A Framework for GenAI Strategy - Part 1 (203)

This week’s podcast is a framework for thinking about AI Strategy. And an intro to the AI Tech Stack.

You can listen to this podcast here, which has the slides and graphics mentioned. Also available at iTunes and Google Podcasts.

Here is the link to the TechMoat Consulting.

——-

I write, speak and consult about how to win (and not lose) in digital strategy and transformation.

I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.

My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.

This content (articles, podcasts, website info) is not investment, legal or tax advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. This is not investment advice. Investing is risky. Do your own research.

Transcript

Welcome, welcome everybody. My name is Jeff Towson and this is the Tech Strategy podcast from TechMoat Consulting. And the topic for today, my framework for generative AI strategy. Now this is actually going to be probably three podcasts on how to think about generative AI in terms of strategy and then in terms of modes and then basically a playbook for CEOs. So this is kind of a lot. I've been working on this for a long time. For those of you who are subscribers, I'm going to send you four fairly in-depth articles laying this all out. The first one goes out tomorrow morning. It's pretty in-depth and I really think it's worth reading. This is basically how to think about all of this. You know, I've written six, seven books on how to think about digital strategy. Well, this is sort of the first pass on how to think about AI strategy and generative AI is really what we're talking about. So yeah, kind of a there's going to be a kind of a lot of depth to this over the next really two weeks. Anyways, this is the first part of that I hope it's helpful. I feel pretty good about it. I've been working on this for eight months, something like that. Anyways, that's going to be it for today. So we'll do a part one of that today in terms of podcasts. Part one in terms of articles goes out tomorrow morning. And yeah, we'll keep going further. Probably the next two weeks we'll cover this. That is the topic for today. Let's see. How's keeping the tomorrow morning. And yeah, we'll keep going further. For probably the next two weeks, we'll cover this. That is the topic for today. Let's see, housekeeping stuff, two things. The first is we have our Beijing Tech Tour, which basically filled up. It's May 22nd to May 26th. It's gonna be a short tour, one city, four companies. We are going to add a couple open spots to that, so probably no more than two or three. We're basically raising the limit on people by two or three. That's going to happen in the next day. If you're interested, give me a call or give me an email, LinkedIn or go to techmoconsulting.com, the webpage, you can find all the data there, all the information, including price and stuff like that. But yeah, I'm getting kind of excited. It's about six weeks out. We're going to visit four really cool companies, and we're going to have fun too. I never talk about that enough, but I should, because I get excited about the content but yeah we're going to the Forbidden City we're going to the summer palace we're going out sort of bars in we I know all the good bars of Beijing well not all of them but I know a lot of them so we're going out to some good bars on my Friday and Saturday night couple great, picking duck of course, we got to do that because everyone expects peaking duck. And they should because it's good, but I actually get more excited about the hot pot. So we're doing that. Anyway, it's going to be a lot of fun. If you're curious, go over techmo consulting dot com. All the details are there. Other housekeeping stuff. That's pretty much it. My books are up as always moats and marathons. There's six of them. I'm going to start rewriting those and updating them to include AI strategy shortly. And then there's one hour moats and marathons, which is the short one, which is a lot more usable for most people. Anyways, that's the basic stuff. Let's see, standard disclaimer, nothing in this podcast run by writing a website's investment advice. The numbers and information for me and any guess may be incorrect. The views and opinions expressed may no longer be relevant or accurate. Overall, investing is risky. This is not investment, legal or tax advice. Do your own research. And with that, let's get into the content. Now as always, we start with sort of the concept library. There's really two concepts to think about today, which I'll add to the concept library today. Number one is just the idea of AI strategy. You know, for the longest time we've been talking about digital strategy, which is what does a winning business look like? Who's going to get stronger over time as a business model, and a digital first business model, versus who is going to have a harder and harder time? That's kind of what we talk about as digital strategy. And then we talk about digital transformation, which is kind of the path to get there, which most companies are doing at this point. And you can usually talk about that from the perspective of a digital native, you know, something that was born digital, or more likely we're talking about an existing traditional company that's adding digital tools, supermarkets, retailers, energy companies, things like that. You know, that's kind of what I've been talking about for many, many years. Well, this