This guy cured his dog’s cancer with ChatGPT + 4 other crazy AI stories

- March 19, 2026 (1 day ago) • 52:07

Transcript

Start Time	Speaker	Text
00:00	Shaan Puri	Alright, there's a lot of stuff going on — specifically AI stuff. It feels like every day there's a new crazy story about AI. So I wanted to do a little bit with you, Sam: crazy AI stories, one crazier than the next... well, except they get crazier with every single story.
00:21	Sam Parr	I was talking to a friend, and he said he was trying to talk to his boss about something. Apparently the boss had created four or five agents that he calls employees. When my friend gave feedback — like, "We need to create X, Y, and Z" — the boss replied, "Can you talk to my chief of staff about that?" So my friend had to message the chief of staff. They were messaging the chief of staff and weren't getting the answer they wanted. They went back to the boss and said, "Your chief of staff sucks — what's their deal?" The boss replied, "Well, can you tell her boss, my head of operations?" So my friend literally had to go to the boss of the agent to say, "Can we get this person on a PIP [performance improvement plan]?" This boss of a roughly 50-person organization was referring to these agents as people and making his real-life employees interact with them. Have you done anything that extreme with AI and agents?
01:25	Shaan Puri	No, but I find the whole "I'm going to have this entire, like, autonomous company of agents and then me on top" thing to be so funny. Because it's like, who says you get to be on top? Why would you be the all-knowing one if AI is so good that it could do all these jobs? You think it just can't do yours — like, why is yours so special? It reminds me of that poem: > "First they came for the socialists, and I did not speak out; > then they came for the Jews, and I did not speak out; > then they came for others, and when they came for me there was no one left to speak out." That's basically what I think is happening with AI.
02:04	Sam Parr	Andrew Wilkinson came out with this personality test, and I took it. It was really cool. Listen to this: my wife and I did it together because there's a feature where you can share it with your spouse, and there's another option to share it with your business partner. I told my business partner to do it, and he mistakenly shared the spouse (the "wife") version. It talked about our sexual compatibility—basically whether we're compatible or not.
02:27	Shaan Puri	Wait, you shared "he and your wives"?
02:30	Sam Parr	My—my, I shared it with my wife and I go, "Joe, this is cool — do this." He did it on his own and was like, "Oh, I did it with my wife too. It was really fun." I said, "Hey, there's a button you can click so we can see how compatible we are as business partners." He mistakenly sent me the "partner" one, which meant a romantic partner, not a business partner. There was another button that said "coworker" — he meant to click that. I'm like, "Well, good news: we are very sexually compatible." And so he sent that to... I have Claude read my Slack, and it gives me advice on how to reply to our Slack messages based on each other's personalities. It's insane.
03:15	Shaan Puri	Wow — it's just incredible. I could see the cynical take; I see it right in front of me. But at the same time, it's something cool, right? So, okay — give me an example: what did it coach you? What did it teach you differently?
03:28	Sam Parr	Well, today it was like: "Joe, if you criticize the organization because he's the CEO, he's going to react defensively," and that's one of his— that's one of his things. He tends to get defensive if you criticize X, Y, and Z. So, instead of criticizing it that way, say: "My opinion is, we should do this instead." For me, it was: "Sam tends to criticize bluntly, but he doesn't mean it in a mean way." That change has helped us communicate. It's been really effective.
03:56	Shaan Puri	"But he does like kisses on the neck, so maybe."
03:58	Sam Parr	"Try. Yeah—if you see him, just smack him on the ass and tell him you appreciate him. What's your first crazy story?"
04:08	Shaan Puri	> "It's a good story. That's amazing. Did you guys just swear, like, to never speak of it, or what?"
04:15	Sam Parr	Yeah, I just— I was like, "Just so you know, you sent it to me." I saw... I saw the letters "sex." I knew what it was; I just scrolled right past it.
04:28	Shaan Puri	That's amazing. Alright.
