a sure thing. So all of these things are like they require the human element to adapt to, right? You to understand where people are going, consumer trends, industry trends, the the incumbents who are your partners back then might not exist anymore, right? Now you have to, you know, align with a whole new group of people, right? And find new ways to market it, right? Like we had media before for marketing, now we have social media. Advertising used to be more Print. Now it's a, you know, actually it's funny it's coming back to TV with like YouTube TV, but that's the point. Like even when it's come back, it's evolved into new distribution channels and other things too. And so the way to look at it is again, like technology is at a fundamental ⁓ it's a tool. And you still have to find like that clarity in you have to go back to what a business is, right? Driving, you know, the right demand and and coming off the right supply. And like there's like that's not, I don't think that's going anywhere, right? We still that's still very much the way that. That the world's talking about business and that's still the way that it's known and and that's just the way that it's gonna be. And I think once you start looking at kind of like it's almost like you have to be if if I'm a founder of a large company right now, right? And I'm still the CEO or I'm on the board or whatever, it's not so much about just adopting the next tool. It's about re-imagining with these tools in in tow and in in the loop, right? It's not just that the tool is gonna solve that problem, which I think is the expectation that people are frequently and and what people are getting sold. It's that I actually have to really come up with, okay, how am I Like how has the industry changed? What is like the signal versus noise? What are where are the consumer trends at? Where's the value at? That's that clarity. And then yeah, the tool is supposed to help you execute on that faster, drive, you know, repeat customers. There's always a goal, right? Reduce churn and it's cheaper to keep your current customers than buy new ones, right? Like I totally get it. And that's another that's another like business fundamental. Like you don't have to be an MVA to understand that's like that is kind of the the point of all this. And so once you kind of look at those fundamental things, once you look at like the reason you even have this business running in the first place. The reason it's in the niche that it's in, like then you use that tool to augment. And I think ⁓ a lot of people are just using it backwards, right? Where it's like, Claude, I work in travel, right? And I want to become the biggest travel company in the world. Of course it's gonna give you an answer. Of course. And the answer might sound inter incredibly intelligent, might sound like talking to a Harvard professor, like no doubt, no doubt. But like that's that's you over relying on like it there's just a there's almost like a psychosis, I is like the word that is in psychology, right? Where you're just like tricking yourself into thinking there's value. But I think most of the other people around you are gonna be like, what are you doing, man? This is like saying buzzwords that you're just it sounds like you're literally repeating what AI is telling you to do, right? Which is not, you know, that it gives great advice, it gives bad advice, but only we know as sort of business leaders and owners and other things, how to really like we need to understand the humanity and understand how people purchase, not just what AI thinks is the way people purchase and how markets move. We can see this with our own eyes, right? AI can help us like bring that information to us, but that actual action is very, very abstract. So yeah, that's sort of my fun, little bit philosophical answer. But I guess that's kind of the difference these days. Like I think if I if I came up with an answer that sounded too technical, it might even sound like I came up with a chat GPT. So I'm trying to I'm trying to be holistic, you know.
speaker-1: Yeah. Yeah. And the one part of the quote that I want to call out there is that yeah, businesses still want us to make money. And I go, Hmm, are they? With how many, you know, billions of dollars they're throwing in the trash? It's like they forgot how to be a business once AI entered the room and they walked out of the room with their new best friend and left all their other friends in the other room. So you kinda gotta get back to that and then, you know, like you said, it it's a tool at the end of the day. If you can justify the tool and the return on value equals plus, you're still learning. If it equals minus, what are you doing?
