Sam and Shaan open with dinner with Hasan Minhaj and a riff on learning to be funny, then Shaan lays out four AI business ideas — synthetic voice, AI-generated student essays, patterns and prints, and AI stock photos — framing them as opportunities to disrupt existing creative marketplaces. The episode pivots to Shaan’s thesis that a Twitter-native media empire has yet to be built, and then closes with an extended conversation about the Teal Fellowship, Vitalik Buterin, Emmett Shear’s pattern of thinking, and what separates truly high-horsepower founders from everyone else.
Speakers: Sam Parr (host), Shaan Puri (host)
Cold Open — AI Is the Multi-Billion Dollar Window [00:00:00]
Shaan: I think this is one of the multi-billion dollar trends that you could be on right now. And I don’t say that for hyperbole — like, literally, this is the window. The tech is just now finally good enough. It’s not quite there, but you need to start now, and you can disrupt all of these marketplaces.
Dinner with Hasan Minhaj [00:00:20]
Sam: All right, I went out with Hasan last night for dinner. Can I tell you about it really quick?
Shaan: Do it.
Sam: It was awesome. So he admires you a lot. Did you know that?
Shaan: I admire him a lot.
Sam: How do you pronounce his name — Minhaj or Minaj?
Shaan: Minhaj, right? I think his name is actually pronounced Hasan Minhaj, and everybody calls him Hasan Minaj because it sounds cool — like Nicki Minaj. So whatever. I don’t know how strict he is about that.
Sam: So basically the story is: Shaan interviewed him last year, I thought I saw him on the street, ran up and said hey, and it actually wasn’t him. I looked like an idiot, tweeted about it, and he texted me — I guess he got my number from you — and we went and hung out.
So this guy’s a pretty big-deal comedian. He is so smart. He’s so much smarter than I am — that’s what I learned. And number two, I was so intimidated being around him because I didn’t know if I could make him laugh. And I really didn’t. I failed. I could not make him laugh.
He showed me these books he’s been reading, and I wanted to tweet this out but I was too embarrassed. Maybe you can tell me — are there good books on how to be funny or clever? I felt like such a nerd. I was like, I’m going to tweet this out, but I’m going to look like the biggest virgin on earth. Like, how do I be clever in a conversation?
Shaan: I don’t know, dude. I feel like it’s a skill that’s only passed down by hand. It’s trial by fire. You have to learn this when you’re 12 — get good at busting people’s balls, saying the funny thing in class.
Theo Vaughn has this hilarious story. He’s like, “I was freshman year in college in Louisiana, and my professor said to the class, look to your left, look to your right — statistics say one of the three of you is going to end up a child whatever.” He goes, “I stood up and said, not it.” And it killed. He’s like, that was the moment.
I think there’s truth to it — you get rewarded for being funny early on and you’re like, more of that, how do I get more laughs? And then you just keep doing it.
Sam: Talent’s real, but I think, like anything, some people will never be able to dunk, but you can definitely improve. You can learn how to dribble better than before. There’s definitely a skill there.
Shaan: There’s a YouTube video you should watch. I’ve gone down this rabbit hole too after I met him. He would tell me a story and I’m so into it, eating out the palm of his hand. Then he’d ask me a question mid-story and I’d just be nodding, still listening like it’s a TV show. He was telling a story — “everybody’s got that kid growing up” — and I was so enthralled I forgot my own name. He’s like, “who was it for you?” and I thought about it and went, “yes” — and then he was like, “who?” And I realized I said yes to having a name when I couldn’t even remember who I am because I’m so into this guy.
The video he gave me is called “Writing a Joke with Mark Normand.” It’s like an hour-long podcast. Mark Normand’s a great joke writer, and he kind of breaks down the mechanics of what makes something funny. Not like “follow these three easy steps and you too could be funny,” but more like — okay, if I consistently practice these mechanics, I can take something that’s not so funny and make it funnier repeatedly. There’s the storytelling, the setups, tagging jokes — there’s more to learn. But that video’s pretty good.
Sam: There’s also this YouTube channel called Charisma on Command. Beautiful name. They’ve been doing it forever.
Shaan: Yeah, I was watching those videos like five or seven years ago. They’re cool. It could be a little lame because it leans nerdy-guy-meets-women energy, but I remember their name, they taught a whole course, they had all this stuff. That’s a beautiful name — Charisma on Command.
Sam: I’ll watch that Mark Normand thing. Anyway, I hung out with this dude last night. It was dope. He paid for dinner.
Shaan: Did you do the thing — try to grab the check?
Sam: Not me. But what was funny — we went to Gramercy Tavern. He likes that place. And I was like, “dude, you come here a lot?” He goes, “yeah, I love it.” I go, “I’m not even going to look at a menu — whatever you want to order for me, just do it.” That’s how I love going to restaurants. I do that all the time. I’m like, I don’t care, I eat meat and vegetables, get whatever.
Shaan: The best is then after they order, you complain. That’s the true power move.