04:30	Sam Parr	If you want to hit your goals, you need a system. I worked with HubSpot to put together a step-by-step process showing you how I use ChatGPT as a life coach. I uploaded my personal finances, my net worth, and my goals. The output is that I have this GPT that I can ask questions that I'm having issues with in my life, like "How should I respond to this email?" or "What's the right decision?" It's literally changed my life. If you want that, it's free. We will send you everything you need to know to set this up in just about 20 minutes, and I'll show you how I use it. Check it out—the link is below in the description. Back to the episode. </FormattedResponse>
05:08	Shaan Puri	Anthropic did $6 billion in revenue in a single month last month, which is insane. I just want to put this in perspective for you. I heard this, I think, on the All-In podcast [podcast]. They said this great line; he goes: > "Just to put that in perspective, $6 billion — what it did in February is more revenue than Snowflake or Databricks." Those are two of the greatest software companies of the last 20 years. These enterprise software behemoths are multi-$100 billion companies. That's more than their annual revenue, and Anthropic just, like, messed around and did it in a month last month, which is just insane.
05:48	Sam Parr	"Did you cancel your ChatGPT?"
05:50	Shaan Puri	"No, dude. I'm a pot [unclear]. One of a polyamorous... I'm maxed out on everybody. I got Gemini. I got Perplexity [the computer]. I got Claude. I got ChatGPT. Give me, give me all the intelligence."
06:01	Sam Parr	I—like, for a year since ChatGPT has existed, I was like, "this is my guy; I'm not gonna switch." But as soon as the new version of Claude came out, I explicitly only use that. I use it all day. It's so... it's so amazing.
06:18	Shaan Puri	Yeah, I want to actually do a survey — like a census-data survey — to 10,000 high performers and ask: "This is the situation — which friend do you trust? Who do you go to?" I had a legal question yesterday and I was like, "I'm just gonna go to Grok," because if I need the right answer or a thorough, rigorous answer, I'm going to Grok. Why, I don't know... this is all vibes. If I need writing help, I go to Claude. If I need to make something, I go to Claude. But if it's just quick and easy, ChatGPT is my best buddy. So you have these different use cases. I don't know if I'm completely alone in that. These are made-up, arbitrary strengths and weaknesses that I've sort of picked up — or maybe everybody's feeling this way. That's actually pretty interesting. </FormattedResponse>
07:02	Sam Parr	So what's funny is, people have been talking to me about Claude and Anthropic forever, and I didn't know what it was.
07:06	Shaan Puri	By the way, we'll just put a survey at the bottom of the show notes. I want to see how people use it. It'll be like five questions or something. I want to know: "Which tool do you trust for this situation?" I want to see if our audience mirrors our usage. We'll send it... we'll share the data next time.
07:21	Sam Parr	"I thought Claude was, like, second, third, or fourth place. I thought it was a joke, but I was wrong."
07:28	Shaan Puri	As per usual, I am the joke. It's how I normally feel about these things.
07:33	Sam Parr	Yeah... I thought that Anthropic was, like, you know, "that's just like using Yahoo." What—what about the founder? Did he work at ChatGPT, or is he just a guy? He's just brilliant, yeah.
07:46	Shaan Puri	"He was one of the core four or five people at OpenAI, and he left."
07:51	Sam Parr	Oh, wow. Okay — how much do you think that company is worth? Six times twelve is seventy-two, so they're doing 72,000,000.
07:58	Shaan Puri	AI over there—art AI. That was a quick little 6 × 12, like that.
08:03	Sam Parr	Yeah, no big deal.
08:04	Shaan Puri	They're currently valued at $400,000,000,000.
08:07	Sam Parr	"That's still shocking, right?"
08:09	Shaan Puri	By the way, other great news: I am now officially an investor in OpenAI. I'm trying to get into Anthropic — you know, at the low, low price of $800 billion. I have invested in OpenAI. "Call me a genius."
08:21	Sam Parr	"Good—congratulations. Who's this guy?" "His name's Dario. Is there anything interesting about him, or is he just a genius?"
08:28	Shaan Puri	I know what you're asking, which is... you know, that's, like, between you and me. That's kind of: does he also have a weird hobby? Did he have some strange upbringing? Was he a child prodigy? That's kind of what you're asking for. I'm sure he's a very interesting guy. I don't know of any of those things that we really like about people. I don't know if he has any of those, you know.
08:45	Sam Parr	Just a typical genius, okay.