speaker-0: Yeah. And a lot of it is equaling minus right now because it's like, ⁓ I'll take the risk before the reward. But which is fine. Like I actually don't I'm I'm a startup guy. So I'm not gonna say no, but how long are you gonna be in the red for? Right? Like what are we doing if you're if you're being if some company's telling you have to be in the red for four years, that's ridiculous. That's like that's literally against there's why are you why are you d investing in a lost cause, right? So that's constantly bringing you down just because it has the word AI and it's because this tool has AI in it, like You gotta move quickly. You gotta learn how to outgrow these kind things. And and that's kind of the the nature of the industry right now. But people are, I mean, sucker born every minute, right? Sales salespeople are ⁓ it's funny because we've always talked about like replacing salespeople, but some of the top sales leaders are still people selling to people. Like that is still that's another fundamental business. There's still very much a lot of that. And AI is not changing that overnight. It might maybe in the future there's might be different things, but we're not there yet. Read the room, right? Much less you will actually earn more. Like I I tell the stars all the time, like you're actually hurting yourself by like having an AI chat bot when you actually need to be a startup that goes to the office of an enterprise and begs on your hands and knees to like have them do a pilot with you, right? Like it's that you don't just send out your little army. Like you can send out millions of cold emails. I'm not impressed if you don't have one customer, right? So I don't know. Maybe I'm too harsh, but I I mentor startups and I I become a bit more of like a feels like Rocky sometimes, you know. Like I have to like it used to be a bit of a industry where like I get it, like there were all these MA. We had a happy golden era of SaaS and MA and all those other things, but things are tougher now. And if you're not gonna admit that, then you're in denial and you're hurting your own business and your and your team, right? So that's another tangent, but just selling like that's another startup versus enterprise thing too to
speaker-1: Yeah. Yeah, yeah, I have a feeling we'll have many different conversations because my my area of ⁓ and A has already came and went, so and I think with that is it's a great time to go ahead and hear from our sponsor. Hello everyone. Ryan here from Stellarhorn. Let me take a quick second to talk to you about something I see all the time. Most businesses don't actually have a technology problem. They have a leadership problem. They've got systems, tools, even teams in place.
speaker-0: Yeah, yeah, of course.
speaker-1: But no clear direction, no strategy, and no one connecting technology to the business. That's exactly why I built Stellarhorn Group. I help organizations through fractional CIO leadership, advisory, and project based work, bringing executive level technology leadership without the full time CIO cost. Whether it's strategy, cybersecurity, AI governance, or leading critical initiatives, it's about helping you make better decisions, reduce risk, and actually move forward with confidence. If that sounds like something your organization is dealing with, please reach out. Let's have a conversation. Stellarhorn Group, rare leadership, real results, technology leadership without the full time CIO cost. And with that, let's go ahead and take a look into how does product thinking change AI adoption?
speaker-0: Good one. Yeah. Yeah. I love my my product. It's funny because I I've been like my product background is always a an ever existing current, but it's like it's evolved so much, right? Because of the way it's gone. Product thinking, what I love about AI in particular, and and why I didn't like I didn't I didn't, you know, fault vibe coding too much is because there is a world now where you can, as a product leader, build what you want to build in a in a very nice MVP or even like the design state, right? Using AI and and If it's an internal tool you're trying to build or something like I don't recommend using, you know, whatever off-the-shelf lovable templates and and and quick one-shot generated things as live products, but like you have a way to work with an engineering team and with a design team and say, okay, here is my framework. And here's like it's not just a Figma prototype, right? It's not just one of those things where you click and it's like a fake. You can do a whole experience, but then you're holding your team accountable to help you really come up with the value of it, right? Okay, how do we, how do we, you know, incorporate the UX and the onboarding and What we have here? Right. Because that's that's what AI I think is still very much missing. Right. And so I think the product thinking there is basically I can actually build a version of what I want to get at as a product owner, what I want to add. And my team, it's not that, ⁓ I'm gonna replace my team overnight. I'm done. They're gone. Right. It's that my team's gonna help me. They're gonna hold me and the company accountable to make sure the security practices are there, right? The compliance for whatever industry I'm in and the and the interface and the way the data's moving, right? It's it's being able to basically create these products and prototypes on the fly. To be able to unify ⁓ a w like an experience that everyone can test and use and then come back with like real feedback. Versus again, like engineers, I've I've heard like they weren't always fans of sifting through Figma and designers weren't always a fans, fans of like having to set up a local machine and set up a website or a web app or something, right? I think this idea that we can have, you know, these bive coded environments we're all working together collaboratively is a very awesome opportunity and can allow people to perceive things in different ways, but