Sam: You know, one thing people liked from a couple months ago — I got a bunch of DMs. They were like, “dude, I love the homie move.” And I was like, I don’t remember what you’re talking about. He’s like, “remember — it’s one of the top ten things you’ve shared on the pod.” It was when you say, “guess how many” — obviously it’s something impressive — and I guess low so you can have your punch line. Like, I say ten, and you’re like, a hundred million. People loved that.
Shaan: I saw the comments on YouTube. People dig that.
Bringing Bigger Ideas to the Pod [00:09:00]
Sam: So I made a little vow to myself to bring more specific ideas and stories to the pod. Because I personally love the free-flowing, wherever-the-conversation-goes style, but I want to make sure we have the entree, not just tapas. Not just appetizers — I want the entrees. So let’s jump into a couple of ideas. I’ve got a few, or I’ve got one big one.
Shaan: I’ll go one and then you go one.
AI Business Idea #1 — Synthetic Voice (Unreal Speech) [00:09:30]
Sam: So we’ve talked a little about DALL-E and how much AI is advancing. I don’t know if it’s just my TikTok feed, but every day I see a new incredible thing AI can do.
Shaan: You have to explain what DALL-E is.
Sam: Oh — DALL-E is an artificial intelligence program. DALL-E 2, I think, is the current name. You basically type in any word and it generates images. You could say “Sam Parr fishing” and it will generate images that look like they’re either hand-drawn or like a stock photo of Sam Parr fishing — even though no such image exists. It will just create the image.
So it flips the whole idea of creating art or taking photos. Like, no — computers and robots can now do almost as good as, if not in some cases better than, what humans produce.
You also see deep fakes. There are videos where you’ll see the Mona Lisa, and someone pushes a button, and she starts turning her head and rapping like Notorious B.I.G. The mouth syncs with the lyrics. It looks real. They took an image and turned it into a video using AI — AI basically says, I know where the eyes and lips are, I’ve been trained on what talking looks like, what singing looks like, and I can turn any image into that.
They could take Barack Obama and make him sound and look like he’s saying something completely racist. That happened with a Ukraine thing — they deepfaked the Russian president. The technology can make you look and sound like you said anything.
Shaan: I heard a podcast about that. Someone made one of these to show how amazing it is, and to show the downfall — they made Obama say something he definitely didn’t say. It seems exactly real. This is where it’s going to get dangerous.
Sam: There’s another one — you can draw a very basic smiley face on your iPad in this app called Procreate. Just eyes, nose, mouth. Then you switch the brush to the AI paintbrush and just shade it in, moving your hand like a fool, and it colors in as if you painted it in fine detail — shadows, colors, everything. It already knows what high-quality art looks like and just turns your crappy art into high-quality art. Every day there’s these amazing magic tricks. They’re not foolproof yet, but the demo is getting ridiculously impressive.
I actually used it to come up with a logo for a business I’m thinking about starting. It was awesome.
Shaan: I do that all the time. If you just Google “AI logo generator” there are like four websites. You say, “it’s a legal startup, the mood is serious but still sophisticated,” and it spits out an infinite number of logo variations. That’s how they make money — print it on business cards.
Sam: Normally I go to 99Designs and have like 30 graphic designers make up a landing page. Then I’m like, I like details from number one, number eight, number fourteen — combine them. That way you can also just be brutally honest and say “this sucks, do this instead.” Whereas with one designer you have to be polite. With DALL-E, I’m excited to use it that way.
So let me give you a couple of ideas I’ve had. These can be very big ideas. The problem — and I’ll say the problem too — is that I think it’s so easy to do that you’re going to have a bunch of competition. You’ll have to find a way to be defensive. But here we go.
Shaan: Okay, four ideas you can do with AI right now.
Sam: Number one: fake speech — synthetic voice. I’m investing in a company called Unreal Speech. This guy sends me deepfake audio of Gary Vee and Jordan Peterson reading the Milk Road newsletter. Let me just play one.
[plays audio clip]
Shaan: Dude, it sounds right. That’s like Jordan Peterson.
Sam: Here’s Gary Vee.
Shaan: That’s amazing. Is this available to everyone?
Sam: It’s brand new. And he was like, “Hey Shaan, you want to never do an ad read again? You have hundreds of hours of audio of you and Sam’s voice on the podcast. Just say the word and I can make it so you’ll never have to do an ad read again.” And like, the Milk Road — it’ll auto-add companion audio. If somebody wants to listen to the news on the go, it’ll do it in your voice. He said, “I can just do that now.”
The way they’re doing it is cheaper, faster, and more accurate than the other models on Amazon or wherever. Dramatically cheaper — and the cost just keeps falling every year. There are also techniques to make the models smaller.
Think about language translation. There’s a world where we do this podcast and Unreal Speech translates it into another language in our voices with our tones — just speaking in Spanish, French, whatever. Our podcast grows without us doing extra work.
I think that’s a great business. Okay — next one.
AI Business Idea #2 — Student Essay Generation [00:18:00]
Sam: Student essays. I told you about that guy who made WritelikeSam.com — or WritelikeShawn.com. I mentioned it in my email and it blew up. Like, caused him a thousand-dollar server bill. He was like, oh, the side project is now a financial burden.