08:47	Shaan Puri	He did say it's a pretty interesting thing: they're growing so fast. They basically have grown revenue 10x every single year for the last three or four years. It was like zero to $100,000,000; $100,000,000 to $1,000,000,000; $1,000,000,000 to $10,000,000,000; and now they're going from $10,000,000,000 to $100,000,000,000. Doing $6,000,000,000 in a month is indicative of that. He said the crazy thing is he has to invest in servers—he has to buy inventory. It's like a clothing brand: he has to buy all these chips, these servers, and these data centers. He's basically being seen as a wuss because he's not out there yelling, "I'm gonna do a trillion-dollar data center." OpenAI's Sam Altman is doing that. On Dwarkash's podcast he pointed out: "The weird thing is we're one of the most successful companies ever, but if I'm wrong—if I bet wrong on that—we'll go broke." He added, "If I bet on another 10x but we only grow 5x in a year, we could go bankrupt." So it's a very strange situation where he has to keep being conservative even though they have the most explosive growth; they're sort of limited on that. Now they're the number one app in the app stores on Google and Apple, and they're just growing incredibly fast.
10:01	Sam Parr	To put this in perspective: when you and I were 22, growing up in San Francisco and trying to make it big, the golden standard was "triple, triple, double, double." It meant your revenue needed to triple, then triple again, then double, then double. That is how you build a once-in-a-generation company.
10:18	Shaan Puri	Now, you were the winner if you were on that trajectory.
10:22	Sam Parr	And now it's 10 × 10 × 10 × 10. It's just—it's just... we're not even in the same ballpark, so it's kind of hard to fathom.
10:30	Shaan Puri	Yeah, yeah — it's great. The bar has been raised. Alright, next story: you probably saw this. This was going viral on Twitter everywhere.
10:37	Sam Parr	Amazing
10:38	Shaan Puri	Man uses AI to cure his dog's cancer The short story: an Australian entrepreneur—basically a data nerd—noticed a bump on his dog's back. Rather than accept "there's nothing you can do," he asked, "we sure about that?" He decided to enter the "high agency Olympics" and manually go cure his dog's cancer. Here’s roughly what he did: - He talked to ChatGPT about potential approaches. That gave him an idea: sequence the DNA of the tumor and compare it to the DNA of the dog's normal cells. - The idea was that, by extracting the tumor DNA, he could find the mutation causing the growth. Comparing tumor cells to normal cells (or to earlier samples) would reveal the anomalous part of the genome that might be responsible. - He paid out of pocket to a lab to get the sequencing done. They agreed to run it under research purposes. - With the sequence data, he used AlphaFold (the Google DeepMind product for predicting protein structure) to predict the folded shape of the relevant protein. Knowing the sequence alone isn’t enough—you need the three-dimensional structure if you’re going to target it with a treatment. - Using Grok, he designed a single-shot custom vaccine. - He convinced a lab to manufacture the vaccine for him. He also noted that the regulatory hurdles were harder than the technical parts—getting approval to actually do this and to have the vaccine manufactured was the main challenge.
12:25	Sam Parr	Approval from the vaccine manufacturer.
12:28	Shaan Puri	Yeah. The vaccine manufacturer needed to know—like, "you don't get to just... you and I can't just go and make vaccines." So there are protocols in place to make sure people aren't manufacturing—whether it's viruses or self-treatments—that could do harm and whatnot. And so they were able to. As he was doing this, someone who already had the ethics clearance to do research on pets heard about what he was doing. He kind of used their umbrella to get the vaccine made, gave the vaccine, and it cured his dog. Legend. Paul, you're a legend.
12:59	Sam Parr	I have a friend who has cancer, and he actually works at OpenAI. I tweeted about this. I was like, I don't particularly know why he wanted me to tweet this, but he was like, "Hey, I need one of these three drugs, and I'm having a hard time getting in touch with the particular person who can get the drug for me." Interestingly, Mark Cuban reached out to me and said, "Give me your friend's cell phone number. I'll give him a call and hook him up — I'm on it." Martin Scarelli did the same, and I was like, "Well, buddy..."
13:29	Shaan Puri	Do you want Batman or do you want the Joker?</FormattedResponse>
13:31	Sam Parr	Yeah, I was like, "I don't— you know, take what you can get here." Yeah. And so he did, and, you know, now he's getting what he needs. But I have to imagine he worked— I don't know if I said this— he works at OpenAI. And I'm like, "If you're reading this, you have to imagine that this is the next step that you're gonna take, right?"
13:50	Shaan Puri	Right, yeah — that's crazy. Is he doing a similar thing? Is he basically trying to self-create, like... is he using AI tools to come up with hypotheses and then possible treatments for himself? What's going on? What is your friend doing? That sounds interesting.