with again with a unified language of like this is actually a usable. product, even if it doesn't have like a back end or something, it is something you can go down a few steps in and we can we can leverage that and use that together. So it puts product people in a position of execution, right? Not just not just product thinking, right? It's product building now again. And and I've always been a technical product manager. So I've naturally gravitated. I've already been, you know, it was maybe not the best use of my time back then, admittedly a few times to build a website on my own, but it also was because I was an engineer. So maybe it wasn't all that bad, right? I could actually speak the language. But non non technical product managers were at a bit of a disability sometimes, right? In in more technical companies or even being able to get into it. And I think now it's not even about being a product manager as much as it is being like a product executor. Right. It's like, okay, now I'm going to connect, you know, the the engineering team and design teams to the sales team. But again, the sales team can come in and use ⁓ this environment, these vibe coded environments, these vibe coded templates and things like we can just come up with this faster and I can adopt the language to different like, you adopt the product rather in different languages to different departments. So there's the marketing, what marketing needs to know, what sales needs to know. And I think it's good, you know, again, I think engineers and designers won't always have that eye for business the same way some business like product owners do, right? In the way that they see the business and the intersection. And so they can be the translator in a way that I think AI still can't quite can't for the same reason of that clarity and that ⁓ you know, abstraction layer, right? That that is the humanity part of it all.
speaker-1: Yeah, no, very Barry right. ⁓ you know, one one of my closest allies and a friend is a is a product manager. So I learn things from him all the time. I'm fascinated with what draws people in and why you would change a button to this color or this, you know, shape or or whatever. And it's always fascinating to me the the psychology behind it really.
speaker-0: Exactly.
speaker-1: what would you say is the fastest way to actually see value out of AI in a business?
speaker-0: Genuinely, like as as as funny as this answer might sound to some folks, and you might get it better, Ryan, because you you are also sort of business owner, right? But it's like automating my own tasks as a leader has been incredible. I mean, I use I use a tool called Motion, which is awesome for product project management. It's kind of replaced like a lot of different tools for me. And I kind of brought them into all in one for my company. And like we're we're not a huge team. We're kind of one of those like, what do you call them? Like the very tactical, like, you know, ten people, but we do the work of like thirty when we that's just the nature of smaller companies and And I'm very proud of my the way that my team is using AI to make that happen, right? But we, you know, we we don't like I hate the idea of having my team already siloed out at that small of a level. So we have tools that literally ⁓ we have Cloud Max, we have motion, we have all these other things as well, you know, Miro and and there's there's other like Slack that still use that too. But I've I've reduced the amount of tools but focused it even on like what I use it for day to day, right? Like what I the the kind of things I need to remember, the way that the tasks surface themselves, like. The tool that I use, like AI helps automate like the calendar for me and make sure it prioritize things for me. Like it's fine to have that. And I think the quickest way to see value is like to give it to leaders and engineers, frank. Like it's to give, it's to give for like the menial tasks, right? So you know, you don't have your you don't have to have your team vibe coding the whole product, but like they should definitely be writing unit tests with AI. Why are you handwriting unit tests and regression tests? Like it's so much and it's still up to you to be accountable and to check these things out. But yeah, I would say it's like it's the menial tasks. That are the absolute fastest. You definitely see an increase in productivity day one. And it's incredible, right? It is for I have things for my email, I have things for my daily reporting and analytics. Like I'm very happy to have those. And it's actually nice. Like they used to require more setup, but now all those tools have their own AI enhancements. And they I can have text messages. I can use Zapier. Like, I think those little elements and investments from NSV are the same thing. Startups, mid market, enterprise, the cities I'm working with, the universities I've worked with, everyone that I can think of is utilizing these, you know. sort of self-helping AI tools for the menial stuff. Like everyone is seeing value there, right? It's overtaking the bigger swings where that clarity becomes a lot more important. But on a day-to-day basis, why do I have to sit there and, you know, organize my calendar repeatedly, like 10 times a day, right? Now why not have AI? I'm a busy schedule. I'm out and about. I'm at events, meetings, podcasts. Like I'm happy you can help me do like that one little thing, right? ⁓ and that's instant ROI. Instant ROI. So yeah. Using as your copilot. Nothing wrong with that. Nothing wrong with that
speaker-1: That. Yeah, no, I'm like minded where that is the most return on value. The more me time that gets freed up to actually work on something or actually produce that value for a client or whatever it is I'm doing, fantastic. Get rid of all that that that time sink stuff that is still incredibly important, but you don't need to do it. That's useful. So what is one piece of advice For leaders trying to implement AI that you could provide.
