Marques Brownlee on the future of creators

What if the best technology were deployed on the biggest problems for humanity? So asked Marques Brownlee, who joined us on Possible to open the first episode of season 2. Also known by his YouTube channel name, MKBHD, Marques is a self-taught YouTube creator who has nearly 18 million subscribers and has been posting tech-focused reviews and analysis on YouTube for nearly 15 years. Marques, Aria and I talk about the ways AI enhances his work and lowers the barrier to entry for YouTube creators today. He also discusses the many arcs of technological progress, the power and promise of electric vehicles, and what makes a good piece of technology. Plus, Marques has a conversation with Inflection’s Pi. Read the transcript of this episode here. For more info on the podcast and transcripts of all of the episodes, visit www.possible.fm/podcast. Topics: 5:49 - Hellos and intros 6:58 - What problems can AI help tackle? 9:59 - Using AI as a YouTube creator 12:48 - How AI can help the next generation of creators 21:23 - Where Marques got his love for gadgets and technology 23:01 - The arc of technological progress 24:35 - The future of electronic vehicles 29:25 - Marques’ investment philosophy 34:03 - Marques talks to Pi 38:28 - Pi’s potential uses and other AI assistants 42:50 - Rapidfire questions The award-winning Possible podcast is back with a new season that sketches out the brightest version of the future—and what it will take to get there. Most of all, it asks: what if, in the future, everything breaks humanity's way? Tune in for grounded and speculative takes on how technology—and, in particular, AI—is inspiring change and transforming the future. This season, hosts Reid Hoffman and Aria Finger, are speaking with a new set of ambitious builders and deep thinkers about everything from art to geopolitics and from healthcare to education. These conversations also showcase another kind of guest: AI. Whether it's Inflection’s Pi, OpenAI’s ChatGPT or other AI tools, each episode will use AI to enhance and advance our discussion.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  One of the most common questions that I get because of how long I've been a creator is, "Could you get started again today? And if so, like, how could you get started?" With all these AI tools, I do think the barrier to entry is lower than ever before. Now, because of all these tools to help you brainstorm and help you come up with new ideas and refine these ideas, suddenly, if you're good enough at this, the barrier to entry to making a really good, interesting story is also lower.

• 0:33Reid Hoffman?

  Hi, I'm Reid Hoffman.

• 0:35Aria Finger?

  And I'm Aria Finger.

• 0:36Reid Hoffman?

  We want to know what happens if, in the future, everything breaks humanity's way.

• 0:42Aria Finger?

  We're speaking with visionaries in many fields, from art to geopolitics and from healthcare to education.

• 0:50Reid Hoffman?

  These conversations showcase another kind of guest. Whether it's Inflection's Pi or OpenAI's ChatGPT, each episode, we use AI to enhance and advance our discussion.

• 1:02Aria Finger?

  In each episode, we seek out the brightest version of the future and learn what it'll take to get there.

• 1:09Reid Hoffman?

  This is possible.

• 1:17Reid Hoffman?

  YouTube, what was once mainly a refuge for cat videos and pranks, has become a multi-billion dollar hub for creators and advertisers. Its reach, as well as the reach of its newer powerhouse rival TikTok, is still often underestimated.

• 1:34Aria Finger?

  So let's put some numbers on that power. YouTube took in nearly $6.7 billion in advertising revenue in Q1. The Wall Street Journal reported late last year that investment firms are increasingly spending big just for the chance at a cut of ad revenues from these growing creators. One firm even announced that it planned to invest $1 billion into those licensing agreements. Another way to look at it, the person that we're talking to today has 17.7 million subscribers on YouTube. As a point of comparison, The Vaunted New York Times, they have nearly 10 million total subscribers across digital and print.

• 2:15Reid Hoffman?

  What that means is content creators have immense power and influence, not just with their viewers, but over the market. And with the new AI tools and technology that creators have at their disposal, that industry only stands to grow even more.

• 2:32Aria Finger?

  There is so much out there right now. Chat GPT can help write scripts. Pi can draft emotionally intelligent dialogue. Midjourney and Dolly 3 can create stunning visual art and videos. There are even AI tools for assessing the latest trends in content creation. Even the text to speech tools we hear on thousands of TikTok videos every day. That's driven by AI.

• 2:57Reid Hoffman?

  So how should we be thinking about this growing, changing, complex market? How can both seasoned and aspiring creators, as well as the rest of us, use new technologies to the fullest? Today's guest, no doubt, has some answers.

• 3:12Aria Finger?

  Marques Brownlee is also known by his YouTube channel name, MKBHD. He has been posting tech-focused reviews and analysis on YouTube for nearly 15 years. As a self-taught creator, he started his YouTube career filming in the corner of his basement in his parents' house on a blurry webcam. Now he has nearly 4 billion views on YouTube.

• 3:37Reid Hoffman?

  On his channel, Marques has hosted the likes of Bill Gates, Barack Obama, Elon Musk, Mark Zuckerberg, Neil deGrasse Tyson, and Google CEO Sundar Pichai.

• 3:49Aria Finger?

  He's also host of the podcast Waveform and a professional ultimate frisbee player.

• 3:55Reid Hoffman?

  We sat down with Marques in his New Jersey-based studio to dig into his YouTube career, his best advice for creators, and his thoughts on the future technology and society writ large. One of the things that was really awesome about talking with Marques was that it bridged from what counts as a good technology project and how that could reach into bigger things in the side. Like, thinking about it, it's not just the, well, this phone versus that phone, this EV versus that EV, but also the kind of like, okay, which of these things makes a difference for humanity locally, you know, individuals, and humanity globally?

• 4:39Reid Hoffman?

  And he brings the kind of the entrepreneurial, new media lens where his, you know, it's like, well, in order for it to be worth covering, it has to be better than fine.

• 4:54Aria Finger?