But basically, this guy trained AI to write like I do, and it was pretty damn good. Now, this has nothing to do with me specifically — you could take any corpus of text data, in this case just my Twitter feed, train it to write like that. The big thing is: you get assigned essays in school. Some people would pay others to write them, some would find one online and plagiarize it. Well, this is going to screw that whole system up.
I’m going to go on there and say, “write an essay about World War II and Germany’s role, at the sophistication of a tenth grader,” and it’s going to spit out a full essay. A thousand versions. I pick the one I like and edit it in maybe five minutes.
Shaan: You remember Turnitin.com?
Sam: Yes! So I used to use a website that gave you free essays, all the time. And you’d rewrite it a little bit. Then they came up with Turnitin.com — you’d submit your paper and also submit an electronic copy, and it would find plagiarism. I have no idea how sophisticated it actually was, but if you just Google a phrase in quotation marks you can find where it’s copied from. The teacher doesn’t want to Google everyone’s paper, so you just trusted the tool.
Shaan: Totally. And I think you’re going to have creation of AI-generated content, and then you’re going to have the fact-checking and detection businesses. Both are going to boom. You’re going to have things that take a video and say, is this a deepfake? You’re going to have things that say, is this auto-generated content or did a human write this? It’s like the opposite of CAPTCHA. Instead of “prove you’re a human,” it’s “prove this isn’t a robot.”
AI Business Idea #3 — Patterns and Prints Marketplace [00:22:00]
Sam: Here’s another area that’s going to get totally disrupted — print and patterns. People talk about using AI for stock photos, for auto-generated music. That’s all going to happen too. But here’s what I think is more low-key.
When Patrick came on, he was talking about quilt businesses. There’s a huge market for pattern-making — for clothing, blankets, curtains, whatever. Patterns and prints are everywhere in apparel and fashion. If you want to make clothing, you use websites that have large libraries of patterns and prints. Today, somebody hand-draws each one. “This floral pattern, this repeating pattern of fire trucks for a kids’ thing.” That person gets paid when you buy the pattern — let’s say $100, and they get that as their royalty stream.
All of this is going to go AI. Somebody can compete with these marketplaces, and the sites that do this are worth hundreds of millions of dollars. We talked about one of them — Spoonflower, I think? Got acquired for something like $400 million.
What that’s going to turn into: I’m going to go on there and type “galaxy-themed repeating pattern,” and it’s going to generate galaxy-themed patterns. They’ll be able to charge one-tenth of what the other sites do because there’s no human involved — and they’ll keep all of that revenue. They become the supply side of the market.
This is like what Uber is trying to do with self-driving cars. We can make rides way cheaper and the system more efficient — we just have to get to self-driving. Tesla with the robotaxi. The big idea in transportation is getting rid of the driver. The same thing is going to happen with artists. We’re going to get rid of the human doing the photography, videography, music, pattern-making, voiceover work, essay writing, logo design.
Every single one of those — 99Designs, Spoonflower, the stock image sites — they’re either going to have to release this and kill their own supply side, or someone is going to come and eat their lunch.
This is one of the multi-billion dollar trends you could be on right now. I don’t say that for hyperbole. Literally, this is the window. The tech is just now finally good enough — it’s not quite there, but you need to start now.
AI Business Idea #4 — AI Stock Photos and Fake Faces [00:28:00]
Shaan: Do you ever recognize stock image models on different websites?
Sam: Yes! There’s this guy from The Bachelor who’s a stock image model — he’s in like a bunch of random kids’ Halloween costumes. You see him on a hot sauce commercial. He’s this generic smiling guy they just reuse everywhere.
I saw the lady from the Google Analytics page in public once — went up and got a picture with her. And there’s another lady who’s on Google Analytics, kind of racially ambiguous, curly hair, could be Italian, could be Black, could be Jewish. She just represents everyone. I see this woman everywhere on all different websites.
I get sick of using the same pictures though, so for some of my projects I’ve been using DALL-E and other services — they’ll give you fake people. I have a couple websites with real testimonials, but I don’t want to use the people’s faces because I didn’t exactly ask for permission. It was like an email they wrote me. So I Google “AI sample faces” and you can get all types — make this person this race, make them happy, make them serious, change it up. I’ve been using those for sample images.
When we were running The Hustle, every once in a while we would use an image that was supposed to be for commercial use but it was categorized wrong, and we’d get one of those takedown notices — multiple times we had to pay $5,000. You’re just like, these guys are going to cause an issue, we just gotta pay them. So I’ve been using AI faces to avoid that.
Shaan: It’s going to be the same with stock images. Designers always use these libraries of icons and illustrations — websites like the Noun Project. What DALL-E did was basically read all of the text on the internet and then use it to create new stuff.
Some people are like, “that’s messed up.” GitHub is doing this too. GitHub basically indexed all of the code that was on GitHub, fed it into machine learning, and now they have GitHub Copilot — I think it’s like $100 a month. You’re typing code and it autocompletes the line. A $250,000-a-year software developer is now more efficient and effective, fewer bugs, fewer mistakes. And a less talented person can create these functions.
GitHub Copilot is going to generate hundreds of millions of dollars for GitHub — it could be a multi-billion dollar product line.
