Remarkable People Behind the Podcast: How Your Brain Works

Foreign. Hi, it's Guy Kawasaki. Good morning. It is Wednesday and I am in Santa Cruz, California. Of course, I tell you that every week because I'm always in Santa Cruz, California because I have not left Santa Cruz, California for about 14 months now. So. Well, it's been a good week so far and I hope it's been a good week for you. You can join my group text at 831-609-0628. I'm sorry, but it only works for Canada and the U.S. that's just the way the company set up. But this is a group text and the way to really works is if I send, once you sign up, I send a text to you, that's one to many. You respond to me, that's one to one. I respond to you, that's one to one. So the only one where you have to worry about it being, well, you never have to worry about sending to a group. The, the, the members of this group always are one to one. I can send just to you or to the whole group. So that's how this works. And I want to mention right up front that this live session, which on. Oh my God, it's on Instagram. No, it's face. I'm spit on. Long day. It's been. It's on Facebook, Twitter, LinkedIn and YouTube is brought to you by Restream. And so we are using the Restream platform in order to send out this live video to four different channels. Restream actually handles 30, but we're only using four. And this means that I'm doing it one time and it's going live to four different places. Also very advantageous. It means that when you comment, and I would love you to say hello and comment just to prove this. When you comment, no matter what platform you comment on, I see them all in one place. And in the old days, if you wanted to do something close to that, you had to, oh my God, you had to like have multiple windows open and try to keep track. And oh, there's Vic from Manila, Mabuhay. And so yeah, this is great. And so it's high definition, great audio, great video. Multiple platforms up to 30. Life is good. I don't know how I would do multiple platforms. I'm not going to do four live shows one at a time. So yeah, powered by Restream. If you're a podcaster or an influencer or trying to make a bigger splash, definitely look at Restream to get your live video out multiple places at once. So this week's guest on the Remarkable People podcast was Jeff Hawkins. Jeff and I go back a long way. He. He was the CEO of Palm, so they made the Palm Pilot, and also Trio. So he's been in, shall I say, handheld smart devices for a long time, probably longer than just about anyone. So between Palm and Trio, that's a long, long time. I guess someone could have been in the Newton group, which would have been longer, but not by much. So he's no longer in handheld smart devices, though now he's a brain researcher. And specifically he has a company called Numenta, and he's written a book called the Thousand Brains. And he is doing a lot of research about the brain, specifically the neocortex, the function of the neocortex, which he describes as the wrinkly stuff on the outside of the brain. It's about the size of a sheet of paper and the thickness of. I forget what analogy he uses. Several millimeters thick. And so that's Neocortex is the new brain as opposed to the old brain. And one of the, I think the most interesting things and a question that I want to pose to you to break the ice here is do you think of your brain as a supercomputer or a maybe not a super space computer, or do you think of it as a modeling system? And I'll tell you right now, what he says is that the brain is not a super space computer. It is a modeling system. And the way he got the title of the book is that he believes that the brain has about 150,000 models. And that is very different than doing a lot of computation fast like a computer. So there is a. And we go into this in great depth. So the brain has multiple models for what is called a coffee cup. And so this coffee cup, it has weight, it has volume, it has a smell, especially because the coffee has been in here for more than a day. It has color. And all these models make up what we think of as the cup. And so that's kind of a different approach. So, yeah, it's be interesting how many of you think of your computer, your brain as a computer versus a collection of models. And one of the consequences of this is how do people get really good at something? So second question for you to consider is if you got really good at something, like really good at something, how did you get good? It's probably not the supercomputer model. It's probably the modeling system that your model, you've been working on, your model for whatever, whatever skill you have for a longer time. And that's a very interesting thing. So now, the third ramification of this is that, you know, people see these great news announcements that Big Blue has beaten the chess master and Big Blue has beaten the go master. And a lot of people look at that, say, oh, my God, you know, pretty soon AI is going to be taking over the human race. And I have to say that what Jeff says is, yeah, you know, it's if that's your project, if your project is to create a computer that can beat a chess master or go master and you beat them, yeah, it's great news. But that doesn't mean that's quote unquote, intelligence. I never thought of it that way. So, I mean, he, he is not nearly as concerned that machines are going to take over the world, at least in the next, I don't know, 200 years or so, maybe, ultimately. But right now he's, he, he doesn't think that, you know, computers are going to decide one day they're going to be pissed off at the human race and takeover. That's not on top of his list of concerns, I