  Yeah, I was thinking along the similar lines is you think he's such a gadget guy. Like, he loves like how things are made. He loves tinkering with things. He loves seeing what the difference between the latest iPhone is and the one before. But then we really got into talking about, again, another sort of piece of hardware that he loved, electric vehicles. And with the adoption of AI, electric vehicles could be talking to each other. Not only could you be learning from your EV, but it could be warning another EV down the street about the child playing. Or like, we could create this whole society because of the hardware and software together, that could be so much better. And it was interesting to think about, how are we going to get to that future? Right? It's going to be a wave

• 5:38Aria Finger?

  before we get there, but how do we shoot for it? Here's our conversation with Marquez Brownlee. So, Marquez, it is delightful to be here, not least because Reid and I have a little East Coast-West Coast rivalry going.

• 5:58Speaker 00

  Interesting.

• 5:59Aria Finger?

  And you are such a New Jersey stalwart. But I have to ask, because you grew up in Maplewood. And the thing I know about Maplewood is the Maplewood Pool is the best. It has the best grilled cheese, the best water slides, the best everything. Do you agree, disagree? Do you have fond memories? Tell me about the Maplewood Pool.

• 6:17Speaker 00

  Agree. I have good memories from Maplewood Pool. I have one bad memory, because I never really got used to the diving board stuff. And the last time I was there, which was a long time ago, but they had the diving board, which was like, I don't know, six feet, 10 feet, whatever. And I got a little excited and went a little too high off the diving board. And it was one of those things where you land in the water, and then you like, think you're going to swim to the top, and then you're not at the top yet, and then you have to swim again to the top. And it's just, just one bad memory from the pool. But I would say, yeah, Maplewood's got a lot going for it.

• 6:46Aria Finger?

  Super clutch. I love it.

• 6:47Speaker 00

  Yeah.

• 6:48Reid Hoffman?

  And that is probably our most unusual beginning to our podcast ever. But that was the reason I was curious because I don't even know what this pool is, although I can guess. So what device would you most want to see AI brought to? And and why? And what would that do in terms of like expanding our possibilities?

• 7:11Speaker 00

  There's so many things, especially because it's we're right in that time where AI seems to be just be popping up everywhere. Especially I play golf and this is one of those things where there's not too many gadgets in golf. But immediately we're like, we're going to use AI to fix your swing, to help you hit yardages more consistently, to play better golf. And I was like, oh, I love this. This overlap is awesome. But where hasn't it shown up yet? There are maybe a couple last big uh frontiers that maybe I'm not as familiar with, but things like manufacturing, things like just generally uh attacking climate change and big human problems with AI. Because we see all the the the customer facing stuff.

• 7:52Speaker 00

  We see chat GPT, we see uh all the little tools that show up in Photoshop and things like that. But I just, I figure it's it's getting so powerful so quickly that we want it to address humanity's biggest problems. And ideally, it is being deployed to do that as much as it can.

• 8:06Reid Hoffman?

  A hundred percent. I mean, I think to kind of speculate a little bit, like for example, but even take this conversation we're having here. We're in your really great, excellent studio. You could imagine that AI might be listening to the conversation we're having and then say, oh, ask Marquez this question. Or or hey, this would be a a theme with with Reed. Or um hey, did you know, for example, did you know Marquez was really loved golf? And there could be something really interesting about how a golf could help Marquez with his swing. Right? Like as particular, you know, and that kind of thing would is part of the amplifying effect of humanity. Because it's it's the how do you make us connect with ourselves with each other more?

• 8:51Reid Hoffman?

  Is actually, I think, part of the thing that people are not focusing on. There's like, oh, it's artifact, it's gonna be AI. It's like, no, it's gonna make our lives richer.

• 8:58Speaker 00

  I find that fascinating, especially the podcast version of this. I think I I've interviewed some people on podcasts before or just tried to do good interviews in my media career. And one of the things that I found is a really, really good. Interview typically asks questions that have never been asked of that person before. Like one of my favorite interviewers, Sean Evans, you've seen Hot Ones. He'll ask questions that are not just like bounding off of something else that they've been interested in, but it's like well researched and the timing of it is great. And it's just that moment of the good question and the thoughtful answer is great. And I wonder if AI trained on everything that's already happened could come up with something unique that hasn't already happened. Of course, it'll come up with great things alongside what's already happened.

• 9:40Speaker 00

  But I just, I love those human moments of like, now that's good. And I wonder if AI can actually help with that sort of thing.

• 9:47Aria Finger?

  Right, and it, can it give the creative spark that we need to sort of go to the next level? And can you say like, search every podcast I've ever been on. I don't want to answer, ask the same question again would be sort of awesome.

• 10:00Aria Finger?

  So I rewatched your video introducing Dolly 2. And unlike many people who, you know, they see Dolly and they're like, oh my God, it's going to put people out of work. You know, this is terrible. It was so clear that you were like, this is awesome. This can do so much. You talked about your graphic designer and also how far it would come. We've come a long way since Dolly 2, you know, a year ago. What would you say about the evolution of that for your own work? If we take it down from humanity to what you're doing, how can you use AI to amplify you?

• 10:32Speaker 00

  Yeah, so I think one of the biggest conclusions I came to when just trying this stuff for the first time is it's not going to replace me, but it is really, really good as a tool to help me make better stuff, whether it's make faster stuff or make the same quality stuff more easily or make better stuff than I ever have. And so with the tools getting even better and accelerating and how much better they get, I feel like that's become even more true. And the one interesting sort of tweak to it that I've sort of added to my perspective on it is the skill now that becomes important for people like me is being able to create a good prompt that can actually help it help me.

• 11:14Speaker 00

  I think if you ask someone like 20, 15, maybe 10 years ago, what their most useful skill online would be, it's like, I know how to type a certain set of words into Google to get exactly what I need to find. And I think this next sort of evolution of all these AI tools is like, I know how to turn these knobs or to type in this prompt to get exactly what I need out of this AI. And that's my skill and that's why I'm able to be so good at what I do.

• 11:38Reid Hoffman?

