Takeaways from the Largest Tech Conferences in the World with Jim Harris, Keynote Speaker & Best Selling Author

(Intro Narrator at 00:00:00) Today, we're talking to international bestselling author, Jim Harris, about his key takeaways from CES, Davos, and Mobile World Congress. You're listening to Joel Beasley, Modern CTO.

(Joel Beasley at 00:00:19) This is, like, time number 700 for you being on the show, so that's good fun. What have you been up to this past year? Tell me about it.

(Jim Harris at 00:00:27) Well, every year I go to three shows in Q1. So I go to the Consumer Electronics Show in Las Vegas, early January. I go to Davos, to the World Economic Forum. That's in mid-January. And then I go to Mobile World Congress, which is in Barcelona, Spain. And the insights from these three shows are, for me, mind-blowing.

(Joel Beasley at 00:00:54) What are you learning? What happened at the first show?

(Jim Harris at 00:00:56) Well, at the first show, so at the Consumer Electronics Show, the focus of all three shows was really AI. AI, AI, AI. And so at CES, the CEO of Intel, Pat Gelsinger, was on stage, and he said, "We're seeing productivity use cases for AI of 10,000x, 10,000x on productivity." And so I began thinking, is Pat just hyping it? Because we've all heard of the Gartner hype cycle. You know? So is he just hyping it, or are there these kind of use cases? So this will prove out Pat's point or disprove it here. So the first point is that Cathie Wood of ARK Invest on Wall Street—I'm a big Cathie fan—she has said her team has said in their Big Ideas report for 2024 that they're seeing use cases of 10x productivity for using AI for programming. So if you code in Python or JavaScript or whatever, I've even seen applications where you just draw, you take a screenshot of a website, and then you slap it in and say, "Write me the HTML to give me this site," and boom, it does it. Now you might have to correct it, but these are kind of incredible use cases. Right? So 10x is the low use case for Pat Gelsinger's 10,000x.

(Jim Harris at 00:02:37) So mid-range, I was at Davos, and I put Pat's thesis to a panel. And this woman said, "Well, I'm in marketing. And the way we used to work in the advertising industry was a client would come in and give us a brief, and then we'd go away and we'd draw Bristol boards of images, and we'd get cut lines and slogans, and it was kind of like Mad Men. And then we'd come back to them and present them this stuff." Well, that process took us two weeks. So now I, all my own, not a whole team of us, with ChatGPT and DALL-E or Midjourney, I can do all that in two hours. And I did the math out on that, and that, depending on how big your team is or how many options you gave them, is a 500x to a 1,000x productivity increase. So that's kind of Pat Gelsinger's mid-use case. And then for the high end, 10,000x productivity, you think, "Oh my god, that is way out there."

(Jim Harris at 00:03:48) So this year at Mobile World Congress, I listened to the co-founder and CEO of DeepMind. Now you'll think back to DeepMind in the game of Go, beat the world's Go champion, Lee Sedol, which nobody thought was possible back in 2016. And for the CEO, playing games was a very important stepping stone in the AI progress, right, to understand how AI works, to refine it, to improve the code, the algorithms, the thinking. So it was bought by Google, and they created something called AlphaFold, F-O-L-D, and it studies proteins. Now, Joel, our bodies are built on proteins, and you can think about it as a long string of amino acids. And each one of these folds a different way, you know, based on different proteins. And so these are the basis of all life, and there's been a 50-year challenge to understand how every single protein folds in the body so that we can understand how they interact with other proteins. Well, like a PhD student will take five years of his or her entire career studying one single protein to understand it. One protein, five years of PhD time.

(Jim Harris at 00:05:25) AlphaFold is able to do that now in seconds. And so there are 200 million proteins that are the basis of all human life, all animal life, and AlphaFold has studied with, has predicted with absolute precision how each one of these will fold. 200 million of them, and it did it in one year. So that is a billion x productivity increase. I feel like, Joel, this is a Dr. Evil moment. Billion x productivity increase. So, you know, some people say, "Is AI hype?" And I just want to say for some companies that I saw at all three events, they're talking a really good AI game, but they're not really serious about AI yet. But the reason they're talking the game is they don't want to be left behind. They don't want to bleed customers. They want their customers to have faith in them and continue to subscribe to their SaaS products. But some are not really walking the walk. So in one sense, yes, those companies are creating hype because they're creating disillusionment. You know? You say it's going to be great and you haven't given me any new features or benefits. Point taken. But on the other hand, when you can say we have a billion x productivity increase, and it's, by the way, a billion years we didn't have to spare. Like, this was not a real exciting thing for a PhD student to study a single protein for five years. Now what AlphaFold is doing is saying, "Okay, how do we take that perfectly predicted structure and say we want to figure out a pharmaceutical that will only attach to this one part of the protein and block it from doing something to prevent cancer at stage zero?" And the reason for only attaching to one point is we don't want side effects. So, you know, we want to block the protein right here in the 3D structure. So now we stand on the shoulders of that billion years of research. So when people ask me, "Is AI hyped?" The answer is yes, because some companies aren't really walking the walk. Is AI offering us huge promise? The answer is yes. It's kind of like a Zen koan, you know? Yes. Yes. Yes. You know, it's both. So these were three really profound events that gave me such incredible—it's such a privilege to go and schmooze with the world's leading thinkers and for myself to have time to think about these issues and really kind of distill what is important and what's going on and cut through the hype from what is really going to change our world.