is AI strategy and AI transformation. And to some degree you can combine these. You can say digital and AI strategy, digital and AI transformation, but in practice, it kind of, AI strategy kind of sits on top of digital strategy. Because the technology stack that you have to build, the technology capabilities you have to build within a company are different for AI than they are for pure digital. We're gonna have to build new databases, new capabilities, new types of staffing, new types of skills. So yes, it kind of goes together, but no, at a certain point you're building something new. And so I'm putting this in a separate bucket. So we'll talk about AI strategy today. This will be the first stab at that. And then within that, we're gonna talk about the AI tech stack, which is where it all begins. That everything from the infrastructure layer, the computation, the types of semiconductors you need, the pure raw computing power you need is dramatically more to do AI than it is, you know, I'm sorry, to do generative AI than it is to do traditional data analytics or even traditional AI. And then there's basically a different thing. So I'll go through the AI tech stack. And I guess the first thing to do is, look, we're really talking about two different things when we talk about AI. We're talking about predictive AI and we're talking about generative AI. Now predictive AI has been around forever. This is when you take your company database, you are YouTube in 2009, you are Google back in 2001, and you start looking at all the data coming in from users and you start to make predictions based on algorithms. Okay, so predictive AI, which I've always kind of described, well everyone describes as basically cheap and fast prediction. That's what algorithms get you. The same way a calculator gets you cheap and fast calculation. Standard algorithms in AI get you cheap and fast prediction. And I've been doing podcasts on that for two years. OK, this is a little bit different. Predictive AI is one thing, but generative AI, when we're generating content, whether it's text, which could be a story, it could be code, whether we're generating an image, whether we're generating now videos, generative AI is a bit different, and we can put it in the same language that generative AI lets you generate content cheap and fast. That's why it's becoming something that everybody can do. You can write books really quickly. You can do marketing really quickly. You can do design and logo really quickly and cheaply because that's cheap and fast generation, the same way we used to talk about cheap and fast prediction. So it's kind of one simple way to think about it. I find that pretty helpful, but those are the two ideas for today, the two concepts. The AI tech stack and then AI strategy, which we can really break into predictive and generative AI. That's not a bad way to think about it. It's obviously a lot more complicated than that. Okay, so we started at the beginning. What is digital strategy and therefore what is AI strategy? And I have traditionally explained this with my little pyramid chart, which has six layers to competition. You have tactics, you have digital operating basics, you have digital marathons, you have barriers to entry, you have competitive advantages. And I've sort of said, look, it's kind of when we're talking about that, there's really two layers to this. We are either building a competitive strategy based on operating performance or based on moats and structural advantage. And you really need to do both. And the symbol I've sort of given is, look, it's Elon Musk and it's Warren Buffett. Right, that's if you've read my books, that's what I kind of, I talk about, look, operating performance. books, that's what I kind of I talk about. Look, operating performance, the reason we use digital tools, digital technologies, and digital business models is to improve our operating performance to make our companies operating model faster, smarter, better. Right, that's really what you're looking for doing things faster, smarter, better. And we can call all that improved operating performance, a better operating model. And I put that all under the category of Elon Musk because he is known for being a crazy operator. He just works faster than anybody. And he can innovate, well well not anybody, but he's a good extreme example. There are others obviously. He can innovate very very quickly. He can create new things very very quickly. Kind of a maniac really, but he's a good, his personality is a good sort of symbol of an extreme version of operating excellence. And there are lots of others. And then the other person will be Warren Buffett. He is only in the business of building moats and structural advantages. Now, if operating performance is about who runs faster on the field. Okay, moats and structural advantages like look, your business is a motorcycle but his business is a Ferrari and that business is an airplane. Certain structures are just better, faster, able to do more. So that's kind of how I break it into my buckets. And I've sort of added one to this and I'm going to put the graphic in the show notes. It's really three questions. It's not two. It's not just the two I've just described. It's three. And number one, which comes before the other two I just mentioned is, look, you gotta get your, you gotta get customers and you gotta grow. All this business model stuff, all this strategy stuff, doesn't matter if you can't