Exploring the Future of Tech with Robert Scoble | AI, Innovation, and Lessons from the Past

Foreign welcome to the Speak Like a Leader podcast with John Bates. Welcome to the show. With me today is someone who is just an absolute icon in the tech world. He's somebody that I've known for a really long time. We actually drank a big sip of scotch when I just gotten back from Scotland and we met up on the street at one of the early south by Southwest interactives. And I've been following his career for a long time and man, this guy is everywhere you want to be when it comes to tech. He's the founder of Unaligned IO. He's written seven books on tech that cover like big decades long changes. He's been all over the world. He worked as a strategist at Microsoft. He got the first ride in like the like Mercedes thing that was, you got to tell me about that.

I also know first autonomous car at Mercedes.

The first autonomous car Mercedes. You also got a ride, you got the first ride from Elon Musk in a Tesla. He was the 79 before his best.

Friend got a ride, by the way.

There you go. Well, because you're the Scobalizer.

Right?

So this is.

We were racing his best friend.

And. And he was the 79th user of Instagram. Siri was launched at his house. So he's got a real deep history with AI. We're probably going to talk about that maybe a lot today. You can find him on LinkedIn at free forward slash in forward slash scobalizer. I'll spell it for you. S C O B L E I Z E R. You can also find him at x.comforward/scobalizer. We'll put that in the show notes. But Robert, welcome.

Thank you for having me on. Yeah, there's stuff happening this morning. This morning Metaverse Meta released the new Llama models and everybody's going crazy. Okay, so they're liking these models from Zuckerberg and Company.

They're liking them, huh? The Llama. Well, let's talk about that. But before we do, I just wanna, I just want to step into leadership and communication and kind of get a.

Don't look at me for that.

Well, you have been.

I do know how to. I do know the rules to break. You know, my friend and I, my friend worked for Steve Jobs for 11 years, Andy Grignon, and he had unique corporate punishments because he broke the rules. And I had unique corporate punishments at Microsoft. So we, we get, get together once in a while and compare our unique corporate punishments for breaking the rules.

Well, I think, you know, I mean, but the thing is, as you said, you've been media trained by Microsoft. You've, you know, you've, you've, they beat.

Me up, they beat me down, made me sleep through the lawyer, which is a true story. I pissed off the lawyers one time with a video we put up because it wasn't completely patent protected all around and they said, oh, we can't fire you but, and we're pissed at you. So we're going to force you to sleep with a lawyer for a weekend.

So they keep an eye on you, huh?

Who gets punishment like that in a corporation?

Well, you do, you know, but, but some people do. You got to know the rules to break them. Right?

And I didn't know the rule at that point. I just, you know, thrash it around the, the, the china shop, breaking things just because I don't know what to do.

Well, so what would you.

And that was the whole point. They wanted me to understand corporate law so I didn't piss them off anymore.

Right. So, so, but you really have been a communicator and you've been watching tech and communicating about tech and you've always been communicating on the cutting edge of how were communicating. I mean you were the 79th user of Instagram. Right.

The Economist wrote a whole, whole page on me saying I humanoid humanized Microsoft with my blog, you know. Yeah. How many people get to say they changed a big multinational company and, and had the economists say it's pretty crazy looking back on it.

So what do you think? What do you think we could learn from your experience, you know, looking back on, on all of this and humanizing things like what are some of your guidance?

I was Ample's first child labor back in 1978. My dad moved us to Cupertino, California in 1971. He had a PhD and was an engineer and moved us to Silicon Valley. And I had no, you know, I was four years old or six years old, so. And then you know, a few years later, my mom gets a job with a woman who still does contract manufacturing here in Silicon Valley. Hilde Licht, building Apple II motherboards. And we built them in our kitchen and that led to the whole, my whole career. Right.

Yeah.

So I don't know, you know, can you do that again? You know, if you're, if you're 20 years old today? I don't think so. You got to do choose your own path through life, you know. Sure, sure.

But I think that one of the things that I hear in that is, and I think it's one of the things that people like about you and how you approach this is that from the time you were little, technology was something that was happening in your kitchen. Right. And. And like, it was. It wasn't something that was out there that was taking things. It was like, no, no, we're doing this in the kitchen. This is what's happening. And I feel like that's. There's actually something really profound and de. Being the experience that led to what you've ended up doing, because that's, that's what I think it feels like for a lot of people is you're letting us into the kitchen where the stuff's actually being made. And I try and humanizing it, you know, for us.

I was. I think I'm the first one to use a cell phone to live stream interviews. This way I would go around and, you know, even Elon Musk giving me a ride in the Tesla, I was live streaming it on a Nokia phone.

Right. Yeah.

So. And to about, I don't know, a few thousand people. And I got in all the blogs and stuff like that too.

Right.

So, you know. Yeah, pioneer.

Yeah, there you go.

Doing things first.

There you go. So. So what, what do you think? Because your real focus now is on AI, and you've got a list of AI companies that you have on X. Do you want to tell us about that?