(Joel Beasley at 00:08:29) Oh, you articulated that extraordinarily well. I was over here taking notes and I'm like, "How much time did he spend to craft this?" I loved the story that you told, but it's a good point. And I'm having these conversations multiple times a week, and I notice a lot of the times I get people that, there's like, "Oh, you know, it's not, it's not what it's lived up to be, and it's just, it's only like 80% of the way. They can only get you a little bit there," and they're very dismissive or downplaying it in very narrow and specific use cases. So I think the first problem that we have is you got a 10,000x productivity on, in what capacity, in what niche, in what genre? Like, are we talking we took a third world country, group of a thousand people who were doing effective OCR and just replacing that? Because if you're such a large company, you've got all of these different situations. So I'm interested to know, like, where are the productivities happening? Like, where is that happening? In what markets and for what specific skills? And so when he goes on stage and says 10,000, you know, does he have context behind that? Or no, he just says 10,000, we're seeing a 10,000 productivity across, like, as a company?

(Jim Harris at 00:09:50) So going to the marketing and advertising example, which was the mid-use case from 500 to 1,000x, she made a really good point that this is just not the entire process, but it was a part of the process. The process of taking the client briefing, breaking it down, doing the creative, and then coming back to the client and presenting it. It wasn't the marketing part, the acquiring the customer, the winning the bid. It wasn't the billing. It was only part of the end-to-end process. So her point was you're not seeing 500x or 1,000x for the entire process. You're only seeing it for part of the process. So that's one thing to note. The other thing to note is in the case of the AlphaFold, like, this isn't replacing people because we didn't have a billion hours to begin with to study protein folding. This now allows us to take information like a 1.0 and take it to knowledge, which is, "Okay, how do we see these exposed parts of the protein that might be causing cancer and address them with pharmaceutical products and predict which pharmaceutical products will work best?" So it's enhancing something we were never able to do before, or if we were able to, we would have had to spend a billion hours to get there. So we're not replacing anything there. We're creating new value that could never have existed without AI. So it's really dependent on the use case. So we have to look on a case-by-case basis, but I'll just give you an example.

(Jim Harris at 00:11:41) Last night, I used Otter.ai. It's a transcription service. You know? And I slapped in some video that I'd created in response to these three events, and it gave me a word-for-word transcription. Now I used to have to hire somebody and give them an hour-long video, and they'd take, you know, a few days or a week, and they'd give me back a transcript, and I'd pay them. Well, this gets done instantly, and I just have to go in and correct it. So it increases my productivity, and it lowers my cost. So I think you have to look at AI on a case-by-case use case. There's one other one I've got to tell you about, though, Joel.

(Jim Harris at 00:12:28) So I was in Barcelona, and there is an influencer on Instagram who's called Aitana López. She is—maybe afterwards, we can cut in the image of her—but she is a pink-haired, 25-year-old influencer. You know it, Aitana?

(Joel Beasley at 00:12:51) She's the AI one, right?

(Jim Harris at 00:12:53) Yeah. She's not real. Yeah. But I went and I interviewed the team of three that created her. I did an hour-long video.

(Joel Beasley at 00:13:02) Oh, nice.

(Jim Harris at 00:13:02) And we haven't produced it yet, but it was fascinating. So she has—when I started following her, she had just over 225,000 followers on Instagram. Now she has 300,000. And they thought about what to imbue her with. So she's pink-haired. She's—oh, there she is. Yeah. This is great. She's pink-haired. She's 25. She likes gaming, like e-gaming. She's into fitness. So they created her with a profile in order to market. And companies like Guess Jeans and Victoria's Secret and others pay her $1,000 per Instagram post to get on her Insta. Now if you think about it, you don't have to pay a model and fly them to, you know, Jamaica for a photo shoot and have a team and, you know, it rains on the day you want to shoot and there's a hair and makeup person. And, like, so all of that changes. Plus, your influencer doesn't say something that's ludicrous and gets the brand in trouble two years down the road. So it is really interesting, and she is hyper-realistic. Like, when people see her, when I saw her initially, I thought, "No, that's a real person." And the funny thing on her account, if Josh in the background can follow this, there was an Italian influencer on Instagram who came to Barcelona to meet her friend, Aitana. So there you have the two of them face-to-face about an inch apart at the nose. Yeah. There's kind of a little—there it is—a little sexual tension there.