Sam: Some people are mad about it because they’re like, you read my code, used my code as training data, I get nothing from this, and you’re going to put me out of a job with my own data. That’s the same thing that’s going to happen with all these pattern sites.
Shaan: Ben found Spoonflower — that’s what it’s called. All you have to do is go and scrape and crawl Spoonflower, ingest all of the patterns with all of the tags, and then train a dataset so that it knows: this is what a pattern with X, Y, and Z keywords looks like. Now give me that keyword again and I can generate 500 new patterns from scratch. Spoonflower is basically going to create its own destroyer.
The data from stuff that’s already on the internet is what’s going to destroy those services. It’s going to be like when chess programs got really good really fast — all of a sudden a human can’t compete. This is going to enslave the gang of graphic design. It’s going to go to your village, it’s going to take everything. That’s my prediction.
Ben, go to this website — can you share this? Sam, I think you’ve used it before. It’s called “This Person Does Not Exist.”
Sam: Yes! That’s what I used.
Shaan: Look at this — this guy just looks like a basic dude. This is an AI composite of a face. You can use it anywhere on your website, copyright free, because nobody’s going to say “hey, that’s me.” It’s such a useful, funny website. And at the bottom right it tells you exactly how the image was created — “imagined by a GAN,” which is a type of machine learning. You can refresh it for new people.
You can also do cats, horses, chemicals — “this cat does not exist.” Click and it’ll AI-generate a cat.
Sam: I usually just take screenshots of those and use them on different websites.
Shaan: So those are three ideas: student essay generation, Unreal Speech for voiceovers and turning your voice into a programmable asset, and patterns and prints. And that’s not even counting the obvious ones — stocks, photos, all the stuff I mentioned. This is great.
Shaan’s Thesis: Building a Twitter Media Empire [00:34:00]
Sam: A million dollars isn’t cool. You know what’s cool? A billion dollars. All right, I have something for you. Have you heard of John Steinberg?
Shaan: Is he the guy who did Cheddar?
Sam: Yeah, listen to this story. So John Steinberg was the CEO of BuzzFeed for a long time. When he joined, it was like 20 people — he wasn’t quite a co-founder but he worked his way up to CEO, was the ad sales guy, was basically like a co-founder. He’s incredibly aggressive.
He started this thing years ago — six, eight years ago — called Cheddar. The shtick was that it was going to be a live TV station for Facebook Live. When he launched it, he messaged me and said, “Have we met yet?” Just such an alpha move. I was like, “Dude, it says Sam…” He’s bad with names and faces. He goes, “Have we met yet? We should meet.” I said okay, cool. He goes, “I’m in San Francisco. What’s your office? I’ll come tomorrow.” That’s what he said.
So I hung out with him — nice guy, incredibly aggressive. He basically told me: BuzzFeed was great but he wanted an exit, and it didn’t happen. He goes, “Here’s what’s going to happen: I’m going to raise a little bit of money, like $30 million. It’s going to do pretty good. And then a big cable company is going to buy us, because it’s going to be cheaper for them to buy us than to go make one of these 24-hour news networks on their own. They’re going to buy us for $200 or $300 million.”
At the time — and even when they got bought — no one watched it. You’d go to the Facebook Live and there’d be like 20 people. I remember thinking, this guy’s crazy. This is not going to work. No one wants this. This is a bad business.
And then in year two or three, he told me: “We’re going to get acquired for a couple hundred million bucks.” In year three, I think he sold the business for $250 million. He 100 percent called his shot. He completely disappeared from the conference circuit after that. Just nailed it and walked away.
I remember talking to other guys in the media industry and they all said, yeah, he told me the exact same thing.
Shaan: This guy is a shark. Not in a bad way — just pure intensity and energy.
Sam: He did it. And another woman, Rebecca — we’re friends with her — I asked her how she got popular on YouTube. She said, “I quit my job, worked 50 hours a week, got up at nine, worked till seven, studied the data, and treated it like a business. And now I make seven or eight figures.” It worked because she was talented and worked hard and treated it like a job.
What I think is going to happen — and this is a little circle-jerky because you and I both like Twitter — is that very few people take Twitter seriously enough to do that. You know, five or eight years ago we’d say, “MrBeast, you’re going to dedicate your life to YouTube? That’s the stupidest thing ever.” Now it’s not funny. We all know that’s real. You can become a YouTuber and make hundreds of millions of dollars a year.
No one takes that seriously about Twitter. But I think in the next two or three years we might see a few empires built on top of Twitter. It sounds silly now — just these Twitter threads of guys summarizing Wikipedia. And they’re kind of right. But I think there’s going to be someone on Twitter who treats it like Cheddar did, like MrBeast did. Takes it really seriously. There’s a huge opening right now where very few people are doing that.
Shaan: This is actually one of the more interesting things you’ve said. I feel like this is different from your usual mode. You usually find an interesting diamond in the rough and break it down — you’re amazing at that. But this is like you see a trend before it’s obvious, which is different.