would say. Climate change is much higher. So, Faisal, according to Jeff, There are about 150,000 models in the brain. So, yeah. So what else? What other questions do you have? Let's see, I don't know. First of all, do you think that Big Blue beating a chess master or go master, does that scare you? It doesn't scare me. Go ahead, ask questions. Let's see here. Looking Waleed, looking forward to today's stream. Heard the podcast and was an early user the Palm and Trio products. So was I, Waleed. Those were the days. I mean, remember when we were just so thrilled? Well, first of all, if you go far enough back, God, Remember the days when a cell phone was basically, it was, I don't even have a box that big, you know, self. Here, here's my rubbish can. So a cell phone used to be about this big and it would be in the trunk of your car, believe it or not. And so it would be in the trunk of your car and there would be this big cable and, you know, one went to the phone in the driver's area and the other went to the antenna and it was about, I swear, seven or eight pounds. And then Motorola came up with a portable one that looked like a brick. And. And now this is my 12. And now we're at this and wow, my, how times have changed. Who would have predicted that? So we got. Welcome all you trio and Palm Pilot users, in other words, all the old people. Nina, Nina from Sao Paulo. The brain is only the hardware and interesting. And as it might sound, I think you get good work by working on the software your mind but more than that, taking care of your in conscient our programming language, that model works for me gets hairy in there, so to speak. So here's a good question. If you could program to be something automatically, if you could program your brain to do something automatically, what would it be? Ah, ridiculous as this may sound, I wish I could program my brain and then my brain could program my body to easily and effortlessly walk to the nose of a surfboard. I have been surfing for about seven years now. For seven years I have been trying to walk to the nose of my surfboard. And I must be reaching the 10,000 hour point, the magic 10,000 hour point of Malcolm Gladwell. So it better happen soon. Now, a few weeks ago, Keith McNamara don't fear big Blue. Big Brother is another story. Well, some people might say that Big Blue and Big Brother are related. Yeah. So Katie Milkman, the God, what was she? Professor of Operations Research from Wharton. She had a very interesting observation and method for motivating yourself to do things. So there, there are these websites where you can say, well, let's take my case. So if I cannot walk to the nose of my surfboard by, let's say, the end of the summer, I have to pay a friend $1,000. There's also the negative way of implementing that, which is in my case, if I cannot walk to the nose of the surfboard by the end of summer, I have to donate $1,000 to the NRA or the RNC, which would kill me. So I guarantee you if I made that wager, I would walk to the nose by the end of summer. I don't know how we got to surfing from Jeff Hawkins and Brains, but Barry Singh. Hi. From India. What part of our lives will not be taken over by AI? Barry, I would make the case that there's no part of our lives that's going to be taken over. Depending on how you define taken over by AI, I think it'll supplement. I think it'll be helpful. I think there's a lot of things that AI could help, you know, that I don't want to do. So. You know, I spoke to the Allstate Claims adjustment department once and the way that automobile insurance used to work is you get in an accident, you file a claim, you have to make an appointment with the claims adjuster. So you drive to that office. You wait in line. Finally, somebody comes out with their clipboard and their Polaroid camera, and they go around your car and they take a picture and they take notes, and then you sign their adjustment thing, and you wait a few days, if not a week, and maybe you get a check. So the next version of this was that insurance companies have an app. The app, you take a picture of your vin, take a picture of the damage, take a picture of your car, you know, etc. Etc. And then they. Somebody looks at that picture and says, okay, you know, 1991 Toyota Sienna, front fender. And I guess they look it up in some database, and it says, all right, that's $2,000. So that's better already because you don't have to go to the claims adjustment. Now, I asked Allstate, well, how do you know that person isn't in a wrecking yard, you know, just taking a picture of a crushed Toyota Sierra? And they say, well, from the geolocation and all that, we know where that person is, et cetera, et cetera. So there's ways of protecting that. And you got to take a picture of the vin, and the VIN has to match your insurance policy. But the next stage of this will be AI where instead of somebody looking at the picture and deciding, that picture will just go to a computer, and the computer will say, Ah, 1990 Santa, front bumper. This is how it is. And it'll be instant. I guess you could say that's taking over, but, man, I. I think that's great. What else? What other questions you guys have for me? Remember Jeff Hawkins, Numenta, how the brain works today? Remarkable people. Nope. That question flashed too fast for much for me. Jeff, what was that question? Laurie, if we have 150,000 programs in our brains, and it's true our brains degenerate with age, I don't like how this question is going. Then is it reasonable to think we can program, reprogram our brains to pivot careers, areas of expertise, etc. What is the limit? How many degrees, programs, skills can we cram