14:08	Sam Parr	I don't ask him. I let him come to me when things are going good. I try not to bring it up too much. This is his second time that he has—he's had cancer, lost his leg, and got cancer again. So I'm like, "Whenever you need me, let me know." He'll visit my home when he's, like, at a New York doctor. But I do know that there's this guy named Sid. Sid is the founder of a company called GitLab. Have you heard of GitLab? The like—know, like...
14:33	Shaan Puri	Big developer tool.
14:34	Sam Parr	Multi-billion-dollar company. Yeah, so Sid and my buddy became friends because Sid wrote this article called "I'm Going Founder Mode on Cancer," and so Sid has...
14:44	Shaan Puri	Did Sam Sid leave GitLab to do this, or is this like a side thing he's doing?
14:48	Sam Parr	Well, Sid has the same cancer as my friend. I don't know if he's left GitLab for sure, but he goes, "I'm going founder mode on cancer." So he's hired a team of doctors. In his article he said they are using AI to do a lot of this work. My buddy connected with Sid because they have the same cancer, so he's using Sid's doctors as well — it's a fancy concierge cancer doctor. They are going through the journey of trying to cure this very, very rare cancer. Sid, who is the founder of GitLab [a technical company] and, I presume, a billionaire, is a very special person who could figure out these types of cases. He's actually documenting on his Substack how he's trying to cure his cancer. It's really interesting.
15:31	Shaan Puri	That's amazing. Alright—well, we're wishing them the best. That's a— I mean, I feel more hopeful about this sort of thing than ever. I think the combination of AI and, on top of AI, the fact that Ozempic and these weight-loss drugs seem to have done something that sounded too good to be true... it's like, is there a magic pill for weight loss? It turns out there actually is a "magic pill" for weight loss. That kind of breaks a mental barrier: maybe you can one-shot certain issues or problems that plague huge percentages of people. So I feel, I don't know, more hopeful than ever that there are new breakthroughs and new inflection points that would change how cancer treatment works over the next one to ten years.
16:23	Sam Parr	And if we totally botched the name—botched the name. So if we did whatever, but the GitHub founders or GitLab founders botched your company too, if this makes it to you guys, DM Sean or me, because it could potentially be fun. If you want to talk about this on a podcast, it would actually be fun to talk about that. That's actually super fascinating. But it is crazy that this stuff still exists. Like, when you think about it—I remember when LiveStrong came out and that was like, "Okay, someone famous finally has cancer when they were young; we're going to bring awareness to this thing." And now, when you think about all the billions and billions of dollars... I think, what's her name, Kate Middleton, had it. It is kind of crazy how this still exists the way it does.
17:00	Shaan Puri	Yeah, yeah. Then you just needed—you know—you needed the technology, and then you needed the sort of… the will. Here's another thing that makes me helpful.
17:08	Sam Parr	And so...
17:09	Shaan Puri	Nat. Who? I think you know Nat well.
17:11	Sam Parr	Yeah, I'm good buddies with Nat.
17:12	Shaan Puri	He just tweeted out this really interesting thing. He goes, "Make $1,000,000 by graduation or get 100% of your tuition refunded." That's the premise of the new high school for entrepreneurs that Cameron and I are launching this fall through Alpha School. We need two or three coaches to make this happen. The premise is they're creating a new high school using Alpha School's technology, which we've talked about before — the AI education software that Joe Lamont has put a billion dollars into building. It is getting tons and tons of buzz everywhere. So now they're doing a high school based off of it, and the idea is almost incredibly incentive-aligned... almost, maybe too incentive-aligned.
17:51	Sam Parr	I think it's too aligned.
17:53	Shaan Puri	Too aligned: "Make a million dollars by high school graduation, or your tuition was free." I think Alpha School's like $70k or $80k a year—the kids' version that's in Austin. So, you know, this is a pretty big bet that they're making. This is crazy. Is this not, like, one of the crazier tweets?
18:15	Sam Parr	I thought that. So, Hormozi talks about "nailing the—nailing the offer."
18:19	Shaan Puri	And, well, he says, "An offer... you know, they'd be stupid to say no to." Right? Like, "It would be dumb for me to turn down this offer."
18:28	Sam Parr	If you are the type of person where this idea tickles your fancy—where you want your kid to build something—this is the greatest offer of all time.