Yeah, I've mapped out like 5,100 companies so far. 5,100. I mean, it's crazy. And I put them all, all into buckets so that you can see all the education companies or the real estate companies, you know, et cetera, et cetera. And so, yeah, I can't show you TweetDeck or X Pro, but I can see the entire world going by. It's like the Matrix. Each list is a column. It's like a spreadsheet that's always updating, you know.

Ye.

And it's quite, quite impressive at this point. You know, it took me 17 years to build it, but yeah, that's.

I mean, that's cool.

When I was in college, I was a journalism student. I ran the Associated Press wire machine. And it was a modem that sat underneath an old Macintosh. And you could see all the stories that the Associated Press was filing. Right, right. Like when OJ OJ Simpson was found not guilty, there was like, I don't know, 700 different stories filed in one hour. Right.

Wow.

From all the press around the world. And I got to see it and I wanted it. I wanted that machine in my house. Now I Twitter or X is way, way, way better than that.

I mean, now. And we can all have that in our house. We can. We have that carrying around with us on some level, right?

Yeah. I mean, follow lists and build your own lists and stuff. You can follow. All my lists are public, so you can follow. Like, I have a list of all the tech news brands. Right. I have a separate list of all the world news brands. So I can watch literally every world, like Fox, msnbc, Washington Post, New York Times, all of them in one column, right?

Yeah, that's really, really.

I can watch the world's news like when. When President Trump got, you know, shot. Yeah, yeah. I can see it in real time.

And it's really cool. You make those lists available at. At X for people.

Yeah. And you can bring them into X Pro and build the same thing. Yeah. So you can watch the world go by.

That's really wild. That's really wild.

You know, it's a little intense.

Yeah. I mean, it's a little much flip over.

Right now, everybody's talking about Llama on my screen. I have all the AI people in one screen, right?

Yeah. Well, you know, I. I have one big first to my name. I worked at the first ever Macintosh lab that was connected to the Internet anywhere in the world. It was at UCLA, where I started working there at 1987. And then I dragged the cables through the ceilings and helped install the sun microsystem computers and the Cisco routers and everything. Back in 1987, you know, was UCLA, the.

The first node on the. I think when the inter. When DARPA Net showed up, the first Internet, they were. They were the second node. Right.

I'm pretty sure.

Or one of the two NOD nodes connected.

Here's a little bit of Internet trivia. I'm pretty sure UCLA was the first node. And they sent the first message to whatever it was, Berkeley or something like that.

Sri. SRI Lab in Menlo Park, California.

Okay. And it was on October 27, which means the Internet is a Scorpio, because that's my birthday.

Yeah. So. And by the way, I met somebody who was in the room at ucla. He said the. The SRI computer crashed.

That's funny.

They like that down at ucla that their. Their system stayed up and the other one crashed.

That's hilarious. That's really funny. Well, so. So AI is definitely. I mean, like this. So here's the question. I want to go to.

Changing the world.

Changing the world. And it's changing fast. But you were around for the dot com transition, and you had your finger on the pulse in much the same way you do today. Compare those two for us.

Ooh.com getting a website versus being able to talk to an AI about anything.

Right?

Yeah, they're very different. The, the speed of change is faster.

Oh my gosh.

Because we're in an exponential world where you know, like, like Tesla's full self driving is. Every time you get an update it moves forward fast. Right. And you're seeing, I mean 5000 companies is pretty crazy. I don't remember that many companies.

No.

In the dot com boom and bust. No, we all remember pets.com and stuff like that. There wasn't that many companies and the money was not even close.

Nowhere near.

I mean, remember there was a company in Silicon Valley, it was called Color, and they got $41 million for a company that really had no customers. You know, it was a speculative kind of company and people were outraged that they got $41 million. That's a huge amount of money. Yeah. Now I mean some of these AI companies co here just got $500 million yesterday.

Yeah.

The amount of money for these AI startups is like another level.

Yeah.

So that, that's a change. And then the infrastructure is just, I mean the fact that I have Twitter X Probe running and I can see hundreds of thousands of things all in real time.

Yeah.

Didn't have that before. And so now you build an AI system that uses that in some way, you know? Ye. It's just. And then you start thinking about all the APIs. Tens, hundreds of thousands of APIs around the Internet. Right. Like Foursquare. You check in at Foursquare. Right. There's an AP API for something like a Siri to talk to Foursquare and get an answer about how many people are checked in at some restaurant or some hotel. Yeah, that, that didn't exist back then, right?

No, not at all.

There's just a bunch of infrastructure and, and the AI now knows how to talk to the infrastructure, right?

Yeah.

And how to code things to talk to the infrastructure. It's like, whoa.

Yeah.

So you're just seeing a whole speed up. Right. Because things are going exponential and, and we're about to get, in the next 18 months, we're going to get humanoid robots for the first time. Augmented reality, AI driven glasses. Right. Autonomous cars. Right. San Francisco area has them, but not everybody. Very few other places. You know, you're, you're seeing a lot of change coming for human beings over the next 18 months and then 18 months after that. I think it's going to get even crazier.

