This week’s podcast is about the arrival of ChatGPT, which is stunning people everywhere with its content generation.
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 Asia Tech Tour.
Here is the link to ChatGPT
Here is the past article on Conversation AI at JD:
From the Concept Library, concepts for this article are:
- AI as Cheap and Fast Prediction
- AI: NLP
From the Company Library, companies for this article are:
Welcome, welcome everybody. My name is Jeff Towson and this is the Tech Strategy Podcast where we dissect the strategies of the best digital businesses of the US, China and Asia. And the topic for today, chat GPT as the future machine human interface. Now you can’t have missed this story. It is absolutely everywhere. GPT 3.5, I think it’s 3.5. GPT-4, which is, you know, this is open AI, you know, the natural language processing, sort of algorithm AI giant, they’ve got their, you know, their new algorithm software and platforms being developed. GPT-3.5 is the current one, GPT-4 is on the way, which is supposed to be a huge step up, but there’s an interface that’s letting people engage with GPT-3.5, which is chat GPT. So people have been playing with it and putting in all sorts of queries about, write me a story, write me a conversation, write me a poem, random questions, random requests, and the answers have really been stunning in terms of text generation and created content by the algorithm. And what you’re hearing over and over, it’s on Twitter, it’s everywhere, what you’re hearing over and over is the phrase, this changes everything. And I think that’s probably true. And I think, at least in my life, this is probably the third time when I’ve seen something emerge and it’s like, I think this changes everything, maybe. Um, the other would maybe be the emergence of the internet, which I really distinctly remember when that happened in the nineties. Uh, maybe you could say automation and self-driving cars. That one’s not as, didn’t have such an impact on me. I’ve always remembered the internet. sort of discovering that in the Stanford computer lab back in the 90s and getting on Mosaic browser for the first time and seeing how everything connected and it was really quite stunning at the time. Didn’t know what it meant, you just knew it was huge. I get that same feeling about GPT. Anyways, if you haven’t played with it, I’ll put the link in the show notes. You can go over there and just start putting in random stuff and I’ll give you some… People are putting stuff up all over the place. It’s really fun. Anyways, that’s gonna be the topic for today and I’ll tell you what I think it means at least at this point and where the impact is gonna be and then big question mark outside of that. So that will be the topic for today. Let’s see, other stuff. One, I’ve been away for a little bit. I’ve been sort of a little sick and also I was just kind of tired mentally. I sort of needed to decompress. So yeah, so I’m back on now but I’ve been gone for a little bit. Sorry about that. And let’s see, Asia Tech Tour, that’s planned for March of this year, if you’re curious, there’s a link in the show notes to that. We’re going deep dive five days in Asia, visiting tech companies, doing a lot of sort of intensive digital training. Gonna be a lot of fun. Anyways, the details are in the link in the show notes. I think that’s pretty much it for housekeeping stuff. And for those of you who are subscribers, I sent you some basic stuff over the last week. I’ve been kind of mentally out of it. Not out of it. Let’s say it’s 75%. So we’ll be ramping up on Microsoft this week. That’s next on the list. And standard disclaimer, nothing in this podcast or in my writing website is investment advice. The numbers and information from me and any guests 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 GBT. Now if you go over to chat GPT and start putting in random stuff or just go on Twitter and you know look at what people are posting and it’s crazy. It’s like simplest things. Let’s say you want to have someone do your homework. You say write me three paragraphs on the history of Napoleon marching to Moscow. It will generate three to four paragraphs in about two to three seconds. That is incredibly well written. Not incredibly, very well written, pulling on all sorts of accurate information, citing dates, times, all of that. It’ll pull it from the web and Wikipedia and all that, and it looks great. So if you wanna have this do a lot of your homework, as a, let’s say, a high school student, or definitely below that, it could do it. If you’re a software coder, if you’re writing Python or Java or something like that, it can write code pretty easily. That’s just more text generation you put in, you know, please write a Python code that creates a sales list for a typical grocery, blah, blah, blah. And it’ll generate the code for you. So software coders are getting stunned by this. Definitely things like copywriters. You know, if you want copywriting for, I know a blurb for a book, write an introductory paragraph to this book. If you want copywriting for products in a catalog on a website, does all that really easily. You can do things like write some sales promotion material for the Nokia, so and so, in the style of a textbook. And you can tell the things like in the style of, it’ll do that beautifully. So, you know, right after the bat, knowledge workers, I mean, this is a bombshell. on a lot of knowledge workers. Because it does, anyone who’s generating content in terms of text, and this is natural language processing, so it understands, quote unquote, understands, and it can generate text. So, and