18:35	Shaan Puri	"I mean, it's crazy. Do you think this is even— I mean, are we just old guys who are not up with the times? Like, is this really possible? High school. So, high school: you're 14 to 18 years old. If I gave—even the most switched-on 14 to 18 year olds—and if I was just coaching them every single day... Let's say I had—let's say that you and I had 20 high schoolers."
18:59	Sam Parr	"So he asked you to be a coach?"
19:01	Shaan Puri	Every day we were their coach. Let's just pretend that was true. How many of the 20 [people] do you think we could get to make $1 million over the four-year [period]? It's not zero, but it's also not 20. What do you think that number is?
19:12	Sam Parr	One. Four years.
19:15	Shaan Puri	Four years is kind of a long time, actually.
19:17	Sam Parr	Four years, actually. Most... I think most.
19:21	Shaan Puri	Right.
19:21	Sam Parr	Well, like, in revenue... we can create some type of app that buys ads to grow and is not losing money. I'm—I feel confident that I could do that half the time.
19:31	Shaan Puri	I'm going to use the assumption that there's external funding available—funding they can go raise if they want—but we can't give them the money. Also, that it's not $1,000,000 of revenue. Maybe... is that what he meant by "make $1,000,000," or did he mean something else?
19:46	Sam Parr	If it's revenue, then there's a quite high chance I could do that a lot of the time. If it's profit, that's much, much harder.
19:53	Shaan Puri	Right. If it was profit, I think we could probably get five out of twenty in a four-year period. I would guess that we'd have some kind of hit rate around 25%.</FormattedResponse>
20:05	Sam Parr	It's pretty magical. So let's talk about why we said it's too aligned — we're joking. I think this is really good for a very particular type of person. But I do think that optimizing toward entrepreneurship for everyone and glorifying it is overdone; it's being over-glorified. I think that's wrong. I don't think it should be that way for everyone. I think it should be for a small number of people who are interested, just like saying we should get everyone to play music, everyone to play sports, or everyone to be good at accounting. So I think this is a bit... I'm not criticizing these guys, because they're saying, "For those interested in this, this is the offer." I just don't think this should be the default for all high schools.
20:43	Shaan Puri	Oh — you feel like, in general, everybody's gotta get on their entrepreneurial grind?</FormattedResponse>
20:49	Sam Parr	"I think that's stupid."
20:49	Shaan Puri	This is not the right message because, obviously, not everybody needs to get on their entrepreneurial grind.
20:54	Sam Parr	Yeah. I just don't think that's necessary. What—what were you right? What were you saying?
20:57	Shaan Puri	I was saying it's too aligned — you didn't need to make it a million. You could've... what's the real thing? Basically, I would say if you did $100,000 this would have been — I mean, that would still be incredibly successful, right? If you can make $100k or $150k or something like that. But of course that's not the marketing message that would stick. The reason this got onto the show today is because he said a million. In that sense, it's the absolute right number to choose. If nothing else, it is a PR offer. I don't know what that does to the business, if that distorts things long term, but in the short term you get the big benefit of putting something out there that is buzz-worthy. It's like when Peter Thiel did the Thiel Fellowship. I think he only spent $2 million out of pocket to do the fellowship — $2 million is not nothing, but in the world of billionaires giving money for causes there's plenty of people who write bigger checks than that. So why did the Thiel Fellowship get so much notoriety? Because he attached it to a very spiky message: universities are a bubble. Like the housing bubble and other bubbles, it's a bubble because nobody else thinks it's a bubble. That's the only way you get to be a big bubble — when everybody sort of assumes education's great. How could you criticize education? Housing's great, everybody should own a home — you can't criticize that. What he said was not just, "I'll give this to people who have good ideas," but, "I'll pay you to drop out of college" — literally the opposite of, "Hey kids, stay in school." Because it was so spiky, he got tons and tons of notoriety because...
22:39	Sam Parr	He got tons and tons.
22:40	Shaan Puri	By gaining notoriety, he attracted the right people because the message reached far enough that those—like the Vitalik Buterins of the world and the Dylans of the world who create Figma—heard the message, were able to apply for the Thiel Fellowship, and came in and made it happen. So I think maybe this might be the right approach: like the XPRIZE, where you have to do something novel, noteworthy, and almost polarizing in order to make it work.