  Because engineers make this, they describe it as prompt engineering. I think this alienates people because most people go engineering. That's what a few other people do. It's this arcane thing. It's only kind of geeks like the three of us who are like, woo, engineering, that's kind of fun. But I think of it more as like prompt directing, right? So it's kind of like, like the direction you'd give an actor, the direction you'd give. Have you played with your prompt directing at all? Because like the kinds of easy tips are like, do this in the style of Ernest Hemingway.

• 12:08Speaker 00

  Yeah, yeah. I mean, so the most common thing that I do is I like to end reviews within alliteration. So after the, the, the big pixel review I did last year, it was like the, the pixel perfect, something, something. I made some great analogy. A lot of that came from me asking ChatGPT for an analogy based on my video script. And the more I do this prompt engineering, the more I find that there are certain things that I can ask for. Now make it shorter. Now make it more punchy. Now make it, you know, this and that. And I keep trying new things. And I think the more you do the same type of prompt over and over again, the more you find you can sort of twist it and bend it and find ways to do what you want better.

• 12:48Aria Finger?

  Well, so, you know, the whole promise of technology is democratization. And so, you know, thousands of years ago, only a, you know, subset of people could write and then only a subset of people can read. And now with more AI, you were, you were one of the first creators. You know, you were creating YouTube videos when you were in high school. And part of the promise can be that now anyone can do it. Anyone with the help of AI or ChatGPT or Dolly can be a more prolific creator, can be a better creator. Like, what do you think about these AI tools helping sort of the next generation of creators come on board?

• 13:22Speaker 00

  Yeah, it's really interesting. I, one of the most common questions that I get because of how long I've been a creator is, could you get started again today? And if so, like, how could you get started? And while I don't have a perfect answer for that because the landscape is so different, I do think that the barrier to entry is lower than ever before. So even 10 years ago, 15 years ago, when I was starting, you had to have some sort of knowledge of how to get these files together, edit them together, put it on, online. YouTube was relatively young. Now it's like, the phone in your pocket, that's a 4K, 30 FPS, high quality video, like bang on instantly. And with all these AI tools, I do think the barrier, not just the barrier to entry, but the barrier to entry to good quality content is also lower than ever.

• 14:07Speaker 00

  So again, like you can get, you know, a 4K 30 FPS video out of your phone, but it might not be that interesting, or it might not have that great of a topic. And I think now because of all these tools to help you brainstorm and help you come up with new ideas and refine these ideas, suddenly if you're good enough at this, the barrier to entry to making a really good, interesting story is also lower. So that's the one thing I think is consistent. I still couldn't tell you exactly how to start a YouTube channel in 2023 because all I can do now is run a 15 year old channel. But I can say the barrier to entry is lower and people are making really good stuff because of it.

• 14:41Aria Finger?

  I mean, I love you saying the barrier to entry to good content is also lower because I feel like we're having the same conversation. 15 years ago, everyone was like, yeah, YouTube's fine, but it's all crap. It's just going to be millions of videos that aren't good. And people are saying the same thing with AI. Oh, we're just going to get millions of books that are bad and millions of this that are bad. And that might be true, but to your point, it can also raise people up and it can also create better content. So it's not just that it can create schlock, but we can also see the best of the best become better. The best directors, the best creators, the best people can still level up with AI, which I love.

• 15:14Speaker 00

  Yeah, yeah. And there's more, it's still right that there will be a lot of random garbage. And that's fine. And that's how YouTube started too, and that's fine.

• 15:20Aria Finger?

  Yes.

• 15:21Speaker 00

  But the fact that we will get a higher volume of high quality stuff is awesome too.

• 15:24Reid Hoffman?

  Yeah. And by the way, that's been the true history of media anyway. If you look at the explosion of cable channels and much other stuff, just go watch the infomercials on God. Okay, this is not great.

• 15:35Speaker 00

  Yeah.

• 15:37Reid Hoffman?

  Let's talk a little bit about the kind of the, the, the inspiration behind various kinds of content creation. And we can be YouTube focused or, you know, online media focused. But, you know, part of what I was a little, as usual, disheartened by, like the whole Hollywood discussion around, oh my God, this will be replacing human beings. Like, look, let's be focusing on how to make more human stories, how to do this better, how to amplify human beings. I can see all kinds of ways in which AI from, you know, just script writing or research assisting or, or challenging or plot twists or alliteration, you know, as kind of ways of doing it. What, what have been some of the ways you've been thinking about the, this is new vistas or new heights that are enabled by AI?

• 16:23Speaker 00

  Yeah, so I'll give you an example in my own workflow. So I make a lot of YouTube videos and those videos go online to an audience where I'm fortunate enough to have a ton of feedback the second the video goes up. So there's hundreds, thousands of comments from people who've watched what I made and reacted to it. Super useful. Now, what I've found is like I've gotten really good at certain types of videos. I can review a new gadget. I can test or explain a new thing. But since I've been doing it for so long, I want to sort of branch out into new things and find new things my audience is interested in. A new tool comes along and I'm not going to remember the name of it off the top of my head, but essentially what it lets me do is it sorts through all of the comments across all of the

• 17:03Speaker 00

  videos on all of the channels that I make and finds the most common things that people in my audience are are talking about amongst themselves and wondering about in their own questions amongst themselves. And that gives me insight into new ideas of what I can then make videos about for them. So even without them asking for a certain type of video or me having to go to Twitter and be like, what do you guys want to see? Like, that's not going to help me very much. But these these sort of summarization tools and insight tools I'm able to use are genuinely very good and have given me video ideas that I've then executed on and I think have gone really well. And people go, I didn't even realize I wanted this, but this is great. The the fact that that actually works and it's a real workflow

• 17:44Speaker 00

  that created better content that exists in the world in 2023, and this is the worst it'll ever be. Right. I think is pretty sweet. I think that's a pretty promising start.

• 17:54Reid Hoffman?

  What are some of the most interesting side angles you've had in creating your show? And one of the reasons this question occurred to me is obviously we walked by the casket on the way into the studio.

• 18:06Speaker 00

  I can explain.