within that you could put code. So, you know, first thing, knowledge workers, you know, this is pretty big deal. But then you can put in all sorts of other stuff. You can say, write me a haiku poem. about Ghostbusters the movie. And it’ll write a haiku and it’ll be pretty good. What are some of the, write me a conversation between Elon Musk and Nikola Tesla in the style of a Quentin Tarantino movie. It’ll do it and it’ll be pretty good. I put in a couple, I’ll tell you the stupid things, I’ve been doing a lot of stupid stuff just for, I don’t know, why not? I put in a couple stupid ones. So let me read you a couple of those. Here’s one I did which was, right, I basically just typed in, write me a story about Godzilla trying on bridesmaid’s dresses. I don’t know why, it just seemed funny. Here’s what it wrote, once upon a time in a land far, far away there lived a giant monster named Godzilla. Despite his fearsome appearance and destructive tendency, Godzilla had a softer side and a secret dream to be a beautiful bridesmaid. One day, Godzilla stumbled onto a bridal shop and, overcome with curiosity, decided to go inside, the shop owner, seeing the massive creature standing in the doorway terrified and ran for cover. But Godzilla was not there to cause trouble. He simply wanted to try on some beautiful bridesmaids’ dresses and see how they looked on him. With great care and delicate fingers, Godzilla began to rummage through the racks of dresses, searching for the perfect one. At last he found it, a flowing pink gown with lace trim and a delicate bow. He carefully slipped it on over his massive body and looked at himself in the mirror. To his amazement, he looked absolutely beautiful. Feeling excited and happy, Godzilla danced around the shop, twirling and spinning in his beautiful dress. He felt like a princess and he was overjoyed. But just as he was about to leave the shop, the shop owner emerged from her hiding place and saw Godzilla in the dress. She was horrified and ran out into the street. street screaming for help. Godzilla realizing he had caused a disturbance quickly took off the dress and fled the scene. He knew that he could never fulfill his dream of being a beautiful bridesmaids, but at least he had a moment of happiness and joy and that was enough for him. That’s what it wrote. Like one, okay, it’s really stupid. I know that, but it’s surprisingly well written for such a stupid little input. And that took all of about two seconds. Now I put in some better ones like, how do I get from Bangkok to the Intercontinental Hotel in Hua Hin? And it basically wrote four paragraphs giving, and I know how to go there. I go there to have drinks sometimes. It gave very accurate information on how to get there by a bus, by a taxi, by a plane, or by driving. It told me the highway, how long it would take, how much it would cost for all of these. and it was correct. That was kind of spooky. That’s why people start saying, this is starting to look like a search engine disruptor. I did things like, write me a one week itinerary for visiting Bangkok, and it will write seven days of things to do in Bangkok. And it is spot on for what the vast majority of travel agents and other books will say, it’s dead on. Another thing, I think I did one more stupid one. Yeah, I did one more really stupid one. I said write a story about a monkey flying out of Joe Biden’s butt It made me laugh. I don’t know why here’s what it wrote Once upon a time a mischievous monkey named Max lived inside the body of a man named Joe Biden Biden had been hiding in Joe’s body for as long as he could remember and he loved causing trouble for the unsuspecting politician One day Max decided to play his most daring prank yet As Joe was giving a speech to a crowded room, Max climbed out through Joe’s intestines and into his rectum. From there he used his powerful monkey legs to push and shove until he burst out of Joe’s butt in a loud explosion. The crowd was shocked and horrified as Max flew out of Joe’s butt flapping his little monkey arms and screeching at the tops of his lungs. Joe was surprised that he forgot all about his speech and just stood there with his mouth open in disbelief. But Max was having the time of his life. He soared around the room playing pranks on the people below and causing chaos wherever he went. He stole hats and glasses, pulled on people’s hair, and even threw a banana peel on the floor to make someone slip. In the end, Max got tired of his mischief and flew back into Joe’s body, where he settled down and went to sleep. But Joe was left with a red face and a lot of explaining to do as the people in the room whispered and giggled about the monkey that has flown out of his butt. I don’t know, is this an advancement of the human race? I don’t know, but I don’t think I could write that good of a story about that particular topic. So anyways, go online. You can find people doing all the sorts of crazy stuff with GBT. It’s kind of funny. A lot of interesting stuff. All right, let me get to the point. What’s the point? I think there’s at least two to three questions. that matter in terms of business and technology. Number one, there’s a lot of talk online about search engines. Is this going to disrupt search engines? Is Google search finally gonna get disrupted? I don’t necessarily think that’s true. My working conclusion is there’s really two ways we get most of our information. I mean, the issue is how do you get connect someone’s brain with all the information on the internet. How do you connect? And right now there’s really two mechanisms. One is search, go online search, which is an active