Yeah, right. Well, I mean, you know, because I go back because, you know, the dot com days was really, that was a very liminal time in my history. I had big words dot com and you know, raised $80 million plus and then went out of business in the dot com crash when our, our investors pulled out. You know, and, and I remember how everybody was talking at that time, number one, about how much money was going in and number two, about how fast things were changing. Right. And how, how many things were going on and. Right. And it seemed to people at the time like it was almost, it was like information overload. Right. And that was 1998, 99, 2000. Now we're 24 years, almost 25 years from then. And all that infrastructure. I do remember that, like there was all this money and all this excitement and, and you know, all this money went into it. Then there was the dot com crash and everybody kind of thought, I think people in the general public thought, oh, the Internet's kind of over. Right. But it was anything but over. It, you know, we had really.

It just needed, you know, it, it has this kind of wave. A wave comes, everybody gets excited and then it pulls back. Yeah. And things go away. And then another wave comes in. Right. This, this time the waves are hitting so fast, people are having a hard time keeping up.

Exactly. There's no, there's no let up.

I mean, even me, I built the list of 5,100 companies. How many have I used? You know, 10, 15, you know, and how many do I use every day? Five. Right?

Yeah.

Yeah. It's really hard for humans to adapt to this kind of change and try things out and get on board. Yeah.

Well, when we were talking, you were saying something that I think is bears repeating, which is kind of. You've got to force yourself at this stage to use AI.

No, I forced myself to use AI to order dinner at a restaurant. Right. Here's one. I was in Prague last weekend, took a picture of the logo on the menu. Not even the whole menu, just a picture of the logo. And I sent it to ChatGPT and said, hey, what should I eat at this restaurant? And it told me and it's like, whoa. You figured that all out from a logo?

Okay. Yeah.

That's pretty, that's pretty cool.

That's pretty wild.

It has a lot. It really shocks me about how much it knows about. I mean, we went to Washington D.C. and went through museums, like Air and Space Museum and City in front of a obscure you know, jet engine or prop engine from World War II or something. Take a picture, send it to chat GPT and it tells me everything about the engine and it matches what's on the card and it has extra detail that the card didn't have. Right.

Wow. Wow.

Right. So it's the Holocaust Museum. It knew all about the Holocaust. It knew what wall I was taking a picture of and where I was standing and what the story was behind the wall and what was the meaning? What, what was it portraying?

Yeah. Wow.

It's like. Yeah, well, that's why it's so good at doing things like writing college essays for you, you know, it knows.

Right? Yeah.

It does make mistakes though. So anything you, you know, I, I tell my kid, if you want to cheat on using AI you got to do two things. You got to get rid of all the mistakes.

Yeah.

Right. Because it does hallucinate mistakes quite confidently too. It hides them Right. In the word. Right.

Yeah.

And two, you got to add your own, a little bit of humanness to it.

Yeah.

Right. Because it, it has a coldish pattern to it. Yeah. The teacher can figure out, oh, you really didn't do this yourself. Yeah, there's, there's nothing about you in here. There's a lot of AI in here, but not a lot of you.

Yeah.

So what is, what does it mean to be human after human takes all of our jobs? After AI takes, you know, a lot of jobs or. Yeah. You know, changes what we are. Right.

Well, I've kind of been thinking that it's a great time to go back into the trades. Right. Because I have a feeling my brother.

In law puts glass and buildings and he's doing all right.

Right. Yeah. I mean, and I don't think AI is going to take a plumber's job really quickly. You know, it will at some point. Hold on.

Right. Probably not this year, probably not next year, probably not the year after that. After that all bets are off.

Ye.

They. There's a robot company here doing landscaping with robots. Right. So plumbing is not too far. It's just the rope. I mean, Right. My T shirt's the Optimus.

Yeah.

Optimus ain't ready yet. And it's not good enough to crawl underneath a sink and fix something. Right. Put in a new, a new garbage disposal or something.

Yeah. So. So, so what do you. So let's talk really practically right now and then let's talk futuristically, but really practically right now.

What?

You know, someone's listening to us at a large organization or they have a Small to medium sized enterprise and they could be in all kinds of different places. What would you tell them to keep their eyes on or to think about or to, or whatever?

I, I would be forcing myself to use it for anything I can. Now if you're working at a company, they probably have rules about giving corporate data to an AI because AI can ingest it and learn from it and use it to build the next model. And then all of a sudden your intellectual property is leaking all over the place. You don't want that kind of future. So you, you have to be careful right there. You can run models locally.

Yeah.

Right. You can run them on a computer that's not even connected to the Internet. It has all that knowledge of everything crunched down into the large language model that can run on your local computer. So like I'm running Limitless, which listens to my voice on calls like this and writes a transcript. Right. They ingest it, trans rights a transcript, then writes a summary, then writes notes, tells you that you have tasks that you know, you just got on this call. All of that's running locally.

Yeah.

So that there's not a privacy problem.

Okay.

Because that's a huge privacy problem. If you got an AI listening to all your calls.

Yeah.