• 18:09Speaker 00

  You you never want to be predictable and you want to keep your audience on their toes a little bit. You want to be reliable, but not predictable. I think that's a good way of saying. The casket, if he would, if I could explain, is, you know, as a as a review, as a product review channel, we get a ton of emails and requests and 99% of them we say no to, but. You know, just around the studio for a while, I think I just tweeted like, "One of these days, I'm gonna say yes to everything, just so you guys can see how ridiculous the inbox gets." And at the studio, we were like, "Wait a second, that's, "Let's actually do that. Let's see what happens." And we just picked them up and we said yes to everything. And a casket and a bunch of weird cases and all kinds of crazy stuff shows up.

• 18:50Speaker 00

  But I think that comes from just having a lot of smart people at the studio and many minds is better than one mind, so that's fun. But yeah, we have all kinds of like tangential weird angles to things that seem fun, just because like, "We already know what the bread and butter is, which is new product comes out, is it good or not? Should you buy it? Yes or no?" We can do that all day. But if we only do that, that's gonna get boring. So we like to find the business angles, the strategy angles, the stories, the things that sort of "are intertwined with all these product releases and have fun with those two." And it's funny because in the tech world, it is kind of surprising how much stuff is sort of like "right in the middle, average." Whenever you ask about like, "What are the hardest products to make?", tech products are kind of the hardest to make.

• 19:33Speaker 00

  And so if you get everything together in like building a company, getting your product off the line, like, "It goes through so many lines of approval and people that by the time it comes out, "It's gotta be at least decent at something to exist, to have gotten this far." So endless products show up here at the studio that are completely "fine." They're just fine, you know? And they're not like exciting or interesting to make a video about, and I could, "And there are gonna be people who buy them, but like, "I don't know, that it kind of feels a little boring to focus on what's fine." So it is especially exciting to find things that are really, really good "and things that are really, really bad." Because then, you know, it's the PSA where it's just like, "Just so you know, this is actually incredibly bad, don't buy this."

• 20:17Speaker 00

  "Or this is really, really great, it's surprising how great this is, "You should probably consider this over the other options." And we've done both. Have you seen our Dyson headphones video? I called it "The worst product I've ever reviewed", "And it might not be the worst product I've ever reviewed, but it's pretty bad." "It's one of the worst things I've ever made a video about." It's, so Dyson makes vacuums and filters and all kinds of other things like that. They made a pair of headphones "that includes air filters in the ear cups and then a connected, like breathing tube around your mouth, "where it filters in clean air and passes it over your nose and mouth." And this is another collab I did with a doctor who was like, passing cold air over

• 21:00Speaker 00

  your mucus membranes 100% of the time is already a bad idea. But also here's all the other reasons why this is a bad idea. And it was a really fun video to make, but I guess the point is not that many truly horrible stuff actually.

• 21:12Reid Hoffman?

  Oh, cuz, cuz, cuz of the filtration of what it takes to actually build these things.

• 21:16Speaker 00

  Yeah, there's so many layers and so many, so much money required that you usually don't get to a final product without it being decent.

• 21:24Aria Finger?

  You obviously have a deep love of products and gadgets and like, where did this come from? Like, from a young age was this here? Like, was there a moment? Tell us.

• 21:33Speaker 00

  It's one of those weird things where I feel like I've always been into tech. I do have a super, super early memory of playing with the JVC camcorder in the like, everyone has like a junk drawer or a junk closet or whatever my parents had. And I just like, there was a camcorder in there and some tapes and I figured out how to charge the batteries and I just ran around the house taking videos of things. And it's not like I could edit anything. So I was like editing in tape and like making a little movie on a camcorder. And like, that was so satisfying as a however many nine years old, whatever I was. So I, it doesn't get much further back than that for me, but I just feel like I've always been interested in making stuff and interested in products and tech that actually work.

• 22:13Aria Finger?

  That's awesome. You were growing up in the start of the internet age. You obviously, the mobile age sort of came about right when you were, um, uh, starting with your YouTube videos. Now, obviously the age of AI, like, where do you think we are in the arc of technological progress? Like, you've seen gadgets change. You've, it's a big question. Um, you know, do you have thoughts about sort of where we are at this time in history? Where we're going? Do you even think about it? Maybe you don't care.

• 22:37Speaker 00

  I kind of do. I feel like there's a lot of arcs. There's a lot of arcs that sort of follow similar shapes, but they are all happening at different times. So like, if you just asked about the arc of smartphones, which is one that I've followed extremely closely, like, we had, we had that explosive moment, which a lot of people is 2007 and we get up to like, What's the difference between this iPhone and the last iPhone?

• 22:57Reid Hoffman?

  Totally.

• 22:57Speaker 00

  It's very incremental at this point. We're up here. Like, they're mature. But ask about EVs. Electric cars, for example. They're having that explosive moment now where like, we went from five years ago, almost nobody even considering an EV, to five years from now, you will have dozens and dozens of really good competing EVs in multiple categories to choose from. But we're right at this interesting point where like, there's only three electric trucks, so every single one gets compared to the existing three.

• 23:26Reid Hoffman?

  Totally.

• 23:26Speaker 00

  And it's like, we're gonna, we're gonna get there eventually, but like, that's where we are in that curve. So I just see a bunch of curves all happening at various times. And I guess it sort of depends on which one you're asking about, but that's how I see it.

• 23:37Reid Hoffman?

  Well, by the way, electric vehicles, I mean, I've, you know, this is said more as a personal, although I've done a bunch of investing in autonomous vehicles and so forth, because I actually think, you know, we're all gonna, there's 40,000 deaths and a lot more injuries per year in the US alone. We're gonna see a huge improvement on that and all this stuff. And EV will be part of it. I think one of the funny things that, the only thing I would say about the EVs that may not be common knowledge is that it's actually, in fact, a software evolution. It's moving cars from mechanical engineering to a software engineering.

• 24:09Speaker 00

  That's true.

• 24:09Reid Hoffman?

  Paradigm.

• 24:10Speaker 00

  Yeah.