(Joel Beasley at 00:14:58) There's a lot—this is—I actually, we did, we're getting automated out of existence, and we did a special video on influencers getting automated out of existence. And all the research that we did, my editor is like, "I don't feel comfortable, like, producing this video." He goes, "I can't have these pictures on my screen, like, during the day with my wife and I'm doing my video editing." So we trashed the video because the influencers were overly sexualized.

(Jim Harris at 00:15:25) Yes. Well, I've got to say, Aitana is very sexualized. Do not miss that out. But that goes to a wider question, like, look at the front cover of Vanity Fair, you know, or Sports Illustrated swimsuit edition. You don't—I mean, it's wider than just the influencers being sexualized. It's the entire media industry, which is a bigger critique. So, you know, I'm not—it is an interesting discussion on the wider societal issues. But I'm just looking, "Okay, this is what exists and this is what AI is doing." Now when we saw that pairing of the two, the one on the right was real and the one on the left was virtual. Now I interviewed Sofía, who's one of the team of three in Barcelona, and she actually stood in front of Elena, who's the Italian influencer, with her nose one inch from each other. And later there's a picture where I think Aitana is kissing Elena on the cheek. And so I said to Sofía, "So, so how was it?" You know? And she said, "Uncomfortable." But the AI takes the real person and turns it into Aitana, the pink-haired influencer. So this I just found fascinating. So it's a whole hour-long discussion with the team that created Aitana. Nobody's ever done that, and so I have to cut the whole video and everything and title it and put it in snips and then put it up somewhere. But it was just really interesting. And, you know, why is this innovation coming out of a little ad agency that's three persons in Barcelona, not out of New York, Tokyo, or London? And so that's question number one. So why is innovation happening at the edge as opposed to in the center of organizations?

(Joel Beasley at 00:17:31) It's risky. That's why it's happening. It's a cottage industry. They're going to do amazing things, and then they're going to watch how the public reacts because, hey, this little three-person agency, if they get smashed and everyone hates them for taking influencer jobs or whatever it may be, whatever the narrative might be, they get the brunt of the force and you get to see what happens. If everybody embraces it, then you have a business case for approaching it. And even if you do get burned, at least you had some success you based your decision off of.

(Jim Harris at 00:18:01) Exactly. And this is why innovation typically occurs at the edges, because this group of three young people is curious, and they don't feel they have a lot to lose. So they innovate at the edge.

(Joel Beasley at 00:18:17) They actually don't, though. They technically don't have a lot to lose.

(Jim Harris at 00:18:21) Exactly. Yeah.

(Joel Beasley at 00:18:21) You know? Yeah.

(Jim Harris at 00:18:22) It's three people in one office in Barcelona, in the north of Barcelona. Yeah. So it was really interesting for me to interview them and have some insights around this. But the reason they began is the price of influencers for some of their clients, for some of their brands, had got so high. You know? And it's not just the $2,000 you pay to the influencer. Again, you have to fly them to Jamaica or do a shoot in Barcelona, and you have to hire, you know, the hair and the makeup and the person with the fans so the hair is blowing back in the wind and all that.

(Joel Beasley at 00:18:58) My job. I do the fans. I operate—yeah. Yeah.

(Jim Harris at 00:19:00) Well, you know, you want your beard flowing, Joel. You know?

(Joel Beasley at 00:19:05) I visited, real quick, I visited a mining, a crypto mining facility that had 42,000 computers in it. I'll send you a picture later.

(Jim Harris at 00:19:12) Oh my god. Where was that?

(Joel Beasley at 00:19:14) In Texas. And this is, like, three weeks ago. And they have two rows of these computers, and those are inside and air-cooled, and then the exhaust fans from both sides are pointed into, like, a center hallway, which is like an atrium. So there's an open-air exit, so all the exhaust comes out and goes up and dissipates. And we weren't allowed to take pictures on the inside. We had to take pictures in that middle section. It was about 150 degrees. We stood there for like 15 seconds to get a picture and my beard's like this because there's so much wind from the fans in the—I'll send you a picture of that.

(Jim Harris at 00:19:50) I want a picture of your beard.

(Joel Beasley at 00:19:52) Yes. Yes. Yes. But that, yeah, that was a lot of fun, man. The mining stuff's getting crazy with the energy, but that's a whole separate conversation.

(Jim Harris at 00:20:01) So this is really interesting, though, and some people just dismiss an entire new area of innovation because they don't understand it or because they're worried about some dimension of it. So, for instance, you know, crypto mining takes so much compute power that we can't do crypto or cryptocurrencies or cryptography. The same critique is aimed at AI. But if you think about Ethereum, Ethereum made a shift to a new way of achieving consensus that reduced the amount of energy use on servers by more than 99%. So these critiques assume that things will stay the same as opposed to innovate and change.