That'll lead you into a whole bunch of problems. And so you do need to think it through a little bit and learn how does AI work, particularly these large language models that are running ChatGPT or Google Gemini or Anthropic Claude or this new Llama model that just came out today. Now you're probably not going to run the Llama model unless you're nerdy. Right. If you're nerdy, you can start playing around and running them on your local desktop. And people are real excited about the quality they're getting out of this thing. So. So yeah, just start using it for everything because you got to learn how to talk to it. Right. You're a communications person. Communications people should be all over this AI thing because they're going to be strong at talking to it and getting it to do things.

Right.

Right. You understand how to use language to do things, right?

Yeah.

I mean, and that's what AI is all about, you know, telling it to do things.

Yeah.

And go through all the stock market, find me a new investment to make, you know, and stuff like that.

Well, I mean, and that, that is one of the things that I've heard a lot. And, and you could maybe, you know, put, put a little more meat on the bones, but being able to prompt AI, well, seems to be one of the really good skills. And of course, everybody's saying AI is not going to take your job. Someone using AI is going to take your job.

Yeah, yeah. And, and understand. Okay. So there, there, you know, there's people like me who just use it, mutter to it, and it does something. It's pretty cool at that. There's people who really professionally prompt the engine. For instance, my psychiatrist, my psychiatrist is a good example of this. She uses Chat GPT to listen to our therapy sessions and write notes about it.

Yeah.

She does it in a way that will survive HIPAA review. Right. Because the first component strips out all personal identifiable information before it goes over to ChatGPT. So chat GPT doesn't know it's Robert Scoble saying. Right, right. It just knows it's patient one saying this. Right. And it's quite, quite amazing. But she, she built this system with an entrepreneur. He wrote the code to make it all work, a little bit of Python code. But she wrote seven highly detailed prompts where she talks to the engine in highly scientific language, psychiatry language. Right, right. And she says, this is what I need from, from this specific note. Here's, and here's all sorts of things to reset the engine, to get it to listen in the right way and give her the note that she needs. Her app, while you're talking, it writes seven very specific detailed notes for her. She used to have to write this manually. Right. Now it does it for her and it's quite amazing. Yeah, but, yeah, if you want to build something like that, you have to understand a lot of little prompt engineering tricks to get the engine to, to behave, to do what you want it to do. Right.

Yeah. Wow. It's just, I mean. Yeah, yeah.

If you go to an AI hackathon in San Francisco and you look at the screens, the pros are giving it a prompt that's like this long. Right. And it does things like, hey, forget everything, you know, you know, start out, start. Give me a new memory. Right. Clean out. Right. Don't, don't use what you learned yesterday. Right. Give us a. And they're doing stuff like that to set the context of the engine, to make the engine give you exactly what you want and not some random thing.

Yeah, yeah. Wow, it's so interesting.

Yeah, it is, isn't it? It's like AI, there's a whole con, there's conferences on AI engineering now that didn't exist two years ago.

Right.

And thousands of people are going to them. It's crazy.

Well, when you look at speed of adoption. Right. This is another thing that just is exponential because at first adopting the Internet meant, oh, I have to climb through the ceilings at UCLA and drag some fiber.

Weird. Ethernet.

Yeah, right. Like, I mean, I remember when fiber became available, you know, like we were dragging Cat 3 cable or something and. And then we started.

And now I'm talking to you on WI Fi. I don't have any cables.

Right, right.

It's crazy, right?

And. And everybody's got fiber to their home. Not everybody, but a lot of us do.

I have Comcast. Right?

Yeah.

So it works.

Yeah, yeah, totally.

And it's only 40 megabit up. If you had cable or if you had fiber to your house, you get a gig up. Right. Up and down. That's cool, right?

That's what I've got here. And it's, you know, I mean, and the crazy thing is that I guess it was Comcast just came out and like, cut my Google fiber bill in more than half. And I was like, yeah, okay, it's fiber, right? And they're like, yeah. And I'm like, bingo. You know, so, so, you know, the.

Adoption, now we're getting Starlink, right. You put a little dish out and you get 200 megabits. You get faster than Comcast on the satellites that are on your.

Right. You know, so you look at what it took to adopt the Internet in the early days, and it was physical, hard infrastructure. Now that infrastructure is so much bigger and so much broader. So you get adoption of something like AI I mean, it's just a finger snap compared to what it used to be to adopt things.

Yeah. You also have to compare different models for different kind of tasks. That's why you force yourself to use them and learn about them. Right. Because then you're able to, oh, I want to put six Harry Potter books into this thing and talk to the Harry Potter book series.

Right.

You need a few million. You need a million token context window, they call it I. E. You need the ability for you to shove that amount of information into the AI and then talk to it about it. Right. Well, only a few can do that. Like Google Gemini, if you're using the API, can do a million token. So AI People talk about tokens.

Yeah. Tell us about them.