build a good product. Get your customers, start to iterate and rapidly improve your product, get product market fit, get your customers, get their engagement, retain them, and then start to grow. That's really question number one was get your customers and grow. And I never really talked about that enough in the past and I really should because you can have the greatest strategy in the world, but if you don't have a product that's getting a lot of engagement and traction, none of the other stuff works. So that's kind of question number one or any strategy has to achieve three goals that are interrelated and goal number one is get customers in growth. And the symbol for that is at least in my mind as I use Steve Jobs. You know that guy he wasn't real good at strategy like he never wanted to build a platform business model Bill Gates ran circles around him because even though Steve Jobs had better products, Apple really did have better products all through the 80s and 90s, but Bill Gates kept whooping him even with crappy products because he had better strategy in business models. He was building platform business models from day one. Steve Jobs never wanted to do that. And when he finally did build a platform business model, which was the iOS operating system and the, you know, basically the app store for the iPhone, he had to be dragged into that. Even at the last moment, he didn't wanna do it. But then when it was finally put in place, it was a tremendously powerful business. No, his skill was building products, you know, even to the point where he probably had obsessive compulsive disorder. And he was so obsessive about building beautiful products. He wanted the inside of his PCs to look beautiful and pretty, even though nobody could see inside the box. So, let's say goal number one, get your customers and grow. Steve Jobs, good example of an extreme personality in terms of doing that. Okay, goal number two, improve your operating performance. Get bigger, get faster, get better, build out an operating model that is continually upgrading with digital tools and capabilities so you can outperform. Good symbol of that, Elon Musk. And then goal number three, build a moat. Build a competitive advantage, build a barrier to entry. Ideally, not only do you want a powerful business model that's entrenched, you want it to be entrenched in a very valuable position within the value chain. You don't really want to be, well, I shouldn't say don't want it. If you're a restaurant, you try to build a competitive advantage as best as you can and build a boat as best as you can It's pretty hard to do but even if you build that you're not in a dominant Position in the value chain where you can tell everybody else what to do. You're not in a point of control You're not commanding the user interface Now if you are Apple and you have the operating system and the user interface. Now if you are Apple and you have the operating system and the user interface, not only do you have a powerful business model, which is an innovation platform, you are right at the point of control because you're making the phones that everyone is staring at. And pretty much everyone who wants to operate on a phone is downstream from you. And you can kind of tell them what to do. They really did control the user interface for a long time. Now, WeChat has kind of trumped them in that in China, where WeChat is much more of the user interface than Apple. So that's kind of interesting. But yeah, so you want to moat and ideally you want it in a real position of power in the value chain. If you can, restaurants simply can't do that. There's no control point for restaurants. Now, if you have a business that also owns a bunch of shopping malls, let's say like central group in Thailand, well they own the shopping mall so they can give their restaurants the primary positions in the shopping mall which is nice. So you can do it a little bit. Anyways that's sort of how to think about digital strategy and I want to kind of go through that again because when we start talking about what AI strategy is, we want to talk about where AI can impact those three goals. That's how I break it down in my mind. Okay. AI is doing a lot of cool stuff. Let's see where it's playing out in goal one, goal two, goal three, Steve Jobs, Elon and Warren's world. Where does it matter today? Where is it likely gonna matter in three to five years? This is how I start to take apart AI strategy. And I'll put those graphics in the show notes. I find that pretty helpful. And the sort of beauty of my little world is I basically focus on how new digital tools and now how new AI tools are changing the answer to that question and how are changing goal one, goal two, goal three within a strategy. It never really ends because every month there's new tools and there's new technologies and there's new business models using new tools. So I said it kind of a very exciting position. Yeah, it keeps changing every year. It's really fun. It's fun for me. It can be sort of agonizing if you're a CEO because you know you think you you won and you did great and then some new Technology comes along and you got to figure out does this disrupt my world or not? Which is something I talk with CEOs a lot of the time as like look don't worry. You're good. You're good You're good this yeah, it's getting a lot of press this new generative, but it's not going to impact you in a major way like that kind of But it's not going to