activity. The other is some sort of newsfeed that you just sit there and it tees up information to you and you just sort of consume it, like the Facebook newsfeed, the TikTok newsfeed. Search plus feed are the two primary sort of interfaces by which we access the information. And if you go onto a company like Baidu or Kakao or any of these search engines and read their sort of 10 Ks, that’s what they talk about. We do search plus feed. We do search plus feed. And. That’s a pretty good summary of the two primary interfaces. Now in theory, what Elon Musk is doing with Neuralink could be a third interface. You know, you put the chip in your brain and you can access the information that way without a search and without a feed, which basically requires you to stare at a screen. So you get rid of the screen. That’s theory, well not theory, but it’s not here yet. So I think at the minimum we can say this is a third mechanism. by which to interface with all the information of the internet, still primarily through a screen. Instead of an active search, instead of staring at a newsfeed and scrolling down it, we will have this sort of question and answer interface where we ask things and it generates an answer. Tell me how to get from Bangkok to the Intercontinental Airport. And instead of giving me a bunch of links, it gives me three to four paragraphs as the answer. And it could, you know, once you have it in text, you can pretty much do it in audio form. So I think we’re looking at a third human machine interface is what we’re looking at, and way of accessing and engaging in information. To some degree that could disrupt search, maybe it will, but it’s definitely a third mechanism. So that’s sort of my first conclusion, second conclusion, which I mentioned before. You know, it’s not gonna hit search. When new technology like this comes along, the first thing it hits is expensive substitutes, which is humans. You know, this is just gonna be a ton of knowledge workers. We don’t need them anymore, or more likely we don’t need nearly as many of them, because one copywriter can do 20 to 30 times the volume that they used to do. One coder can suddenly operate with this and generate far more code than they could themselves. So that… dramatically sort of increases their capacity, prices are going to fall. It’s very bad for knowledge workers. Very, very good, but you know, very disruptive to them. So that’s second one. And then third one, which is my big conclusion. And this is sort of the so what for today. I think the reason this is having such an impact on people is because we are getting the our first real look. at what the human machine interface is going to be one day in the future. Right, we’ve always sort of seen Star Trek where there’s people talking in machines and machines doing it, but it was always kind of theory. And this is the first time I know of that we can see it in real life. That one day we’re gonna just be able to interact with computers this way, whatever we need to know they’ll tell us. whatever we want them to create, at least in text form, they’ll tell us, and once it’s in text, you can make it in audio form, so it’s pretty much the same thing. It just turns out that text, you can convey a lot more information more rapidly than through speech. I think that’s why this is so stunning, because we’re getting the first real look at what the future’s gonna be like. In reality, we can see it happening now. That’s kind of what hit me. Anyways, that’s kind of the first point today. Let me go through sort of, natural language processing as an idea and why chat GPT probably what the first three to four use cases are going to be, which is what I think is the interesting stuff. So let me go through that. Now the concept for today, I didn’t go through my standard here are the key concepts for today because obviously it’s artificial intelligence, natural language processing. And you know, this one I’ve talked about a lot, which is just the idea of AI as cheap prediction. I’ve talked about that before. It’s in the concept library. Uh, it’s, you know, it’s, this is just a, I think a good way to think about this. This is from, uh, Ajay Agrawal and Joshua Gans and Avi Goldfarb, who wrote a book called Prediction Machines. It’s a very good book. They’ve just come out with, and they basically talk, they’re business professors basically, and they talked about the economics of AI and their conclusion was AI is basically cheap prediction. And they’ve just come out with another book called, I think, The Power of Prediction. It literally just came out in the last two to three weeks. I flipped through it. It doesn’t seem to be too much different than the previous one. It’s more like an extension. But it’s pretty good. I’ll put the link in the show notes if you haven’t read that book. But their basic idea was, well, not their idea. Their point was, you know. AI is one of the few general purpose technologies that have ever been created. There’s about 35 of them in human history, electricity, the wheel, steam engines, the transistor, things that have gone into everything and that just changed everything. Well, they argue AI is basically a general purpose technology, which appears to be true. And what does it do? It basically makes prediction cheap and fast. Calculators, you know, made arithmetic cheap and fast. Used to be it took a lot of people with pens and paper doing lots and lots of calculations to add numbers together. Someone invented the calculator, then everyone could just add numbers, multiply, divide, subtract, add very, very cheaply, very easily on physical calculators, in computers, in