The 4 Pillars of Baidu’s AI Cloud Strategy (2 of 2) (Tech Strategy – Podcast 200)

This week’s podcast is part 2 of a deep dive into Baidu Cloud, a leading provider of cloud and intelligence services. And arguably China’s AI leader.

Here is Part 1.

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.

Here are the four pillars:

  • Pillar 1: Baidu AI Cloud Is an Innovation Platform Pushing to Reach Critical Mass
  • Pillar 2: Baidu AI Cloud Is Also Moving to Become a Technology Standard Deeper in the AI Tech Stack
  • Pillar 3: Baidu’s AI-Focused Cloud Has 2 Advantages Over Other Innovation Platforms
  • Pillar 4: Baidu’s AI Cloud Is a Flywheel in Industry-Specific Intelligence

Here’s the mentioned flywheel.


Related articles:

From the Concept Library, concepts for this article are:

  • Cloud Services
  • AI
  • Generative AI
  • Innovation Platform

From the Company Library, companies for this article are:

  • Baidu AI Cloud

——–transcription below

Welcome, welcome everybody. My name is Jeff Towson and this is the 200th episode. Oh my God, of the Tech Strategy podcast from TechMoat Consulting. And the topic for today, the four pillars of Baidu’s AI cloud strategy. This is actually part two of a podcast I did a couple weeks ago, which was podcast 197, which was on Baidu’s AI cloud strategy. That was part one, this is part two. And yes, this is the 200th podcast, which is crazy to think about. Given that I started this about three and a half, four years ago, sitting in a hotel room in Samut Prakhan Bangkok area thinking, “Hey, I should start a podcast sitting on the hotel bed doing about 40 minutes of a lecture on something.” And now 199 really kind of lectures talks later, 199 Really kind of lectures talks later Seven books. I don’t know how many articles three or 400 probably It is amazing how these things accumulate over time anyways, that’s where we are so today I want to talk about Baidu and Yeah, just some fun stuff because hey, it’s number 200 that seems significant so that will be the topic for today. And standard disclaimer, nothing in this podcast during my writing or website is investment advice. The numbers and information from me and any guests may be incorrect. The views and opinions expressed me 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 topic. Alright, let’s start with quick tech news and the big story of the week, which I kind of feel obligated to talk about, is the bill moving forward in the US Congress to force TikTok to be transferred, stolen to a non-Chinese owner. And it looks like it’s going through the house, it’s heading toward the Senate. Now for those of you who maybe know my opinions on these things, I try and stay non-political for the most part. But for those of you who are familiar, big surprise, I don’t buy this in the slightest. I think this is a lot of self-interest dressed up as logic and reason. But I’ll give you my take in two buckets. Let’s just take apart the substance of the argument. Now the argument is about the relationship between the US and China. And the US has to a significant degree been sort of pro-free trade, pro-following a rules-based international order set out by things like the WTO and such. Now that said, the US also kind of wrote those rules in their favor, so it’s not totally fair. But generally speaking, that has been the US’s position. I am generally in favor of that, but I am not a 100% free trader. I think there are certain times when you need to operate according to what is now being called reciprocity. If things aren’t working out in the world trade and all of that, then you just do reciprocity. Whatever that country does for you, you mirror the rules back to them. And the US has largely shifted to that position with regards to China. It’s no longer filing things within the WTO if it doesn’t like what China’s doing, let’s say in steel dumping or tariffs or whatever. Now it’s just being reciprocal. Okay, we slap tariffs on you, you slap tariffs on us. And I think reciprocity is a fine exception You don’t have to be a hundred percent free trader ninety percent enough So you could argue that the whole tick-tock thing is just reciprocity Okay US social media companies Facebook all of them are not allowed in China neither are most media companies So it is totally reasonable to shift to a reciprocity-based relationship with China. And okay, then therefore your social media companies can’t be in the US. That’s not a bad argument. And I think there’s some substance to that. So that’s, call it the logical argument. Now I don’t think that’s what the US Congress is doing at all. I don’t think it’s just the US Congress. I think it’s the security state. I think it’s all the three letter agencies. The position from them for several years has been, we want to ban TikTok. And they’ve come at the same question with lots of different arguments. Well, it’s data privacy. It’s a national security concern. I mean, it’s always a different argument’s a national security concern. I