speaker-0: That's gonna easy answer. Like look at the data within your company, right? Not not just and look at like the way that AI do your research on how AI is taking in that data, right? Especially if you're if you're using, you know, Claude at its most API level, right? And you there's differences in using it between an enterprise plan and not. But in general, like just look at how AI is evolving and look at your data and start to look at the ways that that how that could like what that outcome could look like, right? Try to for like work with your team, work with people around you to figure out like how what the value that you're trying to drive and look at how what's happening to other companies and what people are seeing and what kind of benchmarks. There's people using AI all the time who are coming up with great reports that that t that show you like here's how it's doing math, here's how it's doing data sanitation. And I think it's up to leaders right now to really understand the limitations. Don't just look for the positives, look for the because there are plenty of positives, but look for some of those negatives on purpose so that you can understand, okay, this is the data I need to give it. This is what we need to do with people and with other tools and other resources to make sure that we're giving it the data that actually produces something useful, versus just like here, take all my documents, scram that in there, and just ship me off the nicest slide deck you can, right? Or the nicest ⁓ product feature that you can and ideas. So that's like one part. It's like the data and also just like I was saying earlier too, go back to the beginning a little bit, right? What is this business's vision, values, principles, core value that makes customers say, I want this business and nothing else. Is it the relationships? Is it the trust? Is it the volume? Like the ⁓ kind of flip side of relationships, the commoditization of it. Look at all those things and understand the limitations of that and what you want to be too. But don't AI is not the it can be part of the evolution, no doubt, but it shouldn't be the answer to the evolution. It shouldn't just be like that. All right, ⁓ Claude evolved my company to stage, you know, five hundred of its lifetime, right? If it's if it's hundreds of years of existence or something crazy, depending on how old the company is, of course. ⁓ I think that clarity is just is way too important not to not to mention repeatedly. Same thing with the day with the data problem, right? There's all these little things, but they all go back to this problem fundamentally, these two sort of big items fundamentally.
speaker-1: Yeah. And I I think just ⁓ to leap onto that, you know, have honest conversations two way, honest conversations with your IT team to say, where are we from this standpoint? And if the answer is what I suspect it would be, which is we need time and we need to do a project before we can do this project in order to actually have better data go into that, then that's where you're at. And that's probably where a lot of people are at. So you're not really behind and if you jam like we've been talking about, bad data in. it's not really gonna get you anywhere and you have to be receptive to that as a CEO and I'm sure that's hard talking to boards and, you know, investment cycles, all of that good stuff. But that's the actual answer because if you do that cleanup, probably have a chance at more success, maybe even additional projects. Who knows?