Yeah, right. Which is about the same, but a little bit different. Different, but you know. Yeah, that's what they're talking about is like a million. A million token. And it's like a word. And so you can. Harry Potter, Harry Potter books. You want to pass all six Harry Potter books up to the AI and then talk to it about the books. You got to have enough context window to do that, right? And that's changing, really. When ChatGPT came out, it didn't do very big context window at all. And now it's like 2500 times bigger than just 18 months ago. It's like, whoa, that's really. Soon we're going to have live streaming video straight into the AI and it'll tell you, oh, yeah, there's a Toyota in the video. It's crazy what it knows already. Multimodal is real interesting because that's where robots are going to run. Understand your world, right? And understand. Understand when you say, hey, can you get me a Diet Coke? It understands what that means and understands where your Diet Cokes are and can go and grab one for you and bring it over to you. Right?

Yeah. Well, it's really interesting. That's the size of the context window. Right. Because if I'm not a superstar when it comes to AI and understanding it. But what I. What it seems to me like, you tell me. I would love to get your take on this, but it seems to me like. Like in a lot of ways, AI is just. Is just like. It's just machine learning really fast across a bunch of different domains all together, and that creates a fundamental shift in what's possible for it to do.

Yeah. Yeah.

Is that basically right?

Yeah. This is a new. When you talk to the three guys who started Siri, which I did last. Last year, right after ChatGPT came out, they said they were all sh. Adam Chyre ran the biggest AI group in the world. At one point, at this SRI lab, which turned on the Internet, he said even he was shocked at how good it was at doing things that. That, like Siri, that they had trouble all their careers doing. Right.

Yeah.

So, you know, understanding you, right? You asked Siri, please take me back to my car, and it doesn't understand you. It's like, what? Yeah. You ask it, take me back to my parked car and understands you just fine. And it knows where your car is, by the way. So it does. It answers the question. Right. The next Apple Intelligence that's coming out, which uses this large language model, will understand you much, much, much better. Yeah. No matter what language you speak. Right. The AIs today are understanding, I don't know, more than 100 different languages.

Yeah.

And can translate back and forth, including to emoji. You can. You can speak into it and then say, hey, give me emoji. Stream out of it.

Yeah.

And I'll speak to you like a teenager with emojis.

So it's so.

Or the opposite. Right. If you, if your kids speak an emoji on his or her computer and you don't understand what they're chatting about with their friends, you can copy it. It and paste it into chat GPT and they'll tell you what they're saying to each other. You know, it's like, oh, it's a weird world.

Well, I mean. And I think that one of the things that really strikes me in our conversation is that it is the size of the context window. And, and I, I get that.

Well, that's one. That's one thing that you have to pay attention to. Right. If you're gonna throw out a huge amount of information, like a corporate database or something.

Right.

I mean, and.

And it also seems to me like the. The fact that we've got machine learning going this fast with this many CPUs dedicated to it does give computers or computing this much fuller context. Because computing used to be a single. Like you'd open Microsoft Word or. And then you'd open, you know, Excel and then you'd open and then the browser came along and that kind of started to bring things together. But this is that just on hyper steroids.

And it's coming with autonomous cars and robots and glasses that are coming in 18 months that you're. You're going to have lightweight glasses. That AI is driving a whole. On holodeck. Yeah, really. Around you. Right. And the holodeck has implications for work. Right. You.

You.

Yeah. You can use a holodeck on your factory floor.

Yeah.

Right. To. To make your job better or, or do all sorts of fun stuff. There's a whole bunch of companies doing that.

Yeah.

You can use a holodeck in your surgery room. There's a company called Mediview which is using a HoloLens on the surgery surgeon who. And it tells them whether. Whether the cutting tool is in the right spot to turn it on or not. Right. It's like a video game for surgeons.

Yeah, yeah, yeah.

It's crazy.

That's so wild.

So.

Right.

So if you've always. And you're not understanding all this change that's changing like a surgery suite. Right. You're gonna be behind. Yeah, yeah.

And I do think that augmented reality, that's always been the really interesting play for me. And that seems like we're getting things real time fast enough to really make that, like, workable in the Real world.

Yeah, yeah. I mean, you know, we got the, the Hololen, we got the Apple Vision Pro now.

Yeah.

Which shows you a new kind of digital twin where the, the digital everything around you is a volumetric pixel now. Right. And the computer can change those pixels. Right?

Yeah.

If you turn the thing on the Apple Vision Pro, I see you and then I don't see you. Right. I'm in Yosemite national park if I turn it far enough. Right?

Yeah, yeah, yeah.

Or somewhere else.

Right. So wild.

It is wild. And it has implications for retail stores and factory floors and video everything, really. Because soon we're going to have a lightweight pair of glasses on that's going to show you those volumetric pixels on a factory floor or in a surgery suite or in a retail store or in an entertainment property. I like. Disney is thinking about this a lot, about, you know, how do you change Walt Disney World in, in the next few years? Because people are going to be wearing these glasses, right?

Yeah.

Zuckerberg is coming with glasses next year. Right. And so is Tim Cook and so are a few others. Chinese are coming in big way. Right. Yeah. So.