chips. It’s, you know, and now calculation is in absolutely everything. So once you make something cheap and fast, it goes into everything and it enables all sorts of other things. Okay, this is the same idea, but instead of calculating two numbers you have, five plus seven, we’re doing prediction, which is about filling in information you don’t have, which is where we have four data points, but we’re missing one. Let’s predict what that is based on the other four. And… You know, that’s what business analysts do. That’s what a lot of knowledge workers do is, you know, you take a whole bunch of data from everywhere and you try to predict how many people are gonna buy cans of Coke in Bangkok next year. And that’s something you’d hire an MBA or an analyst to do. Well, that’s all prediction. And we use humans for prediction the same way we used to use humans for calculation. Well, it turns out software and data are getting real good at doing this on their own. And when something gets really cheap, you do a couple things, you put it into everything, and your traditional mechanism of doing that, your substitute, usually drops in price dramatically, which is what’s happening to knowledge workers who used to do this stuff. And the other thing that happens is your complements to that technology increase in value. Now, the main complement of AI is data. So as AI gets more and more used, data becomes more and more valuable and vice versa. So that’s kind of the basics. I’ve gone through this, you know, decent number of times. The pattern that these professors lay out, which I think is fairly good, is when this happens, prediction goes into traditional areas first. So you look at where prediction’s already happening, and then this technology goes right into there, which is. you know, market analysts, business analysts, people like that who are making prediction. Well, they’re all now making predictions on price and demand and logistics and things like that based on AI instead of, you know, analysts. Phase two is it starts to go into areas you never thought it would go into. When calculation got cheap, you didn’t just calculate things like your books like before, you started to calculate and do photographs that were digital. Well, cameras became digital, things like that. Didn’t really see that coming. For prediction, you know, that’s how Elon’s Musk’s cars are driving themselves, is they’re making predictions all the time. Well, we didn’t see prediction going into automobiles, but there it is. And then, you know, after there, it kind of goes from there. So that’s kind of the basics. I’ve gone through this quite a few times. Yes, I mean, it’s pretty devastating. to humans because in this case, we tend to be the main substitute. And unfortunately, there are, you know, there are a huge number of professional workers who generate content all the time. They do copywriting, they write catalogs, they do sales documents, they write, I don’t know, IT manuals, they write code, they do all of this stuff. Well, it turns out the GPT in natural language processing are getting really good at this. That’s gonna be a huge deal. So now that’s not awesome, but it’s hard, you know, it’s hard to look past the fact that a lot of knowledge workers, and I’m a knowledge worker, you know, we have never paid that much attention to when machines and factories and equipment displaced a lot of blue collar and factory workers over the past 50, 75 years. Um, so the fact that white collar knowledge workers are now getting displaced, uh, seems like a bit of karma. Okay. Let me switch back to GPT now, natural language. I mean, there’s a couple of big buckets of AI. I spend a huge amount of time studying AI. I mean, for me as sort of the digital strategy guy, this is, this is ground zero for me. I mean, I listened to a ton of AI in business podcast, trying to figure out use cases, what it can do. I mean, I’m learning to, I’ve been learning to code Python. in the evenings. I mean, it is just such a powerful technology and we’re seeing it everywhere in business. I’m pretty focused on, I’m more focused on this than anything else in terms of digital strategy. Okay, so we look at several buckets of AI. Computer vision is the one that people talk a lot about because it turns out AI is really good at seeing things. And also it was so directly monetizable and commercializable. that there was a lot of traction early on. You can put cameras everywhere, you can put cameras in stores, you can see inventory, you can see logistics. There were direct applications almost immediately. So it took off like a rocket ship and companies like SenseTime and the others have done very, very well with computer vision. Natural language processing was kind of the one that was also there, but people didn’t talk about very much. There weren’t as many obvious use cases. There were some companies out of Hong Kong doing this since 2013, 14. And they were making things like automatic translation devices, where you speak in the device, it translates it into text. There’s a couple versions of this audio to text, but we can see it online with Google Translate and translating one language to another. But it’s basically the way I’ve always thought about it is, OK, understanding text and then creating text. Uh, it’s about prediction. So the easiest way to think about it is I have a sentence, um, could be anything. You know, Jeff sits at his computer and talks in his podcast. And then I delete one word from that. Um, Jeff, you know, Jeff sits at his blank and then blah, blah, blah. That would be a problem for NLP to solve is you give it the data, which is here’s all the text, now