23:03	Sam Parr	So let me tell you a quick story. On Friday I went to Austin Reif's house for dinner. He had over these eight young guys, and I posted a link to who they are in Slack. If you look this up, the New York Post just did an article on them five days ago. It says: "These Gen Z digital moguls are building seven-figure companies while lifting weights." Basically, I met these guys in Hampton and I was like, "Austin, you gotta meet these guys—this is incredible." We talked about Josh. That's a guy—Josh. We talked about him; he has a company called Street Talk. You might know Oliver—do you know Oliver? He had the company called...
23:40	Shaan Puri	yeah he had the tab shock the thing now he's doing a snastling
23:42	Sam Parr	And then that's Seamus.
23:43	Shaan Puri	"Dude, this is like a boy band that you're talking about."
23:45	Sam Parr	Like... that's Joey.
23:47	Shaan Puri	"Joey's cute, Sheamus has highlights, and Brandon is dating Rebecca — what's going on here?"
23:55	Sam Parr	So, listen: there were eight guys. The eldest was Oliver — he was 25, I believe. They're all young guys who share an office, and they've all started companies. One of them was a guy named Victor.
24:07	Shaan Puri	Literally looks like Elon Musk—like, young. </FormattedResponse>
24:08	Sam Parr	Elon Musk — one of them is... Oh, I understand; he kind of talks like him, too. One of them raised $140 million for the company. Anyway, it was all these young guys, and we had them all at dinner. I was like, "I don't want to sound like an old creepy guy, but you guys are so young. Can you tell me everything that 21‑year‑old entrepreneurs are into? Because when I was 21 and I left school to move to San Francisco, I was a freak among freaks. It just seems like you guys are way more normal for how successful you are."
24:32	Shaan Puri	You might not want to call yourself "the freak amongst freaks." It's just... yeah, we're talking about eight young men you had over for dinner.
24:40	Sam Parr	Yeah — if we get a little Epstein [unclear reference], thank God my wife was there. You kinda get away with a lot when she's there. They were saying, "Dude, we are killing it for how young we are. The 13‑year‑olds are now coming after us. The way that you are talking to us — like you're in awe of how young and successful we are." He said, "There are so many 13‑ and 14‑year‑olds that we know who are building stuff that is going to be way better than us. It's just crazy how young these guys are who are building amazing things." "I mean, we had Zach — okay, Zach On from CalAI. He sold his company for tens or $100,000,000,000 ($100 billion) at the age of 20, I think 19. There's a whole generation of these kids. It's just absolutely amazing. Before, it was college dropouts." </FormattedResponse>
25:28	Shaan Puri	We made up that number, by the way. Didn't tell us 100 million.
25:30	Sam Parr	I'm making... Yeah, sorry about that.
25:33	Shaan Puri	Those are some assumptions, but this line—dude—is so funny. We talked about the problems we're facing: hiring, scaling, girls. Seamus told us, "your typical twenty-year-old dude stuff." What is this, the new "Gym, Tan, Laundry"—hiring, scaling, girls?
25:52	Sam Parr	It's just crazy. Have you noticed how young some successful entrepreneurs are right now? Before, it was the very, very rare college-dropout, Zuckerberg-style person, and that was a one-in-a-million case. Now it seems way more common.
26:10	Shaan Puri	Yeah, I told you, dude. I feel like the guy who's just a horse-drawn carriage driver. I've been driving this route for ten years and it's a good thing. I'm standing on the side of the road next to my horse and then a Tesla Roadster zooms by and I'm just like, "What the hell was that?" That's how I feel when I hear and talk to some of the young, inspiring founders — people doing content, e-commerce, AI. There's so much energy, enthusiasm, excitement, and talent that feels unleashed. By the way, the reason why to me is pretty obvious: the information is out there, and we're all memetic creatures. In the same way that, you know, I'll go to my kid's school function and half the moms there have lip filler and Botox and stuff — why? Because it's become normalized. They've seen it everywhere on Instagram, reality TV, whatever. You see it once, twice, three, four, five times and it starts to feel normal and accessible to you. When we were growing up there was no Twitter, right? I didn't have access. I never even met an entrepreneur who told me, "Oh, I have this business and here's how it works." I never listened to a podcast about entrepreneurship. I never watched a YouTube video. I don't even think there were podcasts when we were 14. So when we were 13, 14, 15, 16 — a pretty formative age — the only things I had exposure to were basically the NBA or skateboarding. That was what felt cool and what you could try to model and emulate outside, during summers or after school. Today, if you go home, you have access to Twitter, TikTok, Instagram, YouTube. On those platforms you can get into any rabbit hole, any niche you want. The people who stumble into the tech/Elon Musk bubble are going to see that this is normal — the idea that you can make millions of dollars. People who listen to the podcast we did with Zach hear, "Oh, here's this 19-year-old kid who just sold his company that was doing $30,000,000 in ARR." That now becomes normalized for a huge generation of people who will then do their own version of that. We didn't have that advantage. I don't think the human species evolves that fast, but I do think if you give it new training data — if you give it new stuff to mime and copy — people will do pretty well.