impact you in a major way. Like, that kind of opinion tends to be helpful, actually. So anyways, okay. That's sort of how to think about AI strategy as a first pass. That's a decent framework. Let's get into what's happening today within that framework within generative AI. All right. Now, given that simple framework, we can immediately sort of, I don't know, make some conclusions about what AI strategy is right now. Now, most of the generative AI stuff going on is mostly about the first and second goals. In fact, it's mostly Steve Jobs' world. It's mostly creating new products or infusing generative AI into existing products and doing lots of experimentation. Lots of use cases, trying to see what works and what customers want and does it get adoption and tons of companies looking for product market fit right now and most of it's not working out. Most of it's like just fun little stuff that doesn't seem to get much traction. So that's sort of one bucket of what's going on right now. And then the second bucket of what's going on right now is over an Elon Musk operating excellence. Now it's clear this is gonna change the idea of how you create products and services and get customers, right? It's also pretty clear that as a productivity tool, this is really going to change how the internal operations of companies work. So when we're talking about getting operating performance upgraded, achieving operational excellence. It's obvious to everyone this is going to have a huge impact. You know, coders are becoming dramatically more productive by using this. I've talked a bit about IGE, the sort of Netflix and HBO of China. You know, they're using generative AI not just to be more productive, i.e. your employees can do more with less, but also to just improve the overall quality of the films they're making. So it's not just making things cheaper and more productive, it's making things better. So yeah, that's kind of what we're talking about, and most of the examples of when you hear about people talking about AI strategy, that's really what they're talking about and most of the examples of when you hear about people talking about AI strategy, that's really what they're talking about. They're talking about goal number one, product services, get some customers, get traction, or they're talking about improving day-to-day operating performance. There's not really much going on in terms of Warren Buffett World of, is this going to give create a moat? And most of what I've read, and I kind of read a lot of what is written about this, I don't really agree with when it comes to modes and network effects and data network effects and all of that. Now I'll give you what my conclusion is in terms of where AI actually creates modes and where it doesn't in one of the next couple parts. But that's pretty theoretical at this point. But I think we can put a stake in the ground where I think it's real. So I'll give you my take on that. But the vast majority of what you read out there is we're talking about goal one and goal two. That's pretty much it. Now, venture capitalists are not real great at competitive strategy anyways. They don't really think about long-term competition. They're more early stage. That's where their expertise is. Equity analysts tend to be much better about long-term competition between Bosch and general electric and who's going to win and who's getting stronger VCs which is where the Gen AI world is right now. They're not real great at that The other thing to think about is when we talk about how this is going to change strategy You want to really separate? The companies that are creating this technology which is where most of the focus is today. Everyone's talking about Microsoft and Google and Baidu Cloud and OpenAI, of course. Okay, they're creating the tech. Most of the value is not going to be with the companies that create the tech. It's going to be with the companies and businesses that adopt the tech. I've used this quote before, which is Philip Fisher, the famous investor. He was a tech investor in the 50s and 60s. And he always used to say that like it's as powerful to invest in the company's adopting technology as those creating it. The Chamat Palahapatia, the all in podcast guy who's really interesting to listen to. He just talked about literally the same thing last week. I heard him make a point on a video and I'm like, that is exactly Phil Fisher. And he basically said, this is a direct quote, "the people and the person that invented refrigeration made some money, but most of the money was made by Coca-Cola who used refrigeration to build an empire." It's the same idea. Okay, these companies are creating the tech and it's amazing. Most of the value is not going to be there. It's going to be with companies who use this to create new products or change how they operate or do other things. It's not the people that created refrigeration. It's the people that realized this we could sell Coca-Cola in cans to stores, retailers, and homes. This is the same idea. So when we talk about digital strategy and we talk about open AI, don't get too caught up in the companies that create the tech. We really want to talk about companies adopting it and doing interesting things. Okay, so the first sort of wave of strategy for AI strategy in 2024 would be everyone needs to start putting generative AI into their products and services right now. That's Steve Jobs land. You're either going to create new products entirely