(Jim Harris at 00:20:52) And we're in such a fast changing environment with AI that we can't assume that the AI of the future is gonna look like the current AI.

(Joel Beasley at 00:21:04) That's a — we live in a capitalistic economy. And I mean, it's a weird concept to start saying, like, hey, we're gonna throttle power usage to people who are mining because we don't think it's efficient. It's like, well, what about every business? Are we gonna tell the bakery they can't have power because they're not efficient? Are you gonna tell the compute data centers that because they're processing adult content, they can't have power?

(Joel Beasley at 00:21:24) Right? Like, when are you gonna start deciding who gets power? It should be a free market. People will buy on the market, and that market will then use that money to expand and meet demand. I mean, just let the market operate. I'm a — I know you're in Canada, so there's a lot of government over there, but I'm a big fan of, like, really small government and let the market play out and put some regulation to, like, keep things from going, you know, getting oil in our water and stuff like that.

(Joel Beasley at 00:21:51) But we should — yeah, the amount of — since I last saw you last year, I started investing in mining infrastructure. So not the buying the cryptocurrency itself, just the infrastructure for mining and power and all of that. And what you see in the headlines on the surface versus what the reality is is vastly different.

(Joel Beasley at 00:22:13) Like, it is amazing to me how we can then start making policy on things that are, like, not even close to reality and not even close to truth.

(Jim Harris at 00:22:24) So this is an issue of framing, and we have to be able — one powerful tool for me is to say there are waves in the ocean, which we all can see because they're at the surface, and then underneath, there are deep currents. And what we get in the headlines of newspapers and on radio stations and on TV reporting are the waves. But what are the deep currents that are changing? And by deep currents, I mean, if you wanna look at the total number of transactions using crypto, if you wanna look at the hash rate, all these trends are, like, up, unending regardless of the price of Bitcoin. The price of Bitcoin is the waves.

(Jim Harris at 00:23:18) Right? Sometimes it's high. Sometimes it's low. But the number of programmers who are focused on cryptocurrencies and on, you know, things like Ethereum has been going up all the time. The amount of reusable code in this space has been going up over all time.

(Jim Harris at 00:23:37) The number of wallets globally has been going up. These are the deep currents under the ocean that don't get play in the headlines. It's like Bitcoin's up today, it's down tomorrow. Like, that's what we read about as opposed to the fundamental metrics of what is the transformation and adoption in society. And so I think it's incumbent on us, people, to sort out the wheat from the chaff, the waves from the currents.

(Jim Harris at 00:24:11) And for —

(Joel Beasley at 00:24:11) For the people who don't, great. Because that's where the — that's where the best returns are is before everybody figures it out. Like, I realized when I started uncovering all of this, my instinct was to go tell everybody. Like, this is gonna be great. This is gonna be awesome.

(Joel Beasley at 00:24:25) Like, the world needs to know this. And then I realized really quickly a lot of people just shut stuff down. Like, oh, too good to be true or it's just like whatever or that's risky. And I'm like, cool. That's great because I'm just gonna keep making outsized returns on my money, and because I understand it. I understand how some places have stranded energy you could buy for 3 cents a kilowatt hour and put mining rigs on it and then profit while everyone else is paying 12 cents a kilowatt hour, you know?

(Jim Harris at 00:24:55) So this is — you have to be willing to dig into issues to do what you're doing, and you have to be willing to not focus on the waves to get at the — what are the fundamental underlying currents that are gonna change things forever? And so these use cases we've been talking about AI for me are the deep currents. We're gonna see more and more AI applications. Um, you know, I love playing with DALL-E and Midjourney. It's not just text like ChatGPT that is, you know, creating —

(Joel Beasley at 00:25:37) And I haven't played with Sora yet.

(Jim Harris at 00:25:39) Yeah. But I want to. It's entire video, of course, that we're seeing. And then the questions for society as a whole become really important. Um, you know, if I can stand in front of a video camera for ten minutes and speak a predetermined text and be recorded, and then the AI is trained on how I speak, my hand motions.

(Jim Harris at 00:26:10) And then I can feed it a four-hour training script, and it will create a set of training modules over four hours that look exactly like me that you cannot tell it's not me delivering that four-hour set of training. What are the implications that then I take that four hours of training and run it through a different AI that now it's — I'm speaking Chinese for the four hours, and then I'm speaking German for the four hours. And you might think, oh, Jim, this is so futuristic. No. It's not.

(Joel Beasley at 00:26:45) No. It's happening today.

(Jim Harris at 00:26:47) Right now.

(Joel Beasley at 00:26:48) Yeah. It's happening in learning and development parts of organizations.

(Jim Harris at 00:26:51) Yeah. So there's an amazing use case right there, training and development. Boom. And, you know, you can use GPT for flavoring things different ways.