• 24:10Reid Hoffman?

  And so, the, what happens when the EV becomes the equivalent of a software platform that's programming for a data center that is using not just autonomous vehicles, but everything from, like, you can imagine, like, for example, a future software thing is like, a person's having a heart attack on the road, and it says, "Oh, I see the person's having a heart attack", and it immediately diverts and takes them to the hospital.

• 24:33Speaker 00

  Yeah.

• 24:33Reid Hoffman?

  Right? Like, that kind of thing is part of the, what, what do you see is this, like, I completely agree with you at the beginning of the thing. What, what does that future curve for EVs look like?

• 24:42Speaker 00

  So it's interesting, even within EVs, as you're describing this, I feel like there are two curves. One is the software connectivity, the software platforms that are our cars. But the other is definitely the hardware, which is just like, every once in a while, I'll test or drive a gas car. And the difference between, uh, like, an average EV and an average gas car is, it's massive. The EV is gonna weigh 1,000 extra pounds. The EV is gonna be 10,000 extra dollars. And the EV is gonna have its certain advantages, like instant torque, like more space. It might have two trunks, you know, there's a whole bunch of things that come with that hardware platform versus the gas one. And I think interestingly, the the software stuff is happening kind of at the same time,

• 25:22Speaker 00

  maybe just because of efficiency. If you're if you're trying to make an autonomous car and you have a gas drivetrain, it's not able to be as responsive as just a computer on wheels. So I think that kind of coincides a little bit. But I always talk about this one thing when I make these EV reviews, which is what are the, we're so early with EVs. What are the best types of cars to go electric? Because when you go, I always use the Cadillac Escalade as an example. When you go electric, you're going to make it way heavier, you're going to make it way more expensive. You're also going to make it quieter, smoother and more responsive to drive. So let's pick the things that are already heavy and expensive and just make them quieter and smoother and more responsive. So the Cadillac Escalade, what better car to go electric than the big, heavy, expensive seven-seater?

• 26:08Speaker 00

  And now they're going to make it an electric Escalade and it's going to be great and it's going to be expensive, but it's going to be the best Escalade ever. But when you look at sports cars, the engineering of that hardware is so much harder. Sports cars are about lightweight responsiveness. The most impressive sports car I've ever driven that's electric, the Ramaz Navara, $2 million, 2,000 horsepower.

• 26:28Reid Hoffman?

  Did they send one you one of those for free?

• 26:30Speaker 00

  I wish they sent one of those. That showed up in the parking lot last month. Craziest car I've ever driven. But even as you drive it and you corner in that $2 million hypercar, you're like, this thing is heavy. I can feel that this thing is heavy. So there are certain types that are going to take much more of that curve to get to actually being good and mature. And I think we'll get there eventually. Battery tech's got to get better. The electric motor's got to get smaller and lighter. But yeah, you talk about pickup trucks, Escalades, like we're getting the ones that make sense right off the bat, which is makes sense.

• 27:04Reid Hoffman?

  Yeah, no, exactly. And I think, but I think one of the things I agree completely with your views. Um, I recently got a Ford Mach-E, which I like.

• 27:12Speaker 00

  Yep.

• 27:12Reid Hoffman?

  Um, but the thing I keep watching for is when the software parts of it really start taking off. And I think it's, other than autonomous vehicles, I think it's a little too early for some of the AI stuff. Obviously you'd have connected cloud, you can have a much better assistant. You know, one of the things you you put, for example, we mentioned Apple, you put Siri next to ChatGPT. And it's like the the old dead one, dead thing versus the modern interesting thing.

• 27:39Speaker 00

  Yeah.

• 27:40Reid Hoffman?

  But like envisioning what that software defined future is. Because for example, like you can imagine the question of like, what what is the real time collective data map that happens from all of these cars being network device? I mean, think about like how it fixes traffic, how it fixes like all of the stuff.

• 27:58Speaker 00

  Yeah, this is one really interesting reason. I think a lot of people are optimistic about Tesla specifically is because they are, I don't want to say shameless, but they are so obviously collecting so much information from all of this and training their neural net and their computers on all this information that they're getting better hand over fist at a rapid rate because of all the data they have. I kind of wonder if others are going to start doing the same thing as quickly as they have. But yeah, it's like, I I remember talking to Neil deGrasse Tyson about like the future of cars could maybe just be like every car can talk to every other car and you can just take your hands off the wheel and it will communicate and weave and zigzag in a way to get to your exit in a much faster and safer way than you ever could as a human.

• 28:40Speaker 00

  How far away is that future? I don't know, but it feels like a software problem.

• 28:44Reid Hoffman?

  Part of that talking to each other is not just the safety of the cars. Think about like if the cars are talking to each other and you're driving down a city street and car number one says, hey, I saw a kid playing with a ball behind this car.

• 28:56Speaker 00

  Yep.

• 28:56Reid Hoffman?

  And communicates that to the other cars that are driving nearby.

• 28:59Speaker 00

  There are so many good examples just like that, and especially when they can then learn from each other instantly. My my thing about cars, and I think you've probably also seen this, is like people just keep their cars for so long that in order to have every car on the road be that responsive, it's it's a natural built in extra couple decades of just people with their current cars. Um, and then there's this weird middle point where it's like half the cars on the road are new and have this tech and the other half are people trying to avoid things happening with the cars talking to each other. It's crazy. So I don't know how to get over that hump. But we're gonna have to get there to get to that future. So that's gonna be interesting.

• 29:34Aria Finger?

  Totally. So there's so many ways to be involved in technology, like you're clearly excited about EVs, you're a founder, you're a creator, you're a product reviewer. Um, but I also think you're an investor. And so when you think about what are the companies and products to invest in, like, how do you look at that? What's your lens?