28:55	Sam Parr	They were talking about dating, and I referenced the book The Game. They asked, "What's The Game?" and I said, "Come here, son."
29:02	Shaan Puri	Called the Old Testament.
29:05	Sam Parr	Come here, boy. Sit on the seat.
29:08	Shaan Puri	My lap.
29:09	Sam Parr	Alright, what's the next one?</FormattedResponse>
29:12	Shaan Puri	Alright. I don't know if you saw this: Pokémon Go has one of the great pivots of all time. Did you see the story?</FormattedResponse>
29:18	Sam Parr	"I did not."
29:20	Shaan Puri	Alright, so if you remember, Pokémon Go was the hit game. I don't know when this was, like...
29:24	Sam Parr	Like 2019 or 2020.
29:25	Shaan Puri	Yeah — about seven years ago or something — it was a huge deal. It was this new style of game: you would take out your phone, put it in camera mode, walk around a city, and then discover a Pokémon. The game would overlay the little Pokémon on top of the street where you were standing. That became a one- to two-billion-dollar game. I think they sold the company for $2 or $3 billion because it had this huge hype cycle, and then it kind of died down. There's a really interesting story: they sold the game, but they kept all of the photos and videos. What that means is they had — I don't know — maybe a hundred million players walking around every city in the world with their phones out, capturing what the streets looked like. If you think about it, that's actually really valuable data for AI that wants to do not self-driving but self-delivery — like delivery bots that will go on sidewalks through neighborhoods or down city streets and have to deliver something. They need to recognize “Where am I? What does this look like? What is the terrain? How does this work?” and so on. So they took that data and they're licensing it out to all these AI companies now because they're the only ones in the world who have this data. You've heard the phrase "data is the new oil" — that's a little misleading, because there's just oil and then there's very specific, niche types of oil. You might have a specific type of oil that nobody else has, and Pokémon Go — the company behind it, Niantic — has that data for real-world behavior. It's similar to [Google Street View]. Google would send out a car with a rotating camera on top, taking pictures of everything around it, and they would literally drive every street in the world. They did it roughly every ten years. Side note: one drove by my house the other day and I got so excited I went outside and flashed it. I was like, "Yes — that's internet history, baby. That's for the next ten years." If anyone ever looks up this property, they're going to see my hairy belly out there. I did that — I couldn't think of what else to do; my brainstorm only had fifteen seconds, so I went Mardi Gras on it.
31:44	Sam Parr	Instead of fidgeting, I've hated the analogy of "data is the new oil." For years and years people have said this, but there was no use for it. Actually, it's kind of a good analogy. Did you know the oil industry existed? John Rockefeller started Standard Oil in the 1880s, but cars weren't really invented until around 1915–1918, and they didn't really get popular until about 1920. So up to 1915, what did they use oil for? Well, it was mostly kerosene. Street lamps—someone would literally go and light the lamps in a city—and that alone was enough to make the industry big. They were doing it and killing it, and then Henry Ford created Ford. I think in 1914 the first Model A—sorry, Model T—came about, and the oil industry exploded. What is going on in some cases is kind of like that with AI, where there's data and we're like "data's the new oil," but there's no use case for this oil. Yeah, maybe we think this is useful, but what are we going to use it for? Now they have finally found a good place to actually put this oil.
32:53	Shaan Puri	Yeah, that makes a lot of sense. I feel like that's happened now with a few people: they woke up and were like, "What's under my back? Oh, it's a lottery ticket." That's basically what this is. We talked about Handshake — that company was basically a college-student internship/job-board type of product. It's not necessarily dumb; it was fine. It just wasn't going to be the biggest thing in the world. It was a useful product: college grads who want jobs, and employers who want to hire fresh college grads. Then they pivoted to basically training data [for AI]. They realized, "Oh, we have all these students in different disciplines, and AI labs will pay for curated training data." So whether it's, "Here are math answers — tell me if the answers are good," labeling whether the answers were good and giving expert feedback, or it's labeling data — that sort of thing. They woke up and found a lottery ticket in their back. They had to identify it and act on it, and credit to them for doing so. A better analogy is the person who buys a home and then gets a knock on the door: it turns out the home was built on top of an old mineral reserve and you can get paid far more than the value of the house to just leave and let them bulldoze it. I feel like that's what's happening with a handful of companies right now because of their unique datasets.