I kind of agree with you. Every big social media platform has native stars. Podcasting has Joe Rogan, Call Her Daddy, Tim Ferriss. YouTube has its YouTubers. TikTok created TikTok stars — people who weren’t famous anywhere and got famous there. Instagram made Dan Bilzerian famous. By the end, everybody does multiple platforms, but you’re really good at one medium first.
So who is the Twitter-native star? Sahil Bloom is kind of that. Our friend Trung, who used to work for me, is kind of like that. But we’re still talking 500,000 followers. Nobody has really pushed it to 10 or 20 million. Nobody’s built a proper empire as opposed to a niche.
Sam: And the difference is — basically everything you mentioned is visual-based, and Twitter is copy-based, text-based. So I think there are two routes: you figure out how to add visuals to your thing (pictures do well on Twitter, videos not so much), or you just get really good at writing.
And I was like, well, writing’s at a disadvantage. But then I’m like, wait, I’m an idiot. You know who Danielle Steele is? You know who James Patterson is? Danielle Steele has sold like a hundred million-plus books. In San Francisco, she’s got the biggest mansion on Billionaires’ Row. James Patterson basically writes like 50 books a year — or smacks his name on them — and he’s probably a billionaire.
The power of the written word is powerful enough to get people to buy all these books. Authors still crush it. So I still think it’s possible.
Shaan: I actually think there’s a huge opening. Very few people are taking it seriously. And it’s incredibly easy to capitalize on this window right now.
How to Build a Twitter Empire: The Playbook [00:42:00]
Sam: What would you actually do, though? The guys we talked about — Sahil basically does mental model, intellectual content. Either generic life advice or “here are 30 cognitive biases you want to be aware of.” I don’t know why those do so well, but they do incredibly well. That’s not my cup of tea, but it’s definitely other people’s.
Here’s how you cross the chasm — how you go from Twitter to actually big. Because right now Twitter feels like a one-to-one or one-to-ten. Me and my ten friends consuming this, joking about it. We’re not talking about it on the train or at family dinner.
So two things. Number one: you need to hang out with mainstream celebrities, take pictures, show it. It’s like, oh, this person is bigger than just my little audience — Arnold Schwarzenegger took a photo with this person.
Number two: you need to get off Twitter and into real life. In the early 2010s, when the Caspers of the world were coming out, subway billboards in Manhattan were underpriced. Companies like Casper went and bought those billboards, and people were like, why would you do that? We can track everything online. And they said, well, because if we get off the internet it’s going to seem like more than just an online thing. It’ll feel tangible, legitimate — this ad next to a Ford ad, next to something real.
So you do that by having real-life events, publishing a book, getting a cameo on a TV show. Get everyone in on the joke — it’s like, hey everyone, they asked me to be a background extra on Friends, look you can see my head. Real-life stuff that shows you’re bigger than just this one-to-ten thing. It creates FOMO, it creates social proof, and you start becoming culturally relevant.
Shaan: I think people will believe about you what you believe about yourself. And people will believe about you what they think others believe about you. That’s why YouTube works — people are like taking a picture but the background is the real message. “I’m sitting in this car, at this place, with these people.” Everything said is in the background.
Or Casey Neistat walking through New York with kids coming up to him — “hey, what’s up, you’re on the vlog.” It’s like: oh wow, in New York Casey gets recognized everywhere. He’s more than just what I see on my screen.
Celebrities do this — they tip off the paparazzi to come, pretend they’re trying not to be seen, but it’s actually a deal. I know a buddy who created an online e-commerce brand, opened their first store in New York, not expecting big sales, but just showing “we opened our first store in New York” — a small place, but it makes the brand feel more trustworthy, more established, more legit. He had 15 friends show up for a ribbon-cutting ceremony, pretending to be media — “hey, question for you!” Zoomed out it was 10 people in a small storefront, but from the right angle it looked like a mob was at the store. Perception creates some version of reality.
Sam: I don’t think that alone builds the empire. I think you have to do a few things: one, be in it for the next seven years. MrBeast didn’t start big. It takes five to seven years to build something truly massive. Two, I would stop trying to jump off Twitter and go to other places. Stay. Master this medium. The third is to do the stuff that builds fame — find your lane, your niche. Controversy and collaboration are underrated.
Nobody does controversial stuff on Twitter right now, but it should be like — oh my god, I called into this show and got him to say something crazy, check this out.
There’s a guy who’s doing this — Zach Weinberg, I think. Literally just consistently goes after crypto people, challenges them to debate, dunks on them in public. That’s his whole shtick. He’s getting so much fame because there’s an audience that likes to watch a fight and has anti-crypto feelings. He is their champion.
Shaan: That’s exactly how Ben Shapiro got famous, by the way. And then you have people like Keith — I don’t know his last name — who’ll just be a dick to people. Someone comes in good faith and he’s like, “I don’t share my perspective with losers.” Or, “The data is out there, go read a book.” One time he said, “I’ve forgotten more about startups than you’ll ever know.” Or, “Call me after your sixth unicorn.” Obnoxiously dunking on people. And it’s hilarious. Some people hate him, some people love him, and he uses that controversy to grow.