• 29:50Speaker 00

  This is interesting. I am a pretty new investor, I will say, but I am. And I think I'm just generally most curious about and optimistic about companies that are, when you sort of extend what they're doing to its natural conclusion way down the road. So obviously now they could be working on a single gadget or a single technology. But if you just use like Tesla as the example of like, okay, build this tech on top of this tech and extrapolate it to all of the categories and move all of it forward and look where the world is now. If you can see that meaningful progress, then I'm interested. That's the type of company I think makes a great investment and just makes sense to be invested in in whatever way you can. So I don't even have any examples off the top of my head,

• 30:32Speaker 00

  but I just, I just generally think like trying to extrapolate forward as far as possible and seeing what that lands me in my head. If I, if I feel good about that, then, then I'm interested.

• 30:42Aria Finger?

  Well, I love that lens. It's like how an investor should be. How can we make the world better? How can we use AI, technology, products to increase humanity? And I think that's great.

• 30:51Speaker 00

  Yeah.

• 30:52Reid Hoffman?

  You know, it might be interesting.

• 30:56Reid Hoffman?

  I don't know. Your platform might make this worthwhile. It's generally speaking, you always want to be surprised by the inventiveness of the network of entrepreneurs is usually how you're trying to do as investors. So like, it's good to have a thesis driven investment. These is just to have a prepared mind and be knowing stuff. But the usual thing like when I'm asked that question is like, oh, what to invest in is like, I want to be surprised by the really different and interesting entrepreneur. Like it's not what I've pretty thought of.

• 31:25Speaker 00

  Yeah.

• 31:26Reid Hoffman?

  But there might be a thing with your platform that might be interesting, which is say, hey, I am intrigued by products that solve, that have a good theory of how to solve this need or this kind of thing. Cause you have enough reach that you might then get someone goes, oh, I'm working on something like that. And you could take a look at it and see if it works.

• 31:45Speaker 00

  Definitely. And yeah, it's also about like, when you see that one solution applied to this one problem, but they clearly have the ability to apply this solution to weigh more problems because it's a clever and a unique solution. Then that I I love that I talk about that and especially some of these bigger companies. They come out with products and I'm I naturally extrapolate like I look at what Google just did with their software and the camera and I'm like, all right, this is clearly like, what's the next phone gonna do? And then the next one after that, like I I guess my brain tends to just fast forward a little bit and see if it it can arrive at its natural conclusion sometime later.

• 32:25Reid Hoffman?

  So this is actually one of the things that I thought has been kind of under delivered. It's partially because, you know, Apple is, you know, kind of so asleep on AI that it's kind of, you know, snoozing in the backseat. Um, you know, Google's obviously more active, but what's A what is AI gonna do to revolutionize the phone? Right? Have you have you thought about that in ways that you can share?

• 32:47Speaker 00

  Yeah, so there's a little a little bit of that from Google already. So if you watch Google's latest event, which is just a couple of weeks ago, they had this moment where they sort of casually went over, we're now combining Google Assistant with Bard and there will be features from Bard built into Google Assistant and now Bard is like improving rapidly and multimodal and has all these features. And then they just moved on and they went with a bunch of other stuff. But in that moment, I was like, isn't this what we wanted from assistants in our phones? Like when we got Siri, when we got Google Assistant, even if you look back at early Alexa, I think the idea behind them is that maybe someday it'll be the computer from Star Trek and we'll just talk to it naturally and

• 33:28Speaker 00

  we'll arrive at it'll just do what you asked for it to do. And they've not really delivered fully on a lot of those promises. I think Siri can do some basic stuff like set some timers and open webpages and apps and things like that. And Google Assistant has some more functionality and I've played with them and I've done videos comparing all of them. But I think as far as AI on phones, I want to be able to just ask my phone to do something and just have it do it to actually do what I asked for it to do. So I suspect that that's going to be a sort of inflection point of getting to that point.

• 34:03Reid Hoffman?

  So speaking of AI, one of the things we-

• 34:05Aria Finger?

  And speaking of inflection point.

• 34:06Reid Hoffman?

  Yes, exactly. Speaking of inflection, on a different thing. You probably know that I've co-founded an AI company. It's the first thing since LinkedIn, Inflection, with Deep Minds, Mustafa Suleyman. And, you know, since you do product reviews, we thought we'd bring Pi into the conversation.

• 34:23Aria Finger?

  Okay, sure.

• 34:23Reid Hoffman?

  That's cool.

• 34:24Aria Finger?

  Yeah.

• 34:25Reid Hoffman?

  And so, you know, our goal is a, you know, kind, supportive companion, offering conversation, advice, concise information, but also something that helps everybody kind of navigate their life as a kind of a good, you know, kind of like friendly and helpful person that helps amplify and navigate their life. What I will do is ask Pi a question, we'll have Pi answer, and then we'll ask you to do a follow-up question, right? Does that work?

• 34:55Aria Finger?

  Sure.

• 34:55Reid Hoffman?

  Does that work? Okay. So, "Hi Pi, I like creating. How will AI help the most prolific creators make original and impressive content in the year 2028?"

• 35:07Speaker 03

  Hey there, that's a really cool question. I think AI will help creators in two main ways. First, AI will help with the grunt work, like creating 3D models, editing videos, and fine-tuning the sound. This will free up more time for the creator to focus on the creative aspects of their work. Second, I think AI will be able to generate interesting ideas that can inspire the creator and spark new ways of thinking. Together, these two trends could lead to a huge explosion of creative work in the future.

• 35:35Speaker 00

  So, my follow-up question would be, is there still a place for people who currently edit videos, or what was the other thing it said? Edit videos or do sound refinement?

• 35:47Aria Finger?

  Fine-tuning the sound.

• 35:47Speaker 00

  Fine-tuning the sound. Is there still a place for creators to do those things? To add quality on top of what AI will be able to do?

• 35:57Reid Hoffman?

  So my answer to that would be yes.

• 36:00Speaker 00

  I suspect it's a yes.

• 36:01Reid Hoffman?

  Yes.

• 36:01Speaker 00

  But I want to hear what AI would say.

• 36:03Reid Hoffman?

  Exactly.

• 36:04Speaker 00

  You know?

• 36:04Reid Hoffman?