in there, huh? Well, I. That's a good question. I don't know. I must admit, though, that I never make this really intellectual, organized decision. Like, okay, so, guy, you know, you're 67 years old. You probably used up all the models in your head. You cannot learn a new skill. That just never enters into my psyche. This is what in Carol Dweck's word would be the growth mindset, which I believe that you can always grow. So I have to believe that there's space enough in 150,000 models to take on a lot of new features and a lot of new functions. I'm not too worried about that. I'm going to run out of challenges before I run out of models, I guarantee you. What else? See, I don't know. I mean, it's interesting how many of you are in are afraid of Siri and Alexa. Frankly, I don't find Siri and Alexa that smart. So maybe they're just lulling us into complacency and they really are smarter than we think. And then one day, when we're not looking, they're just going to take over, which would be the dystopian view of the world. Maybe that's Hugh Howey's interpretation, but that's not how I look at it. Thoughts on Elon's neuralink? Sinking human brain and A.I. i must admit, F. Deniz, Denise, whatever. I'm not familiar with that model or. I don't know. You know, Elon is just so all over the map. I don't know what to think of him because I understand the cars. I understand the tunnels to a lesser extent than the cars. I understand the solar panels. I kind of understand going to Mars. I don't understand his whole thing with cyber currencies. Like, what the hell is that about? So I think cyber currencies is crazy. So, you know, not to go down a hole, but why would you want to hold an investment that depends on what emoji Elon Musk uses? I mean, just step back for a second and say, okay, so would you like to be paid in a currency that's going to fluctuate massively depending on what someone who runs a car company uses in his tweets for an emoji? I mean, that just. That just doesn't parse to rational to me. But anyway, what do I know? Someday cyber currency will take over the world and you can say, God, you're such a dumbass. In 2021, you had no idea. Oh, God, I. Oh my God, I can't even pronounce your first name here. I'm going to call you tid. Okay, so tid. Will AI make people think less? Arguably, I would think that AI will free people up from thinking about the mundane to thinking about higher level issues. I'm an optimist, so I would love AI to take care of a lot of the mundane stuff. I would also like AI. Let's take a car example. I would like AI to take and make decisions faster than I can. So I believe that with technology, AI can figure out and look ahead, you know, to traffic signals, predictions of cars coming at me, cars going perpendicular, pedestrians, bikes, weather conditions. I don't know, you know, my emotional state, whatever. And AI can look at all those factors, all those terabytes of data, and hit the brakes way before I would recognize I should hit the brakes. Which is not to say that I don't think that those people who are sitting in the backseat of Teslas are dumbasses, because that, to me, is a definition of dumbassery. But I think that a machine can think faster and better than me for those kind of situations. So I can't wait for the day where autonomous driving is true and it ain't true yet. Speaking of apps, Steve. Steve, I can pronounce your name. Speaking of apps, what is the name of the app that allows you to stream on multiple platforms simultaneous? Steve, Steve, look in the upper left hand corner. You see where it says powered by Restream? That's it, baby. Restream. This. This podcast is sponsored by Restream. I don't want to get into trouble with the ftc. So, yeah, powered by Restream. What other questions do we have here? How do we control our brain? And how do our brains store some memories over the other? That's a good question. I. I don't know the answer to that. Maybe you should read the book. I. I can't tell you that. I know that he handled that specific question, though. I read the book. Obviously, I. You're gonna have to read the book. I'm. I'm hardly a neuro. Neuro. Neuroscientist. Okay? I'm barely an evangelist anymore. How many. Yeah, seriously, how many of you do you think. How many of you think that AI is a threat to humans? I can honestly say that I don't. I think that the greater threat to humans right now is climate change and the Republican National Party. Those are the two biggest threats to humanity right now. I know. I. I know how everybody loves when I get political. And now we're going to get all this. I'm going to get all this hate mail, but, you know, how dare you introduce politics into LinkedIn. LinkedIn is only for professional networking. Says who? So, anyway, got another question for me. How do we categorize our memories? Also, good question. I think it's probably a model for the models. I don't. I. Not that this is an answer to your question, but a friend of mine named Dave Weiner created a category of applications Called Outliners. And this is way back in 1982 and three. And so he had a company called Living Video Text and he had a product called Think Tank. And Think Tank was the first outliner. Well, I don't know if it was the first outliner, but the first outliner that I used and was generally available. And so I thought that was just the greatest thing. So you could, you know, you press return and it makes a head, a header at the same, a heading at the same level. And if you tab, then that makes it go, you know, one level below it and then you tab again so you could go making an outline. And then the magic of what he did with Think Tank is that you