predict what the missing word is. So fill in the blank. And this you could consider this like the easiest version of NLP. So what would it do? Well, it would look at tons and tons of data sets, which is easy to do because there’s text everywhere, right? You can scour the web, you can scour everything. It’s like one of the reasons computer vision took off so fast with JPEGs is there was a million, you know, not millions, a gazillion JPEGs on the internet that you could search. There’s a big data corpus. Okay, same thing with text. Tons and tons of text out there. It’s, you know, it ingests all of that and it starts making predictions. What is the missing word? But if you think about it, it’s not going to look at the words after the blank. Jeff sits at his blank. and blah, blah, blah, blah, blah. More than, it’s gonna focus mostly on the words right before the blank, because that’s gonna give you the greatest indication of what the missing word is. So this is, fill in the blank is like, you give the algorithm four words and it predicts the fifth, because we read in a sequence. But if you give it 10 sentences after that, there’s gonna be value there. but the value tends to be most impactful in the words right before the fifth word, the other four. And you could give it three or four paragraphs ahead of that and that will be valuable too, in terms of sort of pattern matching and things like that. But a lot of, you know, when I think about sort of NLP at the simplest version, I think about, I give you nine words, you predict the next one. Okay, well then you can take that to the next level. I give you three sentences, let’s say I give you 10 words and you predict the next three. That would be more advanced. You’d probably have to have some sort of understanding of what the content was. So maybe I got to give you three sentences and then the sentence and then three missing words or five missing words and then it predicts what that is. Or I give five to six sentences and then the AI predicts what the next sentence is. So it’s always sort of predicting what would come next after the previous words. And that’s kind of the easiest way I think about it. And another version of this that we see early on would be Q&A, which is kind of the same thing if you think about it. Now I give you a question, what day will my package arrive? And then the AI has to predict the answer. So again, it’s predicting the words that follow the words I gave it. Now keep in mind the AI has no concept of what we’re talking about. AI doesn’t think at all. It doesn’t understand at all. But it takes the words I gave it, what day will my package arrive, and then predicts the next six or seven words that would make me most satisfied with the answer or get me the truth, and it might be your package will arrive on Saturday. That would be a typical chatbot, which has been around for quite a while now. And they do what chatbots did, which is, you know, give an answer to a question, Q and A, and try to solve a problem. Where is my package? How do I make a return? I want a refund, things like that. But those are all sort of, in some sense, they’re the same as the fill in the blank answer, which is I give you a certain number of words. You predict the words that come next. Such that. we have a happy conclusion the problem is solved. So the first chat bots, you know, it was a lot of basic stuff and then it sort of moved into what we’d call task completion, which is kind of the example I just gave you. You know, that’s 1990s, 2000, 2010 up until then. Let’s say 2012 ish going on where we start to get things like Siri, where suddenly you’re getting personal assistance where now it’s trying to do more than complete your task. And now we’re starting to get 2015 onward, we get social chat bots. Now, about a year or two ago, I did an interview with He Xiaodong, who was the head of basically AI at JD. I’ve written it up, I’ll put the link in the show notes. But he talked about what they call their dialogue systems, which is what, you know, the newest chat bots at JD. And they don’t call them chat. bots anymore. They call it conversational AI. And basically, you know, if you go online and the little dialogue window pops up and you say something like, where is my package? It’s not just going to give you a simple task answer. It’s trying to one, well it is trying to solve your problem. So it is task oriented. Then it wants to identify your problem and solve it. And it is optimized towards that as an outcome. You know, it makes a prediction based on a certain specified outcome it is given. And the outcome it is given is to solve the problem, task-oriented. However, what he was talking about was like, look, we don’t only have one outcome now, solve the task, we have two outcomes, which is what we call emotional satisfaction, being an emotional companion. We don’t want just the task to be solved, your package will arrive on Friday. We also want you to be in a good mood at the end of the conversation. So we’re measuring how you’re responding and how you’re talking, and we’re trying to infer your emotional state, and then we’re replying in ways to try and put you in a good mood. So at the end of the conversation, your task has been solved and you are in a good mood. So that’s when it starts becoming sort of a social companion or an emotional companion, or what they call conversational AI. And Part of the way it does that is it might generate statements like your package will be there on Friday, that’s text generation, but it can also generate things like poems and jokes and chit chat and it can make photos and their conversational AI can