(Jim Harris at 00:27:06) Oh, Gen Zs learn a different way. Okay. Take this four hours of training and Gen Z it.

(Joel Beasley at 00:27:12) Oh, no. Oh, no. Uh-oh. Yo. Yo.

(Joel Beasley at 00:27:16) Yo. Have you seen the Gen Z Bible?

(Jim Harris at 00:27:19) No. I haven't.

(Joel Beasley at 00:27:20) Oh my gosh. They translated it into, like, a Gen Z, X, or Y Bible. And, like, Jesus talking is just hilarious.

(Jim Harris at 00:27:30) That's hilarious. I gotta get that. I gotta look at that.

(Joel Beasley at 00:27:32) Just watch a couple shorts on YouTube because they're — all these shorts are going viral of Gen Zers reading the Gen Z Bible, and it's just hilarious.

(Jim Harris at 00:27:43) I love it. I love it. But if you think about it, how do we use these tools to tailor to our markets? Because marketing in the old days was one size fits all. And, like, who voted for one size fits all socks?

(Jim Harris at 00:28:00) You know, I certainly didn't vote for that, but really, it was the ease of production for the manufacturer. Like, we're gonna have a standardized sock. Well, now with AI, I can market discretely with zero incremental cost. So I can have an ad for Gen Zs, one for young millennials, one for mid millennials, one for older millennials, one for, you know, Gen Xers, one for boomers, you know, boom.

(Jim Harris at 00:28:30) And I can tailor these basically instantly.

(Joel Beasley at 00:28:34) We're doing that right now at our company with cold email. And so that just gives us a small advantage to, you know, slightly better than just a general cold email. And so —

(Jim Harris at 00:28:43) So this is a really intelligent application of AI. It's very highly specific, and it's very intelligent. This is where people are gonna get great ROIs. I wanna back up for one second to something from Davos, which this has made me think of. They did a survey.

(Jim Harris at 00:29:06) This was Marsh, McLennan, and Mercer, and I was there with the partners. And they said, we went and surveyed CEOs, and 40% of the — or CEOs felt that 40% of their people, their workforce, would need training around AI within the next five years. They then went to the people in the organization, and more than 90% of them said we need training right now. So there is a huge disconnect between what leaders think people need and what people say they need. And one of the reasons we fear things as adults is we don't know them, we don't understand them, and therefore we fear them.

(Jim Harris at 00:29:53) Our natural bias as a human, it's a self-preservation mechanism, is to fear new things. Like you go out in the plains of Africa and will this saber-toothed tiger eat me or is it a friendly kind of animal? Like the natural human instinct is fear first. So until people in organizations have training and the level of training is exponentially greater, the World Economic Forum says people need twenty more days a year of training. So this is one of the huge implications for this AI environment, this rapidly changing environment that we're operating in.

(Jim Harris at 00:30:34) And ChatGPT is the fastest adopted technology in the world's history. A hundred million people began to use it in the first sixty days. Like, there is another Doctor Evil moment. A hundred million people used it in the first sixty days. So, you know, I think it's back to being analogous to the web revolution in '93 with Mosaic and then Netscape, but it is turbocharged.

(Jim Harris at 00:31:03) You know, it took years for the web to reach a hundred million users. It took sixty days for ChatGPT. So we have to pay attention to this. The rate of change in our business environment is far faster than we've ever had. So, you know, you mentioned Sora.

(Jim Harris at 00:31:24) Like, Sora is just brand new. So who's using Sora, right, you know, to actually create video now, just like ChatGPT creates text for us, and just like Midjourney and DALL-E create images for us. It's just taking it to one more level. But who's playing with it? If you're in video, you'd better be playing with Sora.

(Joel Beasley at 00:31:49) Yeah. I wanna short the iStock photo stock. You guys are gone. I mean, you might end up being the intermediary that gets really good at helping people do text to prompt, like, and then licensing it, but your margins are gonna get crushed. It's gonna be highly consolidated.

(Joel Beasley at 00:32:06) I did some math the other day, and I believe it'll be anywhere from three to five years before the number of artificial intelligent agents outnumber humans on the planet. And their economy, because agents are gonna task other agents autonomously and transact business with us, with themselves and with us, will dwarf the size. So we'll essentially — be like an alien species that takes up the footprint the size of a state, but has an economy 10x of the human species. So I think that that's coming, and I think that the banking system they're going to use is the blockchain banking system.

(Joel Beasley at 00:32:43) Yeah. Because they don't need KYC. They don't need to be people. They just can go interact with the system properly.

(Jim Harris at 00:32:48) Way more secure.

(Joel Beasley at 00:32:50) Mmhmm.

(Jim Harris at 00:32:50) So this leads me to a saying, which is the future is already here. It's just not evenly distributed. So I have been talking about autonomous fleets of vehicles. So think about Uber and Lyft, but with no drivers for about ten, twelve years. There's video.