could select an outline and say, make this into a presentation. And so you could say, all right, so all these headers, all these head, let's say heading one, all heading ones means it's a new slide. So heading one, heading hand heading one, heading one, heading one. That would mean you have four slides and then under heading one you would have these sub points and they would be bullet items in your slides. So way back then, using Think Tank and the next version of the product was called More, you could create bulleted slides instantly. And now that's. In those days, that was a big deal, believe it or not. And, oh God, I still remember what we used to do is we used to take stacks of 3m acetate overhead plastic sheets, put them in the laser writer, run Think Tank on it, and we come up for, you know, 50 slide presentations of text with bullets. And that was a state of the art. That was so cool. Now fast forward till today. Fast forward till today shows you how great the world has become. So, you know, many of you are stuck in a PowerPoint world. I'm sorry, I used to be stuck in a PowerPoint world just like you. But I recently completely and utterly left that world behind because I have just embraced completely and utterly Canva. So Canva's presentation, Canva has a design type called presentations. They're 16 by 9 and there are over a thousand designs created in advance. So you go to Canva, you go to the presentation design type, you pick out one of the templates, and then you put in your own text, you put in your own photos and graphics, or you can also search for anything. And I'll give you a good example that one of my key points in my speeches is that to be innovative, you need to be able to change your mind. So the top line statement is in a change your mind. And I needed A photo to illustrate that. So you go to the Canvas stock photos database and you type in change or something. And I found a picture that shows railroad tracks. And you know, sometimes in a railroad yard, the one track is going down and then it goes off this way and it becomes a Y and it goes from one track to two tracks. So, so I found that picture and that's what I use. So now within canva, this did not used to be true, but I'm telling you, certainly true. Now within canva, if you search through the millions of images they have, you will find an image that will illustrate almost every concept you have. Okay, so now I didn't mean to turn this into a Canva commercial, but anyway, you should know because many of you are stuck in a PowerPoint world. So now you might be asking, okay, guys, so great. So you can use canva, you can use the design type, you can use the templates, you can make beautiful designs. But I still live in a PowerPoint world. I still need to use PowerPoint. Well, here's the beauty. Now you can export a Canva presentation into the PowerPoint format. So you build in Canva, you deploy as PowerPoint. That is good in and of itself. But wait, it gets better. So you can also do it where you take the presentation and you record yourself in video. So it's picture in picture. So if you want to send a presentation to people who are not watching you live, you can record an explanation for each slide picture in picture video and then send the video to somebody so it becomes as if you're live explaining the presentation. And I think that is just earth shattering how much better it could be in a virtual world where most presentations are not in person anyway. So I don't know how we got from the question of how do you categorize stuff in your brain to this. Well, the transition was, the question was how do you categorize brains, categorize thoughts? That led me to talking about outliners with Dave Weiner and Think Tank, which led me to talking about presentations, which led me to Canvas. So I have 150,001 models in my brain. So can you give us the book name you're referring to? Huh? Well, what book am I? Oh, oh, if you're asking about Jeff Hawkins books, it's A Thousand Brains. So yeah, which is interesting. So you love this part of the interview. So his book is called A Thousand Brains and he was Talking about the 150,000 models in your brain. So I said to him, well, if you're saying that there's 150,000 models in your brain? Or why is your title A Thousand Brains? Like, whatever happened to the 149,000 other ones? And was one of my proudest moments as a podcaster, an interviewer, when he said, guy, you are the first person to have ever asked me that question, because that is a completely logical question. If you're saying there's 150,000 models, why is your book called a thousand? And so that he has an explanation for that. And I'm not going to spoil the explanation, but I just want you to know, man, I am on top of things when I interview somebody. All right, what was that question? That flashback, Jeff? I didn't even see it, but I saw something happen. Okay, Jeff, bring it back, baby. I'll tell you another proud moment in my history of podcasting. So, in a couple weeks, there's a guest named Joe Bowler, and she is a professor of mathematics education. Not mathematics, mathematics education, at the School of Education of Stanford University. And her whole thing is about revamping how math is taught. Basically, she thinks that teaching math the way we've been teaching math since the 1800s is ridiculous. And making people write out equations and memorize stuff and all that is ridiculous. And so I don't know how we got on the subject, but somehow I started talking about reverse polish notation calculators, and she had never heard about reverse polish notation calculators. So here's a. Here's a icebreaker for you. How many of you know