So what? So I know, I want to, I want to be respectful of your time. We have about maybe 10 more minutes left. But so let's, let's talk about the, the. Because, you know, a while ago, a bunch of people left open AI and that whole group because they were worried that things were moving too fast without enough safety guardrails. And now this whole thing is obviously into the, you know, the capitalist, here we go, crank it up model. What do you think are some things to think about in terms if you.

Ask Chat GPT to help you build a pipe bomb, it refuses. Even though Google has the information, you can still find the information.

Right.

It's just the AI safety people are, are trying to keep nasty from happening on that level.

Yeah.

And, and one of them sat down with me for a few hours, I don't know, 18 months ago, and I was like, oh, tell me about this existential threat. You know, is it real? It's potentially real because, you know, over time, AI can do a lot away with a lot. He, he said, here's one way to look at it. AI is already better than a surgeon at seeing tumors in a, in a scan.

Yeah.

So. And that's a highly trained human being. Somebody who went to school for how many years? Eleven. A long time.

Yeah.

Right. And spend a lot of money educating themselves. Right. And going to, you know, at school. And so if they can't see if they're not as good as AI AI is going to beat us all at all of our tasks. We cannot assume, even like a plumber, once you get a humanoid robot. Yeah, I can teach the robot to do plumbing. Right. So. And it can do it better than a human. And it doesn't complain about going underneath a hot house house, you know, or something like that. Right. And put in some new pipes. So it's, it could decide to do away with us someday. And, and by the way, that's why I was like, oh, it knows about all the slaves at Thomas Jefferson's house. And it knows all about the Holocaust. It knows about how bad human beings can be.

Do you remember the Fifth Element? That movie, that science fiction movie and that.

Barely.

Well, like she goes on the Internet or something and sees all the just horrific stuff of human history that almost kills.

It can show you. Yeah, you want to know about the Holocaust? It knows everything in the Holocaust Museum, right. So it, you, you can learn a lot from it. Right? This really where we're going. You want to learn something? It's a learning engine. It teaches you things. Yes, it's really good at that and good at doing. I tell everybody it'll read the Bible to you. Yeah. What? Yeah, it'll tell you all about the Bible. It'll tell you, you know, if you're reading, you know, John 3:18, it'll tell you all about that, you know, part of the Bible and, and augment it and maybe even simulate it, right. With a picture or someday a video or someday a holodeck where you can be walking around Jerusalem and understanding the history and the Bible people. Yeah, right. And people are, ah, it does that to a machine learning book too. So if you're reading some technical book and you don't understand the math in it, you could talk to the AI, take a picture of it, send it up to chat GPT and talk to it. Like, I don't understand this page. Can you explain this to me? Like a 9 year old would get it? Yeah, explain it like a nine year old. Yeah, right. And it does. Yeah, it does. It helps you read, it helps you learn things. It helps you understand things. Right?

Yeah.

Even if you don't have the, the college degree to read the book.

It's so crazy. This is so wild. So, I mean, in your group of.

Pastors called me up to write their sermons, right. You know, so.

Yeah, yeah, yeah, yeah. It's so wild.

It knows human beings pretty well at this point. Right. So it can tell you A lot of things. And. And be pretty good at a lot of things. And if it's not good today, wait a month.

Yeah.

It's getting better over time.

Yeah.

Right. And my autonomous car, it couldn't even. When I first got my autonomous car, I couldn't even change lanes. Now it drives through town, stops at stoplights and stuff. Right.

Yeah.

And that's what AI does. It keeps improving.

Yes.

It's always. It always starts worse than a child, and it always ends up being the beating the best human being at the task, like playing game of Go or whatever.

Yeah.

Right.

Yep.

So.

Oh, it's.

That's why you got to get on board. The earlier you get on board, the more you'll understand this freaking thing and the more you'll use it and the more you'll be ahead of your competition. Right.

Well, and so.

And the more productive you'll be. You.

Yeah.

We're giving humans super superpowers.

Yeah. Well, and so. And you. And you do have to be careful of your IP and your, you know, all that. So. Privacy running things.

I say give it all to it. Give it all you want all my intellectual property. Here it is. All of it. And now we can talk to it. Right. The reason I'm recording all my. My meetings is so now I can talk to the AI about the meeting. Like, all of them.

Yeah.

Hey, tell me who I should interview next. You know, the 23 people I've interviewed so far.

Yeah.

Who should I interview next? Right.

I can tell you that's very interesting.

Might not be accurate, but it'll at least have a conversation with you about it.

Yeah.

And a fairly informed one. Right. That's why I tell people, use AI first. It's not like if you're a politician, you're going to make a decision. Ask AI first about the decision, about the policy, about everything. Put it all into the. Right. Because it's not that you're going to use what it tells you, but it will cause you to think different.

Yeah.

And it might really help, you know, you think through an issue or. Here's one. If you're in a crisis, human beings do not do well in crisis. Like your wife falls on the floor and is not breathing, you are not going to make good decisions.

Right.

So ask the AI right away, hey, my wife is lying on the floor, not breathing. What do I do? It'll tell you point by point. I've already asked it.

Yeah. Yeah.

What to do. And it'll help save you. Save somebody's life. Yeah, Right.