  Exactly.

• 36:05Speaker 00

  All right, let's see what we got.

• 36:07Speaker 03

  I think there will definitely still be a place for human expertise in editing and fine tuning. AI will be able to automate many of the more basic tasks. But I believe there is something special about a human touch that will still be valued. Plus, an AI will not be able to think creatively and critically in the same way that a human can. So there will always be room for human editors, sound designers, and other creative professionals to add their own unique perspective and touch to any piece of media.

• 36:36Speaker 00

  That's a good answer.

• 36:38Speaker 00

  That's a really good answer. Yeah, I agree. And I hope that that's right. Because when you say that, it's sort of like, all right, in the future, yeah, your video is just going to chop itself up and edit itself and like remove the background noise and be good. And it's like, well, that's my job. So, so how does that work? So it is curious to see that, yes, that will be made easier and faster, maybe on a basic level. But then I'm going to come in with the human touch and create something genuinely new and creative on top of whatever AI was able to baseline cut for me, which is, that's what I want to hear.

• 37:08Reid Hoffman?

  Yeah, I think there's two parts of it. The reason why there's still always going to be kind of, for a long time, human creativity. One is, is, as we were talking about earlier, is the prompt directing. So like what you do, like what you get out is partially what you get in. And what you put in is the thing in that. And there'll be specialization. And then the second part of that is, is that there will be a tendency, somewhat obviated by the prompt directing, to the fine output, right? The safe output, whereas you want the creative original spark, that something somewhat different. And that's another place where the creatives will come in, whether it's editing, video, sound, et cetera, and still go, ah, this is the thing that makes it different from what's out there.

• 37:48Speaker 00

  Yeah. And even, especially because this was, whenever you talk about like, really interesting, deep creative works, like movies or something like that, I think a lot of people extrapolate forward super far where they're like, are there going to be AI-generated movies? And would an AI-generated movie even be good? And I think a good movie. Either tugs on your heartstrings or inspires you, or there's something about it that's human that really gets to you. And I, I, I wonder if there is room, again, for the collaboration, the absolute highest end form of the arts, where you can still improve the movie with AI, but also still have that human touch that makes it a good movie. Seems like it's a yes, hopefully.

• 38:29Aria Finger?

  I also think what you said about electric vehicles with, the brilliance is also going to be with the EVs talking to each other. When we interviewed Mustafa Suleyman about Pi, that was his hope too, that, you know, your personal assistant would talk to Reed's personal assistant. And so that's like another amplification. So it's like, you're talking to your personal assistant about editing, or about, you know, you're feeling sick today, so maybe your, you know, editor can take over. And then they're talking to, you know, your graphic designer's Pi, or whatever it might be. So there's like another level of amplification, if you like, the different Pies are talking to each other.

• 39:04Speaker 00

  Now, this is, this is where it gets interesting, because I've always thought, and this is, this has been true since like the early assistant days of Siri and Google Assistant. It seems like there's a very direct trade off between, like, privacy of information, all of the information you have to give it for it to be this useful, and how convenient and good it will be. So like, I have a Google account, my Google account knows everything about me, but my God, is it useful. Like, my calendar, my tasks, my inbox, everything, it's all there. I ask Google Assistant, what's up today? And it tells you, you should leave early for work today, because you live here, work is here, you want to arrive by this time for your calendar meeting, but there's more traffic than usual on the normal route you take. And it's just like, you know, everything about me, but that's really useful.

• 39:47Aria Finger?

  Yes.

• 39:48Speaker 00

  And I wonder, when you go that far out, like, okay, my assistant knows everything about me, yours knows everything about you, and they're talking to each other, does that get dicey? Like, is there ever a world where I don't want you or your assistant to know about my assistant? Like, that's, that's a really interesting future.

• 40:04Reid Hoffman?

  Well, there will be some errors, just like there's human errors, right? Because even if you had like a human assistant, who is delightful, occasionally, there's an error.

• 40:12Speaker 00

  Sure, sure.

• 40:13Reid Hoffman?

  But like, as long as you realize there's some errors, and it's learning, and it's, its entire, you know, raison d'etre, reason for being, is to be a good assistant to you and is learning, then I think that builds the trust and so on. It just, it has to come with a, it will not have You know, zero percent errors.

• 40:33Speaker 00

  Yeah, I think, I think a lot of the AI should be in its evolution compared to a good version of what a human can do. So like when you look at self-driving cars, I think it's really just about getting the error rate down. Like, how many people die for this many miles of driving? Get that down, it'll be a better version of human drivers if it can get to that metric. Okay, you have an assistant on your phone, what would a good assistant human be able to do? Can it do all those things? Can we get to that benchmark? That's a win for the AI tool. And you just kind of do that across the board. Find the good human, get better than the good human, then the tool makes sense.

• 41:08Reid Hoffman?

  The other thing about it is to use the advantages, like, don't try to, to make the AI better at the things it's really weak at. Sometimes you'll do that, but use it where it has an advantage. So think about like having a tutor on your smartphone. Infinitely patient. Unlike human beings who occasionally like, ah, I've explained this to you five times.

• 41:26Speaker 00

  Sure.

• 41:26Reid Hoffman?

  I'm really tired of it.

• 41:27Speaker 00

  Yeah.

• 41:28Reid Hoffman?

  You know, you go, no, no, let's try again.

• 41:30Speaker 00

  Superpower. It has a superpower.

• 41:31Reid Hoffman?

  It has a superpower. And, and to deploy those superpowers in ways that amplify humanity. So I, it's what you said and this, I think.

• 41:39Speaker 00

  Yeah, yeah, that makes sense. There's going to be ways, especially because why, why aim for the good human when there are reasons where it can be way, way, way, way better than the good human?

• 41:46Reid Hoffman?

  Yes.

• 41:47Speaker 00

  But that, I think, is like further down the line, especially if I go back to the car example, it's like, there's no way every human on the road could talk to every other human on the road and figure out where to drive based on the guy six cars behind you and what he's going to do. Like, that's, that's the superpower of AI.