(Jim Harris at 00:33:09) You can go back and look at it. And people have been saying to me over the past ten or twelve years, oh, Jim, that's so unrealistic. That will never happen. Well, if you go to Phoenix, Arizona, you will see Waymo vehicles, taxis driving around with no driver.

(Jim Harris at 00:33:28) It's the weirdest thing. You know, and I'll embed this in my presentation. Same with San Fran and some other places. I think Austin might just have approved it. So you can actually go to Phoenix and see this. So —

(Joel Beasley at 00:33:41) Oh, I've seen it. One of our employees took video and did it.

(Jim Harris at 00:33:45) It's so fun. Yeah. The future is already here. It's just not evenly distributed. And the implication of this for leaders is you can't be, you know, parochial.

(Jim Harris at 00:33:58) You can't say, well, the only thing that matters is what happens here in Nashville or what happens here in Toronto. You have to be searching the world for case studies as to how this technology will affect your company, your industry. In other words, you have to be forward looking. So just because Waymo isn't in your town yet doesn't mean when we hit some tiny little change in price structure, it isn't everywhere overnight. And this is how blindsiding or the title of one of my books, this is how disruption happens.

(Jim Harris at 00:34:38) It's under the radar and then it overwhelms us.

(Joel Beasley at 00:34:43) That is exactly the case. Yes. That's true with, like, everything, though. Right? A lot of influencers are like that.

(Joel Beasley at 00:34:52) You're like, you learn about them, and you're like, oh my gosh, they've got all this stuff. And then you go back, and they've been doing it for a decade. And you're like, oh, okay. Alright.

(Joel Beasley at 00:34:59) Alright.

(Jim Harris at 00:35:00) Yeah. So — and so to this point, there's something called a cost curve. We know about exponential upward trends like Moore's law, which is the doubling of processor power every 18 months while staying at the same price point. So everybody's heard about Moore's law, but when you actually dig into it and study it, it's pretty mind-blowing in the sense that way back in the nineteen seventies, an Intel chip had 2,000 transistors on a CPU. And today with the Apple M1, it's more than 114 billion.

(Jim Harris at 00:35:37) So from 2,000 to 114 billion, and that profoundly changes the cost curve on compute. So a gigaflop is doing a billion floating point transactions in a single second. And back in 1961, a gigaflop was $164 billion. And today, it's a single cent. And that's a 2016 or so, 2024 figure because it's going down from there.

(Jim Harris at 00:36:09) And so this is one of the reasons we'd have an explosion of AI because it's a compute-intensive process, same thing with cryptocurrency and crypto mining, blockchain mining, very compute-intensive. But because the cost curve of compute power has fallen so profoundly, it makes it economically viable. Now even as such, you know, blockchain and cryptocurrencies use about 2% of the global electricity. So it's still a horrific huge amount, but it will come down because of the example we talked about with Ethereum changing their consensus model and reducing energy use by 99%. So we will continue to see innovations.

(Jim Harris at 00:36:56) But the whole point I'm trying to make here is that if you understand a future declining cost curve and you see that the traditional industry crosses or the new industry crosses over the traditional one. For instance, the cost of electric batteries have fallen by more than 90% since 2010 and all of a sudden EVs become cheaper than gas cars. If you understood that declining cost curve, you could predict with accuracy that EVs would eventually drive traditional cars off a cliff. And we're not quite at the point yet because the production side of EVs can't meet demand. Like, even though the wave stories in the media are declining demand, once Tesla brings on its $25,000 EV, the Tesla Model 2, you'll watch demand spike again.

(Jim Harris at 00:37:57) And once BYD out of China brings vehicles over here that their starting price is $15,000 US, you will watch the traditional car companies fall off a cliff. So people are saying, oh, EV, you know, the waves. Tesla's on a downtrend right now on its stock price because its earnings are being compressed because it is lowering prices, but it's competing against BYD. Wait until the Model 3 comes online on production. The Cybertruck, it will pick back up.

(Jim Harris at 00:38:32) But right now, it's in the down phase. I don't recommend buying it until it's bottomed out. But people confuse those waves with the deep trends. The deep trend is the declining cost curve on battery price.

(Joel Beasley at 00:38:45) That's why I invest on principles and not on waves. Yes. Yeah. Wow. That's a lot to think about.

(Joel Beasley at 00:38:52) By the way, I love that the future is already here. It's just not evenly distributed because that's something that I instinctively feel, but you get a lot of pushback from people about things not being everywhere and perfect, and then them reluctant to give credit to the goal being achieved or at least close to being achieved because it's not perfect. For example, I had a conversation the other day about — somebody was saying that the Tesla, like, self-driving isn't here for us as a society. And I was like, uh, I kinda disagree. I mean, it's not perfect, but I've been in a car where I've pressed a button and a destination, didn't touch anything, and it took me to that destination.