mean, it’s always a different argument, but the goal is always the same. So I think it’s self-interest, and this is just the latest cover argument. It doesn’t mean the cover argument is wrong. I just don’t think that’s what this is about. So what is the goal? Now, this is where we get a little political, or I get a little political. There is a very cozy relationship between the US government and the security state and the entities that control information in the US, Facebook, Instagram, Google. I mean, they are in bed together as much as you can be in together and they are actively controlling what information is seen, what information has not been seen. And you know, you can go through example after example, the things that were not allowed to be said during COVID, even though most of them turned out to be true. But it turned out it contradicted the narrative that the government was pushing election interference, Russia collusion. There’s just been one after the next where you see this, what Glenn Greenwald calls the censorship regime. Now, there’s only a couple companies that sit outside the censorship regime. Twitter is the most prominent. He will not go along with this. Rumble is another one, but they’re not that big. They’re not that influential. And then TikTok is kind of in this camp, not totally, but it’s not as easy to tell them what to do as the US government because they’re ultimately from a foreign company or foreign country. So I kind of predicted a couple years ago that it looks like there’s three main companies that have significant control of information flows in the US and I expect them to be targeted one by one. TikTok, Twitter, and Rumble, they’ve already come at Elon several times. They’ve tried to come at TikTok. I think this is just the latest iteration. And then rumble, I think if they get big enough and significant enough, they’ll come at them. That’s how it looks to me. I could be wrong. And that’s a little bit of a political argument as much as an analyst. So I accept that I might be not thinking as clearly as I should. But that’s kind of how I view what’s going on with this all tick-tock thing. I think this is what was totally expected to happen, which was a sort of assault on one of the two or three information platforms out there that aren’t towing the line, at least not as obediently. But we’ll see. Anyways, that’s as far as I’ll go into that subject because if you think I’m full of it, let me know. If you think I’m totally off, let me know. I think that’s totally possible. But that’s kind of in my working, let’s call it a working hypothesis for the last couple years. And we’ll see. Anyways, okay, enough of that. Now, let’s get into Baidu’s AI Cloud Strategy. I did a pretty long podcast on this before, so this one’s gonna be short. The business, very cool. It’s a cloud business, but it’s very much an AI-focused cloud business. It’s highly differentiated towards AI. And they have built, self-built, self-developed AI tech stack from top to bottom inside of Baidu over the last 10 years. Semi-conductors, computing power, server racks, their deep learning development platform, paddle paddle, their big foundation models, Ernie, which is, you know, Ernie, now it’s on Ernie 4.0. It’s Wenxin in Chinese. That’s been out for four years. They were as early as GPT in building these models. And then on top of that, the application layer, which they’re really industry focused in sort of getting this into usage at companies mostly in China, but it’ll be outside of China soon. So they have got a self-built developed AI tech stack that’s hardware and software that is really unique and they are leveraging that into their cloud business in a very differentiated way, which I think is very compelling. I think it’s a very good strategy. And to pat myself on the back a little bit, I remember four to five years ago I wrote about Baidu when they were doing food delivery and other stuff. And I was like, dude, just do search and do cloud. Just focus on those two. And I actually said this, like, and within your cloud, make it AI from top to bottom. So man, I mentioned the things I’m right in retrospect on. If I’m completely wrong, I don’t bring it up. So there’s a little bit of a survivorship bias in these comments I make. Okay, so what are the four pillars of their AI cloud strategy? Now, their Q4 numbers came out by do’s a couple weeks ago, and it was a little surprising. One, their revenue was up to about 8.4 billion remnB for Q4. So, you know, 1 to 1.5 billion US dollars for the quarter. But what was interesting in that, well I’m sorry, that number I gave you was AI cloud revenue. What was interesting about that was that generative AI and their foundation models foundation models accounted for 656 million remin B. So almost 10% of their AI cloud revenue for Q4 was from generative AI and these new foundation models. Now that’s very compelling because everyone’s flooding money into this, everyone’s building foundation models, they’re