if you're Bosch, if you're a supermarket, if you're Coca-Cola, or more likely, you're going to start to integrate all these new tools that are just rolling out into your existing products and services. Okay, that's what IGE is doing. That's what Netflix is doing. Most of them are not creating the tech at all. They're just putting it into their current products and services. That's most of what you'll read right now. And I would argue that's goal number one, fine. That's Steve Jobs' world. And then goal number two is, okay, we start basically building these tools into our operating activities. Our work flows right now. What does that mean? Well, in practice, that really means two things. If you look at my digital operating basics, which I've been talking about forever, I've basically changed them now. Now I'm calling it starts to change the operating basics, it's basically DOB3 and DOB6, which is DOB3 is build a digital core for your management and operations. If I'm talking to a company about doing digital transformation, we are almost always talking about building the digital core, which is look, you have to become much more data-driven as a company, which means we need to start gathering data internally, we need to start creating ecosystems that connect with other companies outside of yourself to pull data into one single version of the truth that then enables management to become dramatically smarter and dramatically faster than they have ever been before. Right, get speed. Get data. Get speed, start making quicker decisions, things like that. Well, we would call that the digital core for management. And then we would start talking about the digital core for operations. Okay, now that we have a digital core built up, which really means building out a data lake, building out a data warehouse or a lake house and then starting to put various analytical tools on top of that which have been sort of standard data science or have been predictive AI which is looking back at data and then making predictions. Well we have to start expanding that to include generative AI which means you need a digital core for management and operations and you need an AI core for management and operations because the data architecture is different. It's not the same. You are to do generative AI, you have to use tremendous amounts of data and computation far beyond what a typical data warehouse can do in a company. You're talking about tremendous amounts of data, most of which is not going to come from inside your company. So you are starting to build into the data streams of your industry. The data layer of the AI tech stack is really a big deal. And you need a lot more computation. So DOB3 is digital and AI core for management, and increasingly over time, you're going to put that into your operations. So I'll put my list for digital operating basics, and you'll see that I've changed DOB3. The other one you have to think about is DOB6, which is staff, teams, and culture. When you take a supermarket and you start to become more digital in your activities, well, staffing becomes a problem. Culture becomes a big initiative. So does leadership and management and a lot of hiring, a lot of training, a lot of upskilling. That's a huge part of what has to be done. Usually that's the biggest bottleneck. If I talk to a retailer, a bank, an energy company, a healthcare company, that always becomes the bottleneck. We need people and we need to start training systematically because we're going to get caught on DOB6. Okay, DOB6 just expanded. We are not just talking about people who do data science and coding. Suddenly we're talking about people who have AI skills. So that means a new type of talent has to be recruited. New training and hiring has to take place systematically. So that's a big deal. And that's your people. And what happens is you build out the people and then you start, it actually is bigger than this because as you start looking at your culture and your overall staffing, not just your AI people, you start looking at all your people, they're all gonna start using AI copilates in what they do. It's not gonna be one person doing management in let's say marketing. It's going to be a manager with an AI co-pilot making that person dramatically more productive and effective. So this idea of training and people and culture. Suddenly it's not just people. It's people plus AI tools joined at the hip, which we call copilots. That's a big part of it. And it even gets bigger than that because you're going to realize when you talk about the people and the agents that are your workforce. Well, your workforce is no longer going to be just human. You're going to have a significant portion of your workforce that is non-human. You're going to have autopilots. Co-pilots, you have a human flying the plane, but the co-pilot is a generative AI tool or an AI tool. We're also going to have autopilots where there's nobody human flying the plane anymore. So we're gonna have this whole part of our workforce that's gonna be auto pilots. That all goes under DOB6. We are gonna have a workforce that is part human and part non-human, and human parts gonna use autopilots or co-pilots. So that DOB3 and DOB6 are becoming dramatically different with Gen. AI. That's the "So What?" The others are pretty much the same. And you can go through it. I'll put the slide in there so you can take it. So that's first pass analysis for an AI strategy. Goal number one, get