(Joel Beasley at 00:39:30) Destination, didn't touch anything, and it took me to that destination. And so I have done that personally, and I have videos from employees that have done that with Waymo. And I mean, it might not be where you want it to be, but it's happening in a commercially viable aspect today.

(Jim Harris at 00:39:50) This is interesting because it's back to this first reaction of humans as a negative thing to anything new. You know, I love it. People say, "Oh, well, what about the ethical issues?" Like you're driving and there's this group of three-year-old school children crossing the street, there's an old lady, and there's a telephone pole, you know, and your brakes fail. So do you take out the class of three-year-olds? Do you take out the old lady?

(Jim Harris at 00:40:20) Or do you drive yourself into the telephone pole and kill yourself? What are you gonna do? What's the AI gonna do? And this is as though this is some reason not to have AI autonomous driving. But those decisions are being made right now by humans, right?

(Joel Beasley at 00:40:38) Oh, and poorly at that. Like a 16-year-old driver faced with that situation in a millisecond—for all purposes, it's gonna be a random result, right? No one's processing. Like, there is a physiological limit to how quickly we can process thought as humans, and the computers are beyond that limit. So the question isn't whether we let them do it.

(Joel Beasley at 00:41:02) The question is, do we program that in as a specific scenario? Because we can. Like, you can't program that into every 16-year-old in the United States tonight and have them make that decision of taking out the old lady tomorrow and have it happen consistently.

(Jim Harris at 00:41:15) Sure. We know what you're gonna do. Take out the old lady.

(Joel Beasley at 00:41:19) She had a good run.

(Jim Harris at 00:41:21) Honey, it's been great.

(Joel Beasley at 00:41:22) It's been great. I'm not going on the pole. I'm self-preserving in that. Well, in my mind, I have my family in the car, right? Because I'm dad.

(Joel Beasley at 00:41:31) So, yeah, the old lady's getting it. Or the turtle.

(Jim Harris at 00:41:36) Here's the thing that I look at. To separate the waves from the currents, I go down and look at AI and how fast is AI improving. And I also look at what are the statistics. And if you go to—I'm gonna get this acronym wrong—it's National Highway Traffic Safety Association, or it's a government department in the U.S. When you look at Teslas with their autonomous features, they have 85% fewer accidents per million miles traveled. This is not Jim making up the data. This is a U.S. government agency: 85% less. So when you actually look at car accidents in the U.S., 94% of car accidents are due to driver error.

(Jim Harris at 00:42:28) And 45,000 people every year are killed by car accidents in the U.S., and two and a half million people are permanently maimed. Nobody looks at that when you hear this headline about, you know, some autonomous feature had an accident. Like, we're looking at tiny, tiny, tiny data samples compared to this huge data sample of 45,000 deaths a year on U.S. roads and two and a half million people maimed, and 94% of it being due to driver error. So for me, that's a deep current we need to pay attention to, and these media stories are the waves.

(Joel Beasley at 00:43:12) You know what's funny is that that organization you mentioned, so they report those numbers, right? The 85% fewer. They notoriously are more difficult on Musk for a variety of things.

(Jim Harris at 00:43:25) Yes.

(Joel Beasley at 00:43:26) So that lends it even more credibility. They're always difficult with him, and then they come out with this data. Have you seen—so we're gonna take it now from your report to my reality—have you seen the near-miss videos on YouTube with Teslas?

(Jim Harris at 00:43:39) No, I haven't.

(Intro Narrator at 00:43:41) But send—

(Jim Harris at 00:43:41) Send me some links to them.

(Joel Beasley at 00:43:43) There's some compilations where, you know, the Tesla has the recording and, like, for example, it'll just slam the brakes and then a semi truck will go across the front of it. It'll save someone from a semi truck running a red light because it'll do the assist. It does these fantastic saves, and they have it on video. And I'm like, that's what you need. Oh, wait. Josh has got one.

(Jim Harris at 00:44:08) So this guy hits the guardrail, and then you go, okay, I need to slow my car down in case he begins spinning out or something.

(Joel Beasley at 00:44:15) Yeah. I like the one with the, um, there's one with—oh, there it is.

(Jim Harris at 00:44:21) Oh, missing a deer coming across the road. So this is a great example. I love this. So thank you, Josh, for these. I have personal experience with an animal at night coming across my path, and I missed it like this. But what I'm really saying is I don't even want computer vision, which is what Tesla has. I want infrared vision for night where I see the heat signature of a deer coming perpendicular to the road, but in motion right across my future path. Do you know what I'm saying?