posting revenue gains based on this. I mean, they are quite a ways down the path. And the companies that are involved in this, they’re not posting their numbers. Baidu did, it was kind of surprising actually. And they gave some decent stats about, we’ll call it their Gen AI driven revenue in terms of operating performance and pretty impressive. Ernie, which is their open source, their suite of open source foundation models, there’s quite a few of them now. They said they had basically 26,000 monthly enterprise users using that through their cloud service. They said Ernie was getting 34,000 queries per minute in December. The inference cost of Ernie 3.5, they’re on Ernie 4.0 right now. But they say the inference cost has dropped to 1% of what it had been in March. So they’re driving down the cost, the volume’s going up. Now that’s a scale thing. When you drop by 99% if that’s true, well, that has a lot to do with fixed costs. So, you know, that’s what that is. Okay, anyways, it looks like the AI cloud business not only is growing in terms of their operational number, but it is directly tied to customer usage and revenue. It’s not one of these let’s build lots of cool tools and see if we can turn into a business. Now, it’s a business from day one and that’s very much a Robin Lee thing with regards to generative AI. He is very focused as far as I can tell on not doing anything in generative AI that is not directly usable by customers. So none of this building stuff because it’s cool. Very customer focused in this activity, as far as I can tell. Okay, let’s talk about the four pillars. Pillar number one. Underneath all this tech and cool stuff and cool announcements, ultimately, this is an innovation platform. That’s the business model. We’ve talked about multiple platform types. Innovation platforms are fantastic. That’s Microsoft Windows, right? Innovation platforms, the user groups are people that use the platform Which can be enterprise consumer users and then developers and companies that build on the platform. So Okay, everyone who built software that runs on Microsoft Windows Everyone who builds apps that run on Apple iOS those are both innovation platforms pretty much near monopolies. And you could say Google, same thing, Android. So what they’re doing is they’re building, and there’s a lot of other stuff going on on top of this, but the foundation of what they’re building is an innovation platform. That’s what Baidu’s AI cloud ultimately is. Now on top of that platform, there will be lots and lots of apps built and services and features. They will probably keep a couple of the crown jewels to themselves. That’s pretty standard. Okay, so Microsoft builds Microsoft Windows, global dominance, even today, after what 40 years more than 40 years yeah like 45 years but in in addition to windows they kept Microsoft office right that was a crown jewel it turned out that thing was a cash cow we can see the same thing with apples ios for smartphones yes they have the apple operating system for smartphones, but they also kept Apple Maps. They didn’t give that one. Now that one’s not a cash cow, but it’s considered a strategic capability that’s very important. So usually if you own the innovation platform, you’re also gonna keep a couple of the crown jewel apps, which are either cash cows and or strategic capabilities, critical positions, whether it could be mapping, Microsoft Office, payment tends to be one people like to keep. So I expect by due to be the same. I expect AI Cloud to be the platform, the innovation platform, and I expect them to keep a couple of the Crown Jewels for themselves. And the starting point for that is what they’re already doing is they’re taking their entire mobile ecosystem, which is a whole series of mobile apps that they’ve been building for 24 years, and they are infusing AI into almost all of them. Now the flagship app for Baidu is big surprise, Baidu Search. Okay, they’re infusing AI into that. They’re going to keep that in-house. That’s going to be a crown jewel that sits on the innovation platform. The other one that I think, I mean, there’s a bunch of them. They have, I’ll give you the list here. Baidu Search, incredibly important. Baidu maps, that one they will probably try to keep. Now how do you keep it? Well, you give it a more prominent position in the innovation platform. You integrate it into every stuff that people use. Mapping is a pretty critical strategic capability. They have a whole lot of other stuff, Baidu Drive. Okay, probably not critical. GBI on are Baidu Search in terms of crown jewels, Baidu Search, Baidu Maps, and all their marketing services, which, you know, if you’re going to put your company advertising on any of their mobile ecosystem suite of apps, you deal with Baidu Marketing. And they’ve already launched, or at least upgraded, at least three main services within Baidu marketing with their new AI tools. So they’re putting all their new AI into all of their apps. But Baidu