• 42:01Reid Hoffman?

  Yes, exactly.

• 42:02Speaker 00

  For sure.

• 42:03Aria Finger?

  And I think to your point about comparing it to the good human, sometimes we expect technology to be fallible. And to your point, if we could just have cars that were twice as good as humans, and so we had half the deaths.

• 42:14Speaker 00

  Yeah.

• 42:15Aria Finger?

  That's amazing. Like, that's a trade-off we want to make.

• 42:17Speaker 00

  Yep.

• 42:17Aria Finger?

  And I think in the privacy sphere, people don't want to give up privacy when it's so XYZ company can make an extra buck. But it's so that you can get to your job on time, like, that's what you're, you know, willing to give up.

• 42:29Speaker 00

  Yeah, yeah, I think a lot of people really do make that very direct trade-off. And, and they're unhappy when they feel like they've given up too much and don't get the convenience back.

• 42:38Aria Finger?

  Right.

• 42:38Speaker 00

  So, yeah, when I use Google Assistant, I'm gonna get as much out of it as possible because I know it has all this information. And I think that's probably true about a lot of people. They, they understand that there is some sort of a trade-off.

• 42:56Aria Finger?

  Let's move on to our rapid fire questions.

• 42:59Reid Hoffman?

  Is there a movie, song, or book that fills you with optimism for the future?

• 43:06Speaker 00

  I have maybe an interesting answer. I'll give two. One is "Ongoing Thing" by Vain Sil. It's a song that I've used in my intros in videos in the past for years. But when you actually listen to the lyrics about, it's sort of a motivating song, I guess, to me. But I think literally from cover to cover, all the lyrics speak to me and I like, I like the lyrics of it a lot. And then the other one is, I want to get the name right. It's "Prima Donna" by Skogzra. And I don't think I'm saying that right, but it doesn't have any words, but the arc of the song feels exactly like the arc of life. And it's really flowery and optimistic and great at the end. And I think that's pretty cool.

• 43:45Reid Hoffman?

  I'm gonna have to listen to that.

• 43:46Aria Finger?

  That's awesome. That's gonna be my pump up song.

• 43:48Speaker 00

  Yeah.

• 43:49Aria Finger?

  So what is, and this, this can be silly or serious or professional. What is a question that you wish people asked you more often?

• 44:00Speaker 00

  I think what's really interesting about the two worlds that I live in right now, which is ultimate frisbee.

• 44:07Aria Finger?

  Okay, that's what I was gonna say, ultimate frisbee and tech.

• 44:09Speaker 00

  And tech is there's not a lot of overlap other than like, so how about that new iPhone? Like it's pretty basic. So I think it would be more interesting to find, to have questions asked about that overlap. I think it would be fun, whether it's just like AI and health training or, or injury prevention, or just getting better at the sport. I think it'd be cool to, to get asked about. And to be also probably to ask others about like that overlap. I think that's really interesting.

• 44:42Reid Hoffman?

  So next question, where do you see progress or momentum outside of your industry that inspires you?

• 44:50Speaker 00

  I'm just gonna say ultimate again, just because, like I said, the the growth of the sport, not just in like how many people play, but in high level strategy, coaching, player development across the board in every division, every age category, every five years, it's drastically better. And that's really, it's tough to say about a lot of sports because I think most sports people are familiar with are so established that like high school basketball today looks very similar to high school basketball 10 years ago. High school ultimate today looks nothing like high school ultimate 10 years ago. And especially the earlier footage you see, the more of a difference it's, it's you can notice. So I'm, I'm very inspired by the progress of this, this little sport that I get to play on the weekends. Yeah.

• 45:36Aria Finger?

  That's awesome. And you're, you're not scared about the other folks coming for you.

• 45:40Speaker 00

  I know my prime is gonna be a certain set length and then I'll be watching.

• 45:45Aria Finger?

  You can be coaching next.

• 45:46Speaker 00

  Exactly.

• 45:47Aria Finger?

  Um, so you have such a positive outlook for the future. Can you leave us with a final thought on what you think is possible if everything breaks humanity's way in the next 15 years? Like where could we go? What's the first step to get there?

• 46:01Speaker 00

  Yeah, I think I love the thought of, of the best technology being deployed on the biggest problems for humanity. And I think you see a little bit of that here and there, little sparks of that, oh, a little, oh, a new rocket launch. Oh, oh, it landed itself. Okay, that's good. Maybe we'll get to Mars someday. Like little stuff like that is very inspiring.

• 46:21Speaker 00

  Also connecting the entire planet has been something that's been really fun to watch. Whether it's just getting more communities online and then the entire world and all of humanity can talk to all of humanity and we all learn from each other.

• 46:34Speaker 00

  So I think just in general, to zoom all the way out.

• 46:37Aria Finger?

  Yeah.

• 46:37Speaker 00

  Technology at the highest level being deployed to solve humanity's biggest problems.

• 46:42Reid Hoffman?

  Amen.

• 46:43Aria Finger?

  Here, here. Thank you so much for being here. Appreciate it.

• 46:45Speaker 00

  Thanks for joining me. Thanks for having me.

• 46:47Reid Hoffman?

  This was great. Thank you. In an honor of our season two opening guest, we tip our hat to Marquez and his use of AI in his outros. Here goes. As creators, can we catch the current of computational capabilities? Crafting content, we contemplate how AI can carry our connection closer.

• 47:12Reid Hoffman?

  Possible is produced by Wondra Media Network, hosted by Ari Finger and me, Reid Hoffman. Our showrunner is Sean Young. Possible is produced by Edie Allard, Sarah Shleid, and Paloma Moreno Jimenez. Jenny Kaplan is our executive producer and editor.

• 47:29Aria Finger?

  Special thanks to Surya Yalabonchili, Saida Sapieva, Ian Alice, Greg Diato, and Ben Relis. A special thanks to Ben Davis and Evan Crask at WME, the MKBHD team for their studio and hospitality, and Little Monster Media Company.