customers, products, growth, Steve Jobland. Put Gen. AI into everything you possibly can. Start experimenting. Goal number two, operational performance, excellence, upgrades. That's Elon Musk land. Do B three and do B six are going to be dramatically different now. Start moving on those two as fast as you can. That would be a solid first pass of an AI strategy for most companies. Like that would be a good solid take to begin talking about. Okay, that is most of the content I wanted to talk about today. But there's one other big thing to talk about, or at least tee up, which is you gotta start getting comfortable with the AI tech stack. We kind of know what it's going to look like now. Everybody just has to understand this. You can't do AI or digital strategy, or you can't do AI strategy or AI transformation without understanding this new emerging tech stack. And I've talked about this before quite a few times in different podcasts and articles, but I'll put in three versions of the tech stack in the show notes. Really take a look at them. Now, the standard one people talk about comes from Andreessen Horowitz, which they call their preliminary generative AI tech stack. I've posted this many times. Basically, it breaks it all into apps, models, and infrastructure. So infrastructure is the bottom layer. That's computation. That's NVIDIA chips. That's massive server farms. That's data centers, that's cloud services. Because as mentioned, you need a tremendous amount of specialized computing power to do any of this. Okay, and the next level up is the models. The foundation models, everybody talks about all the long time, the large language models, GPT, the generative image models, Dali, and now we're seeing things like Sora and others that are doing video generation models, but we also have ones for sound, for voices, multi-modal models that combine all of that, that's becoming a big thing. So everyone's talking about the models and they can sort of be close source, which proprietary models, which is OpenAI, even though they're called OpenAI, they're really closed now. But then you have all the sort of open source models that people are using that are coming out of. Pretty much everything out of China is open source. Baidu, the Hwauwe, Pangu models, Ali Baba's open. There's really only a couple closed. And then the next layer on top of that would be the apps, which everybody is building like crazy chat, GPT, Git Hubs, Co-pilot, mid-journey, and sometimes they're integrated with the model and sometimes they sit on top of the model. Anyways, that three layer model is not a bad way to think about the tech stack. And you should look at that model, that graphical, put it in the notes. I don't actually use that one 'cause I don't think it's awesome. The one I like better is the one that comes out of Huawei, which I've posted before, which they basically lay out five layers to the AI tech stack, the gen AI tech stack. The bottom layer, the infrastructure layer, they've broken that into two layers. Instead of just having infrastructure computation, they break it into basically cloud service and a data layer. Now the cloud service is really servers, data centers, the semiconductors, the cloud service, that's all kind of one thing. But the layer that they break out is basically a data layer. And they argue rightly, I'll put the graphic in the notes for this one, that the data requirements of generative AI are so huge and they have to be real time. Data has to flow into these models all the time for it to do inference. And then once it runs on the data, it has to flow out and give orders and conclusions. So that data flow layer is incredibly important and it's not gonna be within a company. It's gonna be, well, I mean, certain companies can do it like Google, but most can't. It's gonna have to flow within an industry. We are going to see data ecosystems within industries like automotive or let's say Shanghai as a city or let's say industrial robotics or mining or all that data has to be tagged. It has to flow. The security has to be in place. The permissions have to be in place such that the data can flow and you can get high quality inference and generative AI. So that data layer is really complicated and it's cuts across industries and geographies. So that's kind of one layer that I think is valuable. The others, when you talk about the foundation models, they refer to that as L0. So that's GPT, LAMA, which is, you know, these models that stretch across lots of industries. They call theirs Pangu, the NLP model. But then on top of it, they talk about L1 models, which are really industry specific, which is I think the right approach. You take a general broad model, like GPT, but then you have to build a customized one for the mining industry, or for the airline industry, or for the automotive and transportation industry, where it becomes very customized for that specific situation and you basically build those on top of the foundation. So L1 gets built on top of L0. That I think is true as well. So it's good to break those and then they also talk about L2 which are sort of scenario specific models like media or marketing or customer service, things like that. You could kind of say those are apps, but you basically want to break out the models into a couple of layers and then you have apps. That's a good way to think about it. The last one, which I won't go through, but I'll sort of point out is McKinsey has a pretty good generative