marketing is actually pretty important. For those of you who are subscribers, I sent you one article. I’ve got four, I mean this is bad. I’ve got four articles on Baidu’s AI cloud strategy coming your way. They’re already written. I think I’ve sent you two already. There’s two more on the way. I kind of went to town on this. I kind of went way too far. Anyways, there’s more than you’ll ever want to know about this. So all the apps are laid out, what they’re doing with AI in their apps is all there. But the ones I’m keeping an eye on are their marketing services, Baidu Maps and Baidu Search. Those look to me like the Crown Jewels. So that’s sort of strategy pillar number one, which is ultimately, this is an innovation platform. They’re going to build that and then they’re going to keep a couple of the crown jewels in house. And right now there’s just a race to get to critical mass. Whoever gets to critical mass first as an innovation platform, that’s Microsoft Windows. So that’s kind of pillar number one. Now pillar number two is, okay, Baidu AI Cloud is clearly not just staying at the innovation level. You know, they are clearly not just Microsoft Windows sitting on top of all the hardware and PCs that other companies make. No, you know, they’re going all the way down the tech stack as well. So yes, they have an innovation platform on top, but they’re also going down the tech stack and trying to become the technology standard, one of the technology standards all the way down. So they want their NVIDIA chips and the architecture to become a standard. They want their development deep learning platform, which is paddle paddle, to become a standard that everyone uses. That’s like being the group that created Python. You’re the programming language that everyone uses. They’re trying to play at the chip level. They’re trying to play at the architecture level like arm holdings. They’re trying to play at the programming level like arm holdings. They’re trying to play at sort of the programming level like paddle paddle. So that’s you know pretty interesting. So if you’re trying to become the technology standard at multiple levels of the tech stack, not just at the top where the apps sit, that becomes a game of sort of standardization network effects. And you have to not just get developers and companies to build using your technology. You’ve got to get your technology deeply integrated into all the other technology. I mean, everything that people build for semiconductors has to sort of interoperate seamlessly within Vidius chips. has to sort of inter-operate seamlessly within Vidius chips. And it has to work with, let’s say, the arm holdings architecture. So there’s such tremendous interoperability required as you move down the tech stack that that’s how you become the standard. And if you do it well, like arm holdings, yeah, you can have a beautiful business for 25, 30 years, which they have had. So that’s kind of strategy pillar number one, too, which is look, Baidu’s AI cloud is clearly moving to try and become one of the technology standards at multiple levels of the AI tech stack that’s just emerging. And their competitors here are You know tensor flow by Google and You know what Facebook is building. You know, they’re playing it multiple of those levels. They’ve got to get That’s why all their software all their tech is open source They want everyone to start integrating this and adopting this they’re encouraging developers and companies to build on their tech stack. They’re having training classes. They talk about this initiative they call 5 million AI talents where they’re basically doing training and tournaments and projects and working with hundreds of colleges to get people to learn to do AI. Now you can use other tools and technologies besides theirs, but obviously theirs is one of the ones that’s taught. So that’s kind of strategy pillar number two, and the fact that they’re infusing all of their apps, their Baidu apps with their AI tech, that’s a kind of way of jump-starting that as well. So that kind of fits together. Those are the first two pillars. All right, strategy pillar number three, which is, look, Baidu has built an AI-focused cloud service. That gives it at least two in two sort of advantages over other companies building innovation platforms. They’re not the only ones trying to become the Microsoft Windows of AI clouds. If there’s a handful of companies out there, the fact that their AI focused really born AI native in their cloud services offering gives them at least two advantages in that. Number one, look they’re an early mover. They got you know if you’re gonna try and become the Microsoft Windows and become the standard in the tech stack, you know getting there early is critical in a new technology. That’s literally how arm holdings became the architecture for smartphones is back when smartphones were just emerging as an idea in the 90s. Actually, even before smartphones, they were really writing architecture for devices