Sam: The main thing is: find your angle, your hook, your niche. Like Casey Neistat and the daily vlog. Whatever that is for you on Twitter — do it for five to seven years and have faith. Just like it wasn’t obvious that being a YouTuber was a big deal back then, it’s not obvious you can do this on Twitter either. But I think you can.
Anyway, that’s my pitch. It could be cool.
The Teal Fellowship — What It Actually Produced [00:52:00]
Sam: All right, can I tell you about something that I kind of knew about but has recently blown my mind?
Shaan: Yeah.
Sam: The Teal Fellowship. Do you know what it is, Ben?
Ben: Just my impression — it’s for essentially high schoolers who are super geniuses, and Peter Thiel gives you money to not go to college and do something cool instead.
Sam: Kind of. I think it’s actually for people in college to leave college for a year — $100K to basically leave college for a year and work on something interesting. And ben, is it successful or unsuccessful?
Ben: I know that Vitalik did it. The guy who did Ethereum.
Sam: Exactly — Vitalik. By the way, I still haven’t told my Vitalik story on air. I was in Amsterdam, I saw him, I couldn’t think of what to say, and he’s walking right past me. I just turned and went, “You’re Vitalik.” He didn’t look or say anything. He just kept walking. But it was him.
Shaan: Did he have security?
Sam: No, he was walking alone in Amsterdam at 8am. Just us.
Shaan: I really hope he’s got security so good he just can’t see it.
Sam: The Teal Fellowship is amazing. Peter Thiel — original co-founder of PayPal, first investor in Facebook, put $500K in early which became hundreds of millions — also invested in SpaceX and other businesses. He comes out and basically says: college education is a bubble. It hides under this banner of teaching you something, but really what you’re getting is an insurance policy — “I got this degree so I won’t fall through the cracks of society.” And you’re getting a status symbol, like a luxury product. Your Harvard degree is like your Louis Vuitton bag.
He says: costs keep going up and up, but the salaries you make don’t go up. It’s an overpriced bubble, like the housing bubble. So he says: I’m going to offer $100,000 if you drop out of college and go work on something interesting.
A bunch of people criticized him — “you’re telling kids not to stay in school, you’re like a vape.” And some people were just like, all right, you’re contrarian, that’s cute, but who knows if this works.
It worked. Let me tell you what’s come out of it.
Ethereum — $170 billion market cap. Vitalik is probably this generation’s version of Facebook. That alone.
Figma — somewhere between $5 and $10 billion, taking on Adobe in the design space. Dylan, the founder, was a Teal fellow.
OYO Rooms — India, Airbnb-ish hospitality, multi-billion dollar company.
Luminar Technologies — the LiDAR company for self-driving cars. This guy went public via SPAC, I think he’s 21 or 22 years old, worth like a billion dollars. One of the world’s youngest billionaires, if not the youngest.
Polkadot — another crypto protocol, valued at like $7 billion.
So just those — I just told you basically $200 billion worth of business value. And then there are dozens more in the $100 million range. I invested in one that recently raised another round — Owner.com. The founder Adam was a Teal fellow. That could be a unicorn.
The reason I got reminded of this was because I invest in Italic, and Jeremy Cai is the founder, and he was like, “Oh yeah, I was a Teal fellow.” I was like, how did we first meet? He’s like, “Oh, you gave a talk and I thought it was awesome, so that’s why we kept in touch.” I was like, dude, was Vitalik in the room? He’s like, “I don’t think he was at that event, but dude, he used to crash on our couch.” And I was like, what was he like? He goes, “He’s genuinely the weirdest person I’ve ever met. Nothing like anyone I’ve ever seen. Just a different cat.”
Shaan: I like that you’re workshopping “cat.”
Sam: Yeah, I take it out of the garage twice a week, see if people flinch. I tried “lit” and had to put that one back in storage. Got rid of it on eBay.
Shaan: Yeah, “lit” is not coming back. But “cat” — I’d stick with that.
Sam: It is crazy, man. The Teal Fellowship. I thought at first it was kind of stupid — like, these 18-year-old nerds aren’t going to do anything interesting. Obviously I was wrong. When I rethought my opinion, I was like, wow. High IQ just matters. These guys have more horsepower. They’re higher functioning. Your oven burns hotter. It actually truly makes a difference.
Emmett Shear and High-Horsepower Founders [01:02:00]
Shaan: When I was at Twitch I had opinions from the outside — what are the people at big tech companies like? How hard do they work? How smart are they? Do they have processes that are way better than the messy stuff I was doing at my startup?
For the most part, I walked out a little cockier than I came in, and I came in pretty cocky. Like, okay, they don’t have some magic process, the people aren’t that much better — except like two or three people at the company. One of which is Emmett, who’s the founder.
We’d sit down, data science would be presenting some six-page memo on trends, and I’m just trying to stop yawning — I’m like, why do I keep yawning? I’m just trying to understand these charts. And then you’d hear a click. Someone’s pen goes down. Like when someone in a math test finishes early. And it would be Emmett. He’d be playing Hearthstone on his phone.