do all of that as it’s part of the interaction. Now that’s all pretty similar to GPT in some sense. It’s… You know, that sounds a lot more like what GPT is doing. It’s not just trying to give you a simple answer. How do you go from Bangkok to Hua Hin, Intercontinental, it’s trying to come up with a story that I kind of think is funny or interesting. So it’s obviously optimizing on a couple of dimensions. And one of them is just like, hey, this is fun. Is it interesting? Do they engage if we do this? So they’ve basically given their chatbot emotional intelligence and What he was saying, which brings me, I guess, to the final point for today, is that conversational AI is not customer service. It’s much more than that. It’s not just about solving the problem. It’s more than that. It’s not just about making you sort of in a better emotional state. He called it basically a new machine-human interface that you are engaging with the computer. and the computer can engage back with you and it can try and engage with you emotionally, it can engage functionally, it can engage on lots of dimensions. And it’s really about the interface and the ongoing interaction because when you log off and log back in six months later, it remembers everything you talked about. It remembers everything you said. It remembers everything it said to everybody. So it really almost starts becoming a relationship that can go on in time. between this machine and you. So he kind of called it, you know, this machine human interface that you can have with the system. In this case, it’s JD’s conversational AI. But you could put this capability into products where you could have a conversation with the new Nike shoe and it could act in certain ways. It could have a certain personality, could be sarcastic or whatever. You could put this into anybody’s website. And what’s interesting about GPT, which people haven’t, I haven’t really had people talk about this so much. I mean, GPT is basically an API, right? They built their system, but they have an API and you can plug it into anything. And the amount of memory it takes to plug this into something is on the order of a hundred megabytes. I mean, it’s really surprisingly small. that you can put this unbelievable capability into anything. You could put it into your laptop, you could put it into your mouse, you could put it into your microphone. Obviously I’m looking at things on my desk. That’s when it starts to be like, this is starting to look like sort of a universal conversational AI capability that’s gonna change how we engage with all the information on the internet and other things. That was kind of my point for today. That’s what this looks like to me so far. And we can use it to write fiction, we can use it to write articles, we can use it to write letters. So and so AI, write a note to my mother apologizing for being late last night to dinner. And it’ll shoot it back in two seconds. I take a look at it and I send it. I can say, AI, please write me a Python code for my website that does this. It generates it, it puts it, I look at it, I tweak it, I put it up a little bit. Please write an introduction to my podcast, blah, blah, blah, it does that. I mean, you can just see the use cases everywhere. The one I’m actually paying attention to is Microsoft PowerFX. You know, they’re, I’ll write about this in the next couple of days. You know, they’re low code and no code, basically developer tools that lets anyone sort of become a developer. by giving them tools that can basically code pretty much on their own. And that’s pretty amazing. And this really adds to that. And then there’s a very important relationship between Microsoft and OpenAI. OpenAI is a bit of, as a company, it’s a bit of a question mark. There’s this idea that, oh, it’s open source and it’s a foundation. No, no, no, it’s actually not. I mean, it’s a company. and they’re letting people use the APIs, but the only group I think that is licensed to their core technology, and the only people that can see their core technology as far as I know is Microsoft. I think they hold the only technology license. To everyone else, it’s a bit of a black box. But Microsoft, I would expect them to go put this right into their Power FX, which I assume is what they’re doing. And I’m gonna write about that this week for those of you who are subscribers. Anyways, okay, today’s… bit of a high level discussion. I’m still trying to get my brain around this, but I think this is really worth your time to go over, use the link in the show notes and play with this a little bit. If nothing else, it’ll probably blow your mind. It really is pretty phenomenal. And let me end it there for the content. I’m still sort of ramping up here myself. So we’ll just keep it short and do that for today. Yeah, I was under the weather, which was annoying, but not serious. So I was pretty much camped out in the apartment for a week, not doing much. I ended up buying the condo next door, which was really not the plan at all, but it just sort of became available and I sort of, yeah, jumped on it. And that actually may turn out to be a spectacularly good decision. We’ll see how that one plays out in time. So that was me, yeah, sick for the week, just. emailing and texting back and forth trying to get that done, which it actually got done pretty quick, which was nice. Yeah, so that was my week. Anyways, that’s it for me. Next couple days I’ll be back at full speed and articles and podcasts should be coming on a more regular basis going forward. So that’s it for me. I hope everyone is doing well and I will 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.