because the architecture at the time was all written for high-powered PCs. So they were already sort of ready to go when the new technology of smartphones emerged. They already were there. Okay, that’s kind of like Baidu, that they were already building AI-focused cloud services before the world shifted to AI. And getting their first being an early mover, getting everyone on your platform and getting your network effects going, that’s a lot of how you win. That’s number one. Second advantage they have as being an AI-born cloud service, is their AI tech stack is really kind of different than, let’s say Amazon Web Services, which was built as a complete web service for everything. And now they are trying to adapt that to use cases for AI. Right, this is a little bit like, you know, Intel, when they had their CPUs, and they started trying to adapt those to do AI services. And it just turns that out that NVIDIA’s GPUs were just a better technology for this. And it’s been very hard for CPUs to adapt to these new use cases and be competitive. And now you could say that these sort of TPUs that are emerging may make it hard for GPUs from NVIDIA to sort of compete. Well, this is kind of the same idea. Their self-developed sort of tech stack, which is their own software and their own hardware, from top to bottom, all the way from chips to apps, they may just be better positioned and can offer better solutions than companies, let’s say like AWS, who are trying to repurpose their existing technology to this new world. Maybe that looks like what’s gonna happen and they’re doing the hardware and they’re doing the software and they’re doing it for the whole tech stack. So they’re vertically integrating, and they’re integrating between the hardware and software. Now, in theory, being an AI native cloud service should make their services much more efficient and more effective. One, they should be cheaper and faster to develop, and they should in theory perform better. If nothing else, the fact that they have a sort of an integration of cloud and AI from day one should mean their rate of development is going to be faster than everyone else’s. That is probably true. Here’s a quote from Baidu. This is a direct quote from one of their documents. “The integration of cloud and AI is a unique advantage of Baidu AI cloud. As a cloud tuned for AI, it provides a safe, stable and flexible infrastructure for digital transformation, while the AI engine provides leading innovative technologies and platforms to drive intelligent upgrading. Buh-buh-buh, skip-sip, continuing on. It is the seamless integration of cloud-tuned for AI and the AI built for common scenarios. So they literally cite this as their biggest advantage as a cloud service business is that they are a cloud that has been tuned for AI. And that makes it safer, more stable, more flexible. And that’s gonna give them an advantage as companies do what they call intelligent upgrading, which is instead of digital transformation, you’re moving to intelligent transformation, which is intelligence upgrading. That’s pretty compelling. So that’s kind of number three. And the metrics I’m sort of watching is what is their cost structure for running various scenarios? How efficient are they? How good are their solutions versus other solutions that are maybe not run on a cloud native system, or sorry, AI native system? Three, this is the most important one. What is their rate of development? Are they just creating more features and more services and better features and services faster than everyone else because their technology is just better at this? That’s probably the number one thing I’m looking for is the rate of advancement because these are still early days. Anyways, okay that’s strategy pillar number three. And the last one, just about done here. Baidu’s AI cloud is a flywheel in industry specific intelligence. Now this is the last podcast I did on this part one. That’s what I talked about. The idea that they are basically trying to create a flywheel in their AI cloud services that increases the intelligence of their system specific to industries. So what does that mean? Now, Robin Lee, a very cool guy, has been very sort of not blunt but direct about large language models and foundation models as somewhat being a waste of time and that people shouldn’t focus on these as much. These cool language models can do all these funky things. Now, the goal is to models can do all these funky things. Now, the goal is to whatever you’re building, make it directly useful and usable by enterprises. Make it have value in usage, not just a cool tech you’ve built. Now, by staying so close to the customer use case, that will create information and usage that it then feeds back to the foundation model and the applications built on it to make it better. That increasing intelligence effectiveness of these models, which are let’s say industry tuned, will make them better and smarter, which will get you more usage and more adoption