Then they’re like, all right, we’re ready to discuss. And he would be like, “So on page three, this chart — the axis is a little wrong, because if this were true, then this other chart on page seven would show this.” And I’m thinking: how is your brain processing this much information this quickly, and always getting to the harder issue?
It’s like we all spent our time and energy trying to parse out what’s important. He immediately got to what was important, had the conclusion, had the follow-up question, and also identified why the paper wasn’t very good in the first place. Once I saw that, I was like — oh, okay. There are just some cars that have more horsepower. His oven burns way hotter than mine. Good to know.
I always say: I like to see what level 12 looks like. I want to know what volume 12 even is. Feel my own chest vibrating with the bass. I don’t want to be somebody who never sees it.
Sam: What was he like in real life? Have you ever seen some of these intense people outside of work? I told you about one friend who literally only owns two T-shirts, one pair of underwear, one pair of socks, and his laptop. That’s it. Worth hundreds of millions of dollars.
And I know a bunch of these guys — for hobbies, they’re all in. If they’re into optimizing their body, they’ve got every gadget imaginable. My friend Jack, who’s in supplements — he’s got literally $30,000 worth of supplements in a closet. He just tests everything.
Shaan: Before I tell you about Emmett, I just read this thing about Sergey Brin. The co-founder of Google, in the news because of his wife and Elon. Someone asked what Sergey is doing nowadays — I don’t think he’s at Google anymore — and it said over the last year or two he’s become very into personal fitness. He’s been trying to master several Olympic sports.
Sam: What Olympic sport?
Shaan: I don’t know. Ping pong? Curling? And it also said he’s currently writing a physics textbook.
Sam: Of course he is.
Shaan: Right? Like, “yeah, I’m writing a physics textbook.” Come on.
So anyway — Emmett. Honestly I didn’t hang out with him a hundred times outside work, maybe five times over the couple years I was there. But a couple things stood out.
One — he’s like us. He’s an idea machine. I need to get him on the podcast.
Sam: You’ve been saying that for three years.
Shaan: I know, I know. When I was still there I didn’t ask him because doing the podcast wasn’t exactly career-enhancing. But I think I could do it now.
He told me this gym idea. He’s like, “Gyms need to be made like video games. If I was going to leave Twitch I might start a gym structured like a video game.” He said: you show up, we give you a basic uniform, and you get a wristband that tracks your fitness level as you work out. You level up. Every time you come to the gym you’re earning points. As you earn points we give you better gear — more fitted clothes that look cooler. You can visibly see people’s different levels. You get access to sections of the gym that were previously locked to you. You get protein shakes for free.
He’s like, “I think if you gamify a gym people would come more.” And honestly that’s kind of played out with Peloton — leaderboards and gamification. But he was like, “This would be the World of Warcraft of gyms. People would grind years of their life to level up their character to 60.” Like people do with karate and jiu-jitsu — blue belt, brown belt, black belt. You keep showing up for 10 years because of the progression. But you don’t want the increments to be a year before you see progress. Every time you come, you need to be earning a little bit more and getting a little more unlocked.
Sam: I thought that was interesting.
Shaan: The second one: he was trying to plug in a computer to screen share one time — one of those situations where normally someone else sets it up, but it’s just the two of you. He’s wrestling with the cables. And he goes, “These cables — you know what, I think I could create the Apple for computer accessories. The best connecting cables, the best cable management, the best plugs. Nobody has done this right. Look at every office in the world — these cables look bad, function poorly, they’re confusing. I think you could build a really nicely designed cables and connectors company for every office in the country and people would pay for it.”
And I’m like: why do you think of these things? This is weird.
Sam: That’s what people say about you. He had a dozen ideas like this.
Shaan: What else did I notice? There was this famous story. Chris Sacca — popular investor, invested in Instagram, Uber, one of the best angel investors of all time. He told this story about being in Tahoe with his dad and Travis Kalanick, right when Travis was about to start Uber. They’re on vacation and Travis and Chris’s dad start playing Wii Tennis. They’re being casual, and Travis goes, “You want to step it up?” He steps it up and just crushes this guy — beats him so badly.
Mr. Sacca is like, “What the hell, man?” And Travis goes, “I’ve got a surprise.” He moves the Wii controller from his left hand to his right. “I was playing left-handed the whole time. I’m actually right-handed. Let’s really play.”
But then he says, “Actually, I do have to tell you something.” They go to the ratings board. He’s number two in the whole world at Wii Tennis.
Chris Sacca said: right then and there, I realized I never want to compete against this guy. Because his drive, his hobby is winning. He just proved that in his free time while starting this company he’s just going to be number one in the world at Wii Tennis just because he has to. He hustled my dad thinking he was left-handed. This guy is a winner. And that’s why Chris Sacca invested.
Sam: Peter Thiel — same way. He’s a master or grandmaster at chess. And in the PayPal Mafia, which is now the most successful entrepreneur group ever, I think several of them were very high-level chess players. There’s something to the “oven burns hotter” thing. It’s a combination of intensity and capability.