what RPN calculators are? RPN reverse Polish notation. So, Jeff, put that question back. Man, I like it. When I'm talking, I can't talk and read at the same time. Which is more dangerous, AI bias or human bias? Well, if you're asking my opinion, I would say human bias. I think that I'd be more willing to take a chance with AI bias than human bias. One of the key messages in Jeff's book is that the understanding that I don't think it goes this far, but it's almost like there is no right and wrong. It's just people have different models, so people could look at the exact same facts and come to very different conclusions. And it's not that one is right and one is wrong. It's just it depends on your model. And that. That is a key, key thing to think about when you, you know, try to understand how, for example, some people look at vaccination as the holy Grail, and some people look at it like, you know, Bill Gates is injecting a chip into you. It's. It's a matter of the models. So this is another good icebreaking question. So do you use AI for any part of your job? I can tell you with my podcasting, I use a service called Descript. And what Descript does is it takes the audio and it transcribes it for us. And that transcription has got to be doing it with AI and machine learning and all that stuff. Having said that, I will tell you that I don't think the AI is too smart when it transcribes. Because let me tell you something, I'll give you some insights into how the sausage is made. Let's say I conduct an interview now, we upload the audio. The audio is then transcribed. And I will tell you that to go from the machine transcribed transcript to something that I consider usable and readable as a transcript, that is something that someone would read and get more or less an accurate representation of what was said and as grammatically correct as possible. If the interview is one hour, I probably spend two to three hours going over it, making changes, cutting out extraneous stuff, cutting out filler words, but also actually changing the transcript as opposed to the audio because the transcript is simply wrong. And so I spend two or three hours doing that. And after I do that, I send it to several people who are working as contractors for us and they spend another three hours. And, you know, I tell them to basically think of the Chicago Manual of Style as the Bible and you make this transcript match the Chicago Manual of Style. So I want, I want commas and conjunctions between two independent clauses. And I want EM dashes. I don't want commas when it's, you know, a thought, a pause for a thought, or going, going slightly in a different direction. I want an EM dash, I don't want a comma. And I want numbers under 99 spelled out and over 99 as numerics, unless the number is the start of a sentence. And then if it's the first word in a sentence, I want it spelled out even if it's under 99. And I don't want things to say gonna and dunno and coulda. I want it in the transcript to say, you know, proper English. And that takes another three hours. So there's probably six hours into making a decent transcript from a one hour recording. And I will tell you that you're going to hear about it soon because I interviewed two people. One is Martin Lindstrom, the author of Biology B U I O L O, L O, G Y. And Haben Girma. Haben Girma is the first person who is deaf and blind who graduated from Harvard Law School. And so after talking to them and talking about how much effort it takes to make a transcript that I'm not sure anybody is reading, I was convinced that I have a moral obligation and it's good marketing to have a great transcript. So we invest, I would say, six hours into making a great transcript for you. So when you go to remarkable people.com, the website, and you scroll down through an interview, you will see full text transcription. And I only started this about four weeks ago to really put any effort into it. And I hope you appreciate that transcript because, man, there's a lot of work in that transcript. I don't know. So, anyway, so that was a long answer to question of do you use AI in your work? And I could tell you the state of the art of automatic transcription. AI is not so great right now. So, yeah, this is. I didn't realize this. Somebody told me this. So if you say, hey, sir. Oh, if I say that now, my phone might do it. Hey, Siri. Or I know Alexa, play Guy Kawasaki's Remarkable People podcast. It will do it. My goodness. Oh, okay, this is funny. Okay. Oh, my God, that's funny. So in my little studio, this is my Alexa. I don't know, what do you call this? Which model is this? Hawkins. Computer architect, neuroscientist, and author. Huh. Just so you know, I can help if you ever misplace your phone. Try asking me to call your phone. Huh? Alexa, call my phone. I didn't know Alexa could do this. Hello, this is Guy. Oh, my God, that's funny. Okay, so anyway, hey, I really. That's funny. That's great use of AI. Oh, man, if now AI should scare you. So anyway, that's this week. That's Jeff Hawkins on Brains and models and neuroscience, Science and neocortex. So if you want to make your head explode this week, listen to Jeff Hawkins on the brain. You want to join my ohana group. 831-609-0628. Remember that this live stream is brought to you by Restream. And. Oh, my God, it's too late anyway, so I can beat all the AI back into their little boxes here. I think we're going to end this interview. So have a great week. Remember, Remarkable People dot com. I hope you enjoy it. And please, please, please, please, please read the transcript because, man, there's so much effort in those transcripts. Take care. Thank you. Bye.