(Jim Harris at 00:45:03) I can't see the deer because it's outside of my headlights, but I can see—my vehicle is so intelligent it says if this deer continues running at this speed because it picks up an infrared signature, a heat signature. So if you do focus surveys with customers, they will never say that they want this feature. They will only be able to tell you in the negative a bad situation they had, like the video we just saw. I was driving late at night and a deer—in my case it was a moose—came out on the road.

(Jim Harris at 00:45:45) It weighs 2,000 pounds. If I hit it, I'll not only likely kill the moose, I may kill myself and damage my rental car. So customers can't tell us in the positive the feature or benefit they want in our product, but they can share with us the classic fears and frustrations that they've had as a driver. And it's our job as an innovative company to take those negatives, take AI, take infrared technology, blend it all together in a new feature for our car that increases safety when driving through wilderness, which, you know, the majority—we don't think about it, but the majority of the U.S. and Canada is wilderness.

(Jim Harris at 00:46:33) We only think about big cities like Nashville or New York or Austin, Texas, but the majority of the country is wilderness. So you get these situations.

(Joel Beasley at 00:46:43) Yeah. You've been out to my house.

(Jim Harris at 00:46:47) Yeah. We didn't come across any deer or elk or anything on the way to your house. But we did come across a very good sandwich place.

(Joel Beasley at 00:46:56) Yeah. Those were delicious. I can still remember those sandwiches. So can Josh. Alright. We've got a few minutes left here. I wanna wrap up. Oh, I actually did have a question for you. Did you see that Ballie, B-A-L-L-I-E thing by Samsung, that Samsung robot in your house at CES? Did you get to play with that?

(Jim Harris at 00:47:12) I didn't get to play with it. I think I saw it on social, but it's really hard at CES. There's two and a half million square feet of exhibit space.

(Joel Beasley at 00:47:23) Oh, wow.

(Jim Harris at 00:47:24) So, no, I didn't. But we're gonna see in Japan, for instance, they have a very restrictive immigration process. And they're not having enough children, and who's gonna take care of them from a healthcare perspective? So we're seeing robots that can lift people that are 175 pounds to move them about because healthcare workers in Japan are aging and they can't lift a patient. So we're gonna see AI and robotics assisting people in different ways that we don't think of right now—nontraditional ways.

(Jim Harris at 00:48:06) So AI robots to keep people company. And music is something that's very important to people with dementia and Alzheimer's. So we're seeing robots that are soft and furry. My own mother had dementia, and music and furry animals soothe some people with dementia and Alzheimer's. So we got a little dog for mom that was white and fluffy and really cute.

(Jim Harris at 00:48:42) But we're seeing fluffy, small dogs and seals that are animals with music embedded in them—not encrypted—to help with Alzheimer's. So we're gonna see all sorts of specific use cases around AI and automation to aid in healthcare and creating greater end-of-life quality for people who have dementia or Alzheimer's.

(Joel Beasley at 00:49:20) Oh, man. This is great. Jim, it's always a pleasure. Any other big events this year that you're going to?

(Jim Harris at 00:49:26) Well, I go to a conference called Collision, which is in Canada. It's the sister conference to Web Summit in Portugal. There's a Web Summit in Rio. So I tend to focus on high-tech conferences, but I'm not interested in technology just for the sake of technology. It's all about how is technology gonna change business models? How is it gonna increase revenue, decrease cycle time? And cycle time, we didn't talk about, but that is huge. Imagine that it takes me, on average, three weeks to do some process, like the advertising marketing example, and I can do it now in two hours. And my competitors are all still locked into the three-week process.

(Jim Harris at 00:50:22) Well, we can win more business because we get back to the client faster with options, and they love something and they say, "Yes, we're gonna go with it," because the other people aren't gonna be ready to present for three weeks. So time can be a huge factor in winning, gaining business. But there has not been a lot of discussion that I have seen around the time impact. So real leaders need to get—oh, and I wanna just say for people who are watching this. I've had the honor—I don't know if you can see that, Josh, in the background—but the largest global organization for mentoring and coaching CEOs globally with 45,000 members globally, it's called Vistage.

(Jim Harris at 00:51:15) I know Vistage. Alright. Well, the Canadian arm of Vistage is called TEC, and it just named me the Speaker of the Year because of the AI that we've been talking about just now. It is so explosive. So if your listeners have a conference or a seminar that they wanna bring in a speaker who's gonna be eye-opening and fun and engaging, please reach out to me at JimHarris.com. And the largest CEO mentoring network in the world just said I was the Speaker of the Year.

(Joel Beasley at 00:51:53) Well, yes. And I personally recommend Jim. He's been to my house. We've done episodes together. I followed his content online. It's why we keep having him back on the show. He's always got interesting ideas, beautifully intelligent, and any opportunities to work with Jim, I jump on them. So woo-hoo! Thank you for doing this, man.

(Jim Harris at 00:52:13) This has been so much fun. Yeah. Yeah. An hour just went by like that.

(Joel Beasley at 00:52:18) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email: [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.