AI tech stack they talk about and what they do is they break out not just models but they break out things like the tool kits because that's kind of what if you're a company you're gonna have to build in sort of tool kits that let people access these models they talk about things like policy management and governance layers that give permissions to various levels. The customers have certain permissions. The senior management has certain permissions. Others don't. People within the company will have different permissions than, let's say, entities outside your company, which is when you do data sharing is very important. You know, if you're a A, you're not gonna let bank B see client specific information, but you are gonna share a lot of information between the banks. So they lay out, I think, a more robust version of that, and I'll put that in there, but it's kinda too complicated to me explained by, you know, describing a graphic, which is not great. Anyway, that is kind of the last idea for today I wanted to talk about, which is, get comfortable with the generative AI tech stack. And I've given you three versions. For those of you who are subscribers, I'm writing a lot about how to take these apart. But yeah, those are the two concepts for today. The generative AI tech stack, and sort of this first pass of AI strategy based on three goals as a framework and then from that you can make some preliminary conclusions on where generative AI is going to change the game, which I've kind of said multiple times. Anyways, that is it for the content for today. I'm a little bit over about 35, 36 minutes, I think. Sorry about that. I'm trying to stay right at 30, but I always seem to go over a little bit. But yeah, this is a huge topic. Super important. And hopefully, this is helpful, and I'm going to help you get ahead of the curve, because most people do not understand this stuff yet. And I'm sort of struggling to stay on the frontier of putting this in a usable form for investors and also for CEOs. That's kind of who I think about. Anyways, that is it for the content for today. As for me, just kind of a busy week overloaded. A lot of projects, which is fun. I'm planning a trip to the Philippines. I'm gonna be there about a week. That'll be fun. I like going to the Philippines. I don't really do a lot of business there. I've given talks and keynotes there over the years. Usually when I'm there, I'm writing motorcycles. That's kind of my thing. Like, I don't like the cities as much. Like, I really like cities in Asia. I find them really dynamic and exciting. I don't really like hanging out in Manila that much. It's kind of, oh, it's not the most fun place to be. Sebu as well, also not awesome. But I love getting out of the city and going into the provinces and sort of riding up to the north, like motorcycles going up to Baguio is fantastic. Or you can kind of red west from Manila over to Batan. And that's really fantastic. So I've ridden over quite a bunch of the islands. So I'm gonna try and sneak out and do that. It ain't the safest activity in the world, but if you're up in the provinces, it's pretty good. I've almost hit water buffaloes a couple times because you come kind of rolling around the hill pretty fast and then the water buffalo, they just kind of stand there in the road. And I have no doubt, like if I run into a water buffalo, I'm pretty sure he does fine and I'm pretty screwed. So yeah, I do think about that when you're sort of rolling through the hills, like what happens if certain animals jump out? Like I know if it's a water buffalo, I gotta swerve and hit the brakes or I'm in trouble. I know if it's a chicken 'cause they will run out on the roads every now and then. I've already mentally decided, if it's a chicken because they will run out on the roads every now and then I've already mentally decided if it's a chicken I'm just gonna sort of you know stabilize the bike and go straight through like I'm not swerving I'm just like balancing and I'm going right over as I've mentally sort of game this out in my mind I've never actually hit one or swerve, but I've come close to a water buffalo before. So anyways, these are the things I think about. I also tend to ride, I tend to ride bigger bikes like not these little 150 CC bikes you see in Vietnam and places, but no, it's like 650 700 CC like really big heavy bikes So I'm pretty sure I'm in good shape for most things minus probably the water buffalo is the one that would take me out. Anyways that's what I'm thinking about. I'm gonna plan a little motorcycle trip and yeah I don't know why I think about these things but I do. Anyways that's it but more usable information if you can ever take a motorcycle trip up north to Dagupan which is sort of you go Manilla, you go a couple hours north, and then you go up to Baguio, which is beautiful city up in the mountains. That's kind of my favorite ride. That or either the baton area, which is sort of beautiful on the water. Anyways, that's what I'm thinking about. If you have any recommendation for good places to go in the Philippines for like motorcycles and stuff, I know all the standard, you know, Boracay, and I know all the standard tourist spots. But motorcycle trips or hikes, I'm going to do some hiking in the Philippines later in the summer. So yeah, if you have any suggestions, let me know. I'd appreciate it. Anyways, that is it for me. I hope everyone is doing well, and I'll talk to you next week Bye-bye