and more APIs and more use cases. And that’s your flywheel. Here’s how they describe it. Quote, the core advantage of Ernie lies in its knowledge enhancement and industrial level application. Now, what they write about quite a bit, which is actually very cool, application. Now what they write about quite a bit, which is actually very cool, is this idea of knowledge enhancement, which is foundation models, which are then tuned, customized to various industries and then apps are built on top of them. How effective they are is a function of quote unquote knowledge enhancement that certain AI is are going to get smarter than others and that’s the phrase they use. Now what is knowledge enhancement? They describe that as the combination of two things. One it’s the you need massive unstructured data. You need tremendous flows of data running into these AI that are coming mostly from customers, from people deploying this in the field, industrial applications. So you have massive unstructured data that combines with large scale knowledge maps. Knowledge maps, that’s a whole concept. But when you put those two together, that gets you a better knowledge. That gets you knowledge enhancement, basically. It gets you more efficient learning. It gets you better performance and all of that. Now, what is a knowledge map that’s? It’s like when human beings learn things. We consume information all day long. We have information flowing into us all day long through our eyes, through our ears, conversations, looking around the world, reading books, listening to podcasts, whatever. That flow of information is overwhelming to the human brain, which is an incredibly efficient computer when it comes to energy usage. It’s stunningly efficient and how little energy it requires. One of the reasons it requires so little energy and so little data is because there are knowledge maps in our heads that the data is combined with and that lets us make decisions. Now some of those knowledge maps are passed down through us genetically. Language appears to be innate. You know, you can put a hundred human beings on an island as babies with no contact with anyone else and they will develop a fairly sophisticated language very quickly. Grammar and everything. Language and grammar appear to be innate. Most everything else comes from learning and passing on knowledge from generation to generation. So we have knowledge maps in our head that let us combine with information we take into make decisions and do things. This is basically the same idea except the amount of data being consumed is massive and it’s unstructured and the knowledge maps are what Baidu is building within its foundation models that are highly tuned to specific industries. That’s really what they’re doing. Here’s another quote. “It also aims to promote the intelligent upgrading “of industries by constructing a foundation model system “that is more suitable for specific scenario requirements. This includes providing tools and methods to support the entire process and creating an open ecosystem to stimulate innovation. So basically, knowledge maps plus massive unstructured data get you knowledge enhancement. and they try and basically tie that knowledge enhancement as close as possible to industrial level applications and usage. That’s the flywheel. OK. For those of you who are subscribers, subscribers, I’m going to send you a lot on that. It’s really worth thinking about. I think there’s other ways to think about this, but I think there’s a ways to think about this, but I think there’s a fairly impressive way to do it. And they basically, in addition to their big foundation models, which they talk about Ernie, which they have a image generation one, they have LLMs, they have multimodal, they have quite a lot of these, but they also have four to five major industry groups that these models are being customized to. So there’s Kairu, which is their smart manufacturing group. So they’re basically tuning Ernie and bots to industrial manufacturing cases. They have a smart finance group. They have a smart healthcare group. They have a smart transportation group, which brings in smart cars. But that’s how these models are all being tuned. And the metrics they talk about, which is really interesting, is how do you measure whether the knowledge and intelligence of your model is increasing? And they actually have pretty good metrics they talk about, memory, creativity, content generation, reasoning. They have certain metrics they talk about and they look for a rate of advancement in those metrics within these industry specific flywheels. Anyways, that’s number four. That one’s a little bit complicated. For those of you who are subscribers, you’ve got four articles on this subject and two podcasts coming. So that’s six pieces of content. I kind of went over off the rails on this one. Anyways, that’s it. I’ll put the four pillars in the show notes, but here’s a recap. Number one, Baidu’s AI Cloud at its core is an innovation