Shaan: Another observation: I asked Emmett what he does for fun. He’s like, “I love board games, you should come over for board game night.” I said, what are we playing — Monopoly, Twister? And he’s like, “We only play deep strategy games.” Not regular Catan — Catan on hard mode, or these other games I’d never even heard of that are multi-level, multi-hour strategy games. And basically it was him, the brothers from Stripe, and the founder of Reddit, just getting together to play these games on the weekend.
Sam: That’s exactly how I want them spending their time.
Shaan: The other one — we were in a meeting at Twitch, talking about some policy question. A girl streaming, her shirt was a little revealing, she lives in Denmark, what are the rules, should we ban her? These policy questions kept coming up and you could never win. Whatever Twitch did, people were like, you’re censoring us, you’re inconsistent.
And then Emmett — he has these hobbies where he reads obscure old books that no normal person would read, and he’d be able to connect the dots on two completely unrelated things. Sort of like what Ben does on his podcast. He’d be like, “Do you know how the potato farmers governed their farms in the Netherlands during the potato drought?” And the head of policy is like, no. “They created this governing council, and the cool thing about this rule of law was — maybe we should do this. Tell me what you think about it.” And they’re like, where do I find this out-of-print book you randomly read?
He would always bring these unrelated examples in. And the other thing — he would debate you over semantics until you couldn’t take it anymore. In an exec offsite, someone would say, “I just feel like we don’t have enough trust on the team,” and everyone’s nodding, yeah, trust, collaboration. And he’d be like, “Is it trust or collaboration? Because trust means this and collaboration means this.” And you’d say, you know, just working together better. And he’s like, “But define for me what you mean when you say trust. I want to make sure I have it right.”
And it wasn’t a gotcha — he’s a hyper-literal person. “When you say it’s failing, I take that to mean it’s failing — is it failing, or is it this other thing?” And people would get lost in the sauce of this semantic debate. But he really cared about words, what they mean, whether we’re using them correctly to describe what’s going on. Everybody else was kind of hand-wavy about stuff — the higher you go up an organization you sort of get paid for being a politician. He was the opposite of that.
The Pattern of Elite Founders [01:14:00]
Shaan: Now when I invest in founders, I look for this stuff. I invested in this guy who created something called Ski-O. He’s like, “Yeah, I was a professional Call of Duty player before this.” He was top three in the world in some game on the east coast server or whatever. And I’m like: okay, you’re a winner and you’re obsessive. You find the rules of the game and then optimize to win. Sure enough, with his startup he’s a wrecking ball with customer acquisition. He finds the rules and maximizes his edge in everything, to the point where he’s kind of a nut job. But it works.
Sam: At the end of episodes like this I’m exhausted. I feel like I exerted so much energy both listening and speaking. Do you take a nap after?
Shaan: Not at all. I leave these on a high.
Sam: Me too, but then I need to go rest. You never even have water. I usually drink three drinks during these. But I’m exhausted. I get exhausted just listening. That’s probably my fault — sorry.
Shaan: No, it’s a good exhausted. Like, god, I just learned so much. I need to get better at this and this and this. I was so intensely listening and enthralled that I need a nap.
Sam: Balaji went on Tim Ferriss’s podcast and did like a three-hour episode. I listened to it last night. Literally while it was happening I could feel my brain fatiguing trying to digest what this man was saying. He’s so — and by the way, Tim Ferriss sounds like an absolutely defeated man on these podcasts with Balaji. Balaji would be like, “Anarcho — on one hand you have the world capitalist, on the other the anarcho-capitalist, do you know what I mean when I say that?” And Tim’s like, “No, tell me.” That happens 40 times in the episode. “Are you familiar with the Schrödinger window?” Tim’s like, “No, but I assume you can tell me.”
One time when Balaji was with us, he was like, “You know, Bitcoin is kind of like the Battle of the Bulge — you remember how the Germans were doing it…” And I’m like, dude, I don’t. I needed an analogy for your analogy. Don’t reference the Battle of the Bulge to explain a complicated topic. Or, “Bitcoin is kind of like neuroscience — you know how the chemicals interact this way? Think of Bitcoin as the amygdala of the crypto ecosystem.”
Shaan: I don’t know. Yeah. That was happening. But you did that to me in this case, and I could totally keep up. It just was a good story. You talked a lot.
Sam: That was a compliment — I get exhausted because I’m so into it. I love hearing these stories.
We used to do these mastermind dinners — me, Sam, and like three other guys, every two or three weeks. Each person gets 20 minutes to say what’s going on in their business, what’s working, what you want help thinking through. But everybody runs 10 minutes over, we start 10 minutes late, it’s three hours long. We’d start at 7, finish at 10:30. Skipped dinner. Nobody went to the bathroom.
Two reactions: some people were exhausted and like, great but can we be shorter next time? I get that logically. My reaction — we used to do some of these at your office, the Hustle’s office. I would leave your office at 10:30 and go straight to my office and just work for four or five straight hours. I was so inspired. A combination of inspired by your guys’ awesomeness, panicked that I have so much I need to do that I now realize, and like oh, I need to do all of this tonight.
Shaan: I need to go drink Gatorade and take a power nap.
Sam: Yeah, I need to fuel up. I better eat a Clif Bar. My blood sugar’s low, I can feel it.
Shaan: I guess that’s the episode.