speaker-0: Yeah. And as someone who's worked with executives, like there is a way to, there's a narrative for sure. I get I get how large companies work. And even startups too. Like I know there's hubris and ego and all these other things. Like I understand it's very, very hard to like seem like you're taking a step back to step forward, but it's also like you you could lose a lot of money and then no one is happy. Right. Like there is there is a way to find a middle ground. And I think each company is very unique. There's not and there's not I that's why I'm in business, right? There isn't an easy answer that I come in with, but there's frameworks to use. There's understanding of the human, like I do a lot of ⁓ you know, I I use this principle very frequently called most advanced yet accessible, which is I or yet acceptable rather, but it's also accessible too, where it's like, you know, the I like you could come up with a flying car tomorrow, right? You could literally present the flying car to the world, but we have no infrastructure and a lot of mistrust in the idea of just flying around and there's planes in the sky that are, you know, having issues and stuff, right? Like we we have way more to go before we have that become a normal thing. And so you have to start with like other ideas first and other sort of evolutions, same way you would building a startup, right? Your MVP. It shouldn't be that you're like necessarily building a car in pieces. It's like the scooter, then the bicycle, then the motorcycle, then the car, right? There's like a famous diagram. And that's what this Maya thinking is. And it's it's it's like the secret to a lot of brands that honestly seem unfazed by this AI, boom, not because they're not driving value or anything with AI, it's just because they are using it to always do what they do best, right? not trying to reinvent themselves or or or just do things just to be like, you know, ⁓ crazy overnight. I mean, it's in fact with everybody, but there are companies that are surprisingly like bulletproof in this, but I really think it's because they have they understand the movement. They're not trying to rush anything. They're not trying to just be in the press and the media all the time by doing these these crazy leaps, right? Like I said, it looks a lot worse if you do a leap and then you have to go back on it. Would not want to be Disney right now with the Sora and Epic Game situation, right? Like very hard to look at that and be like, ooh. That's like one of the biggest entertainment companies who made two major mistakes in AI, in believing in AI recently like a week, in a span of a week. I know they'll do fine. I know they'll do fine, but it's like it's not a great look. And so we have to kinda look at these things, right? Is that the reputation you wanna have? Yeah. Something something that I think resonates with CEOs for sure.
speaker-1: Yeah, and I'm just gonna put this on a limb. I I think it's pretty accurate. Not every company is Disney and has Disney money. So maybe maybe think about that investment a little a little closer. Maybe ⁓ actually pay attention to what your own teams are telling you and have that communication early. Have it now. Yeah. Like ha like listen to this, go have that combo. At the end of the day. Slowly.
speaker-0: Absolutely. It's a big secret to a lot of things. Good communication, right? Like definitely true.
speaker-1: It goes back to basics in a lot of ways that we seem to just run past for some reason.
speaker-0: Hey, you know, AI AI kinda had ⁓ it's like I said, there's a bit of a psychosis with it, right? Like you really get into it and down the rabbit hole and you kinda forget, We we got time. We got time. There's there's there's better better wins for all of us if we take our time correctly and and marathon not a sprint, right?
speaker-1: Absolutely. So we're at the end of our meeting here. So where can people actually connect with you if they want to get in touch, hire you, meet you, have you on future podcasts?
speaker-0: Sure. Yeah. LinkedIn is the best place to reach me at the moment. I I try to look at like I I see all the connections come through. And then I also ⁓ I I'm at nvall dot com is my website and I have like ways to contact me and sign up for my newsletter, which I talk about ⁓ more leadership stuff than but it also kind of links to some of the AI stuff. ⁓ but yeah, those two places are definitely the best. I have other plenty of other social media channels, but the spam that seems very AI driven is so bad on some of them I can't I can't r ⁓ promise I'll be responding. But luckily on LinkedIn I have a I have a good filter system now. So yeah. Yeah, that'll be the best way.
speaker-1: Wonderful. So with that, the closing thoughts are, you know, AI doesn't fix broken businesses, it exposes them. And the companies that understand how they operate are the ones that actually get results. ⁓ and with that, our meeting's been adjourned. And I so appreciate you joining the podcast today, Nikhil. And I imagine we're gonna be in a bit more of a communication going forward now that we've met one another. And ⁓ I hope that the viewers out there have taken some value because you said a lot of valuable things today. They should take car.
speaker-0: No, it's it's been a it's been an absolute honor. And I'm I'm glad like I said, like I I think before this ⁓ conversation even started, like we had a another podcast together. It was a good like taste of what both of us are good at and how we think. And now we've been able to expand on that. And and and you brought up some really great points too that I think like even if we had two, three hours, we could have just kept going, right? Could have kept the could have kept the tape rolling and just talked all day 'cause I've I have a lot of perspectives on on where like the CIOs might sit in IT people, a lot of people that I tend to work with. ⁓ there's a whole new world opening up there too, which AI is helping and also Diluting as well, right? So see, I almost almost went off on a different topic there alone. So, anyways, yeah. No, it's it's been a pleasure today, Brian. I appreciate it.
speaker-1: Absolutely. With that, ⁓ that so I screwed up about that at that. So with that, that takes us to the end. So take care everyone and see you in the next one.