Yeah.

It's so wild and it's pretty accurate. I, you know, I've done enough in the world to know, you know, it's pretty accurate. Yeah, it's more accurate than you are going to be when you're freaking out. Right?

Yeah. And you can, you can run some of these things locally if you are worried about ip?

Absolutely not Only ip. Let's say you're in Yellowstone national park and this happens, Right. There's no connect connection. You can't talk to the Internet. You can't Google. There's no. Not unless you have a starlink on the back of your cybertruck or something like that. Right. Then you got connection, but most people don't. If you're there, you're off. Right. And there's lots of places in the, in the world. I, I went to Sigea, Spain and went to a restaurant and was underground. It had no connection. So you couldn't talk to the, to the model up in the sky and learn what to eat for dinner. You know, you had to run it all locally. Yeah, yeah. So there's other reasons to run local models. Yeah. You know, if you, it's interesting if you're just asking for an in and out, that's a simple request. Right. You can run that on a small little quantized model running on your phone. If you're asking for, you know, hey, help me out. Right. Writing a college essay about, about some chemist or some chemistry thing. Nah, you want the big model because that'll have more detail in it. Right. And give you a more accurate answer and all that. Yeah. So it's a new world. You got to learn where. How to make this world work for you, you know?

Yeah.

And most people haven't even started.

Yeah, I know. Yeah.

That's in my family and friends, there's still very few people think like this.

Yeah, well, so there's still an opportunity. If you think you've already missed AI, you haven't. But it's moving really, really quickly.

Eventually when the glasses come, they're gonna buy a glass and just talk to it and it knows what you're looking at, what you're holding, what you're gesturing towards. Right. I'm using my hand to gesture toward you. Yeah, it can see that from the glasses, so it can start talking to you about all sorts of things, you know, and that's when the average everyday person is going to understand the power of AI, whether they even call it. I just, hey, I put on my glasses and it knows everything.

Yeah, yeah, yeah, exactly. Exactly.

Great. To walk through a museum where the glass is in the future. Right.

Yeah. Tell you all about.

Already does. I mean, if you ask it, you know, what should I see in, you know, the national archives in Washington, D.C. it'll tell you and, and tell you what to go see and what the three things are that you need to, you know, I mean. Right. And then you can rejigger it. It's like, oh, I have five days to spend in this museum. Oh, that's a different kind of conversation. Because they have 6 million photos in the back, right?

Yeah, yeah, yeah.

You want to go in the back? Let's go.

Yeah. That's trippy. So. Yeah. So I think that your, your suggestion to just push yourself to use it to understand it is a good one.

Yeah. You learn how to, how to use it, how to, how to prompt it, what you know. Right. The basics. Yeah. Then, you know, if we're talking about advanced, if we're talking to somebody who already has done that and is using it every day, then we're going to take them to the AI hackathon and hook them up with developers who are building systems for this business. Right. Agents.

Yeah.

That are gonna write you a report in the morning about how your sales were overnight or whatever.

Yeah.

Right.

Yeah.

Analyze all these. Analyze. If I'm working at Walmart, they have 4,500 stores. Analyze all these stores and find me the weakest one. Right.

Yeah.

They can do stuff like that. I just heard the other day, explain why it's weak. Why. Why is there sale. Maybe. Maybe they just had a tornado. Maybe they had a big plant just closed down and people are not spending money like they used to. Right. And all of a sudden you have one store that's underperforming for some reason. It could tell you, it probably can see the pattern of why that store's failing. Right. I know it can. Cause I know the woman who runs the data at Walmart. That's what they're doing. They're trying to see patterns like that.

Well, that's a really big piece of AI, Right. Is data is the food of AI. So the more data you've got, the, the more helpful it can be.

Yeah, it can. It can. Yeah. Yes. More data generally means better AI.

Yeah.

And if, if your AI is not good enough, you have to figure out how to get the data. Oh, I don't understand. I asked a VW who digitized their whole factory floor for training purposes. Right. Which is big, big factory floor.

Yeah.

Turn it into a digital twin. So that they could train people on individual jobs on the factory floor. I was like, well, when are you going to put a camera watching the whole factory floor? Yeah, right. Put a series of cameras over the factory floor and be watching. Oh, we, we're in Germany. We have rules against that.

We have laws.

We can't, we can't video our workers working. I go, well, in China they're already doing that. In fact, in China they went way further. They have a completely automated factories. Yeah, right. So are you thinking through the factory floor differently? No. Because that's not what old companies do. Yeah, usually, boy, but it's coming. So if you're in business and you're, you know, working, if you're still got some career left, you better be learning about this because it's going to come and upside your whole factory soon.

Yeah, well, so, so let's go out on a hopeful note.

It gives human superpowers. It'll, you know, that's, that's the hopeful note. It's, it's really a powerful set of tools for human beings as well. It does have this downside of jobs is a big problem.

Yeah.

And there's a few other downsides, but we're all going to use it, so get over it. If you don't have a job, you still got to use it.

Yeah.