platform that is pushing as hard as possible to reach critical mass. Number two, Baidu’s AI cloud is also moving to become one of the technology standards deeper in the AI tech stack. Pillar number three, Baidu’s AI focused cloud, AI native cloud, has at least two advantages over other similar innovation platforms being built. And those were early mover and the fact that their AI native and their rate of development should be faster. And the fourth one, Baidu’s AI cloud also has what looks to be a flywheel in industry specific intelligence. And that is the content for podcast 200. That’s really crazy. I’ve you know what I feel I feel good about it. For those of you who are maybe back with me 200 podcasts ago, I did make some comments early on that I’m a big believer in setting a big goal and then turning it into a daily habit and just sticking with it for a long time. Like I’m a huge believer in like just daily discipline plus a daily habit. Like if you go to the gym every day for 30 minutes, in a year you’ll be in stunningly good shape. If you learn 10 words in Japanese every single day and you stick with it, two to three years later, you’ll turn around and be like, I speak Japanese. That’s kind of a lot of what this podcast was about is I created sort of a weekly, really it’s daily, you know, to do with a podcast and two articles per week, two to three articles per week for subscribers. It’s almost a daily habit where I have to sort of basically give lectures. And if you force yourself to give two to three lectures every week for three to four years about one specific subject, you get pretty good at it. Like it really does add up. That’s like my holy grail in life is daily habits plus time. Can accomplish almost anything. That’s my little strategy. Anyways, I hope this has been helpful. I’m gonna think about turning all of this. When I started this, I wanted to focus on one specific question, which was, “How do you build a competitive advantage in digital businesses?” Which I focused on for a long time that, you know, there’s six, seven books now about that subject. And a lot of the podcasts, that’s what that is. I’ve been looking for the next question for the next three to four years. And I’ve really settled on kind of the question that goes hand in hand with that, which is how do you build a competitive advantage in intelligence and AI? AI is sort of the next wave. Digital transformation and strategy was the first wave. AI transformation and strategy, which is really intelligent transformation strategy, is the next wave that’s built on top of that. And moats is a good way to think about it, but really it’s about winning. And that’s kind of what I advise companies on, which is look, how do you win? What does winning look like? Well, here’s about winning. And that’s kind of what I advise companies on, which is look, how do you win? What does winning look like? Well, here’s the path. If you stay on this path for one to two years, this is what winning looks like. Oh, and by the way, here’s what losing looks like. So it’s kind of the same thing, but that’s kind of the next question I think I’m going to focus on over the next two to three years is I’ll take all the existing frameworks I’ve been building for digital strategy and add on top of those extend those to AI strategy and intelligence strategy. And those things really go, they’re really not separate. They’re dependent on the same data. You end up building more architecture, but you can’t really do AI strategy if you don’t have a digital strategy. So it’s kind of the same thing Just sort of the next evolution. Anyways, that’s what I’m gonna focus on. I think for the next two to three years Which is super fun Like it’s just a blast and it’s so interesting Like there’s something cool all the time the business thinking is fascinating And then every now and then the politics aspects bubbles up. And I, you know, I will occasionally succumb to a bit of ranting in that area. But I try not to do that too much. I think it can annoy people. And I don’t necessarily think I know what I’m talking about when it moves into the politics. I think my thinking is fairly shallow. So working hypothesis is a better summary of that. Anyways, that’s where I am. I hope everyone is doing well. I hope your year is going well. If I can be of help, don’t hesitate to reach out. Especially when people reach out to me, maybe it’s an advisory, a consulting thing, or maybe they just want to say hi and talk about something. And they say, I’ve been listening to you for a year. It really means a lot. It makes my whole day. I feel great all day. I mean, that’s great. Now, if they say I listened to you for a year and I think you’re full of it, OK, then I don’t feel so good. But generally speaking, I get a lot of satisfaction at it. So anyways, don’t hesitate. And that is it for me. I hope everyone’s doing well and I’ll talk to you next week. Bye bye.


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.

Leave a Reply