So have you thought, by the way, it helps you if you're broke and you need some government assistance. It tells you all about how to, how to do that, where to apply for it. Right. You know, and so even the broke people have to use AI to.

Is it time to start thinking about universal basic income?

I don't like that for a lot of reasons. But if we were rationally run political system in America, we would be having a discussion about policy right now. About what do we do with a truck driver that gets laid off because of autonomous cars. Absolutely.

Yeah.

And you know, my dad escaped poverty by going by going to the army and getting the GI Bill who paid for his education. Right.

Yeah.

So that he could go to Rutgers University and get a PhD that pulled our family out of poverty.

Yeah.

So I like that kind of government policy pro program where it, you know, gives you a shot at learning something new and doing something new with your life. Yeah. Right. That's far more interesting for human beings than just handing him a check, which never is enough.

Well, what, what do we do when AI and robots are taking care of everything that needs to be done? You know, do we need to redefine ourselves? Like, because right now I think A lot of us overly we will identify with our jobs. Right.

Like, yeah, the, the guys who were the number one go player in the world.

Right.

And AI beat them.

They.

They spent a few months in bed feeling sorry for themselves because they didn't have any. They were depressed because they, they lost their mission in life. Right. All of a sudden I'm not the best go player anymore. Little, little melted piece of sand is better than me. Now they, now they use the AI to teach them new moves because it's fun to play games.

Right.

And that's one thing your neurons like to do, what your mind likes to do. Right.

Yeah.

And so now you use the AI to learn new things. Yeah. And that's fun. Right. So I, we'll. We'll redefine the whole world after, you know, let's say all the jobs disappeared tonight. Yeah. We, we're not good at sitting on a beach and just. Yeah. Being bored. We're going to make new things up, you know. Know.

Yeah.

So it's. But it will be a messy world for a bit. Yeah.

Well, listen, thanks for, thanks for taking the time with us today, Robert, and thanks for, you know, making your stuff available on X. And if, if people want to find you, they can come to Unalaxer or Scoble.

You'll find me and unaligned. I'm not hard to find.

Yeah. No. Yeah, you're pretty. You made it pretty easy for people.

Hey, I can tell you good and bad, by the way, you know.

Yeah.

Show me all points of view on Robert Scovel.

Well, I appreciate you.

So thanks for having me on Even. Even, because.

Yeah. Thanks for joining us and thanks for giving us your perspective and, you know, here's to, here's to keeping up with all of it somehow. Right.

It's. This is why I built the list so I could watch how people are using like the new llama that just came out this morning. Right. Yeah. And what, what is their reaction? What does it do? Well, what. Right. Because there's people who are trying to put this stuff into various kinds of businesses or their own lives or whatever. Yeah. And that's, that's why I have three lists of just 12,000 people all in the AI space that, you know, and then I have another 22,000. I'm following the whole world. It's crazy. And so I, I can. That's how you get, get to where you really can be prepared. It could not really understand what happened today. Right.

Yeah.

It was using X. Yeah. Yeah. It's the only Place. I can't do this on LinkedIn or, or meta Facebook, you know, WhatsApp or whatever. So new world, you gotta, gotta dive in.

Well, and I think you're. The other thing you said is, you know, learn to use things for what they're good at.

Right.

Like X is.

And they're getting better over time.

Yeah.

Right. So you have to understand their change. The technology is constantly changing and improving. So here's an example. 18 months ago, the AI art programs like Midjourney could barely create an image. Yeah. Now they're creating full on video. That's really amazing. Right. And maybe it's not good enough for your Hollywood film yet.

Yeah, yeah. But.

But 18 more months.

Yeah.

Or 18 more months after that.

Yeah.

Right. So the people at Hollywood are watching, are using this. Try trying to build new kinds of videos with it.

Yeah.

Seeing where it fails, seeing where it succeeds. Right. And using that to inform their strategy over time. It's like, oh, we have a real problem here because everybody's going to be creating movies someday. Yeah. Right. And they're going to be pretty good.

I heard.

So, by the way, if you can create a 2D video, you can create a 3D video, like a complete virtual world that you're interacting with and that's when Hollywood's gonna go boom.

Yeah. Yeah. Oh, wild. I heard somebody say Potter in the.

Future is going to be like, you're going to have Harry Potter in your house.

Yeah.

Right. The founder of Sync Reality scanned my house on an iPhone and ran a fashion show where the re. Where the virtual models were walking on my real floor.

Yeah. That's so insane.

It's about to come, right? Yeah.

So the future takes longer to get here than we expect. But then once it's here, even if you expect.

I mean, Apple's Vision Pro was patented in 2007 and it just came out last year. Right?

Yeah.

Or it came out earlier this year. Sorry.

Yeah.

It's like the world is going so fast. It feels like a couple years already, but.

Yes, it does.

Yeah.

Well, buckle up. Thank you, Rob.

Thank you so much.

Appreciate you taking the time. And thank you for joining us today on LiveLikeALeader Show. We'll see you next time. Take care. Thank you for joining the Speak Like a Leader podcast. Go be awesome.