34:25	Sam Parr	Yeah, I don't know if it's going to happen to a lot, but it definitely is a good example of this happening in a wonderful way. What's the next one?
34:31	Shaan Puri	I don't have much to say about this. This guy used ChatGPT to sell his house. So: "Florida man sold his house in five days using ChatGPT without real estate." I don't really have much to say about this, but there is a funny "Florida Man" game that I played the other day. Have you ever done the "Florida Man" and then your birth date?
34:45	Sam Parr	No.
34:47	Shaan Puri	Well, let's go. What's your birthday, Sam?
34:49	Sam Parr	June 15.
34:49	Shaan Puri	What's that, June 15?
34:51	Sam Parr	June 15.
34:53	Shaan Puri	So, Florida man — June 15. Here's the top story: Florida man pushes pastor, knocks churchgoers to the ground.
35:03	Sam Parr	What's the second one?</FormattedResponse>
35:04	Shaan Puri	Florida man said, "It's 30 months for illegally transporting an alien."
35:08	Sam Parr	"That's like something I would do."
35:12	Shaan Puri	"That's the plot of E.T. I'm pretty sure he was a good guy in that. Alright, I got one more here. So, have you seen this Karpathy Jobs thing?"
35:23	Sam Parr	no but I saw the anthropic version but what's the karpathy one
35:27	Shaan Puri	Do you know who Karpathi is?
35:28	Sam Parr	I don't even know who he is, but I know that when people say, "Karpathy says 'blank' about AI," you're supposed to take it seriously. He...
35:37	Shaan Puri	He was one of the OGs of OpenAI. He was then poached by Tesla and ran Tesla’s self-driving program for about a decade. He went back to OpenAI after that, and now he’s an indie hacker — like Peter Levels with six brains instead of one and a billion dollars. He’s just fucking around and posts random AI experiments on Twitter. He made something called "karpathy.ai/jobs" [visible on the screen here]. Basically, he took all of the U.S. jobs data from the Bureau of Labor. That’s about 143,000,000 jobs in the U.S. economy. Each square’s size represents how large a part of the U.S. economy that job is. For example: registered nurses — 3,400,000 in the United States; construction laborers — 1,600,000.
36:31	Sam Parr	Secretaries 3.5, so it's percentage by...</FormattedResponse>
36:34	Shaan Puri	Size by.
36:35	Sam Parr	Quantity of workers.
36:37	Shaan Puri	Popularity of... because it.
36:38	Sam Parr	It says, "Cashiers are a huge percentage of that," which is insane. So cashiers account for 3.2 million jobs. That's wild. That's cool. Okay.
36:47	Shaan Puri	Correct. And so then, why is it—what is green and red? It looks like a stock market chart. Basically, he told AI to estimate the likelihood that a job is going to disappear based on AI's capabilities. So, based on what the job does and what AI can do, how disruptable is this job? How at risk is this job? For example, construction laborers are green because they're more in the safe category from AI—AI today can't do what those people do. But customer service is in red, obviously. He went through every job in the U.S. market and you could almost see how turbulent this will be for the labor market, because you can see how much red will appear on the screen. If AI is going to disrupt a very small niche job, that's not a big deal. But if it's going to disrupt secretaries, admin, cashiers—I'm surprised waiters and waitresses are here, and high school teachers—I'm surprised that one's in red. Kindergarten and elementary are kind of in an orange state here. [Context: mapping job roles in the U.S. labor market to AI disruption risk]
37:54	Sam Parr	"That's insane. What?"
37:56	Shaan Puri	And then he has a caveat. He says "caveat on digital exposure", so let me just read the whole thing: > "The digital AI exposure option is one example. It estimates how much of current AI—which is primarily digital—will reshape the occupation. But you could do this for any other type of thing, like offshoring risk, etcetera." He even includes the prompt here. This is a cool little hack project by him on the side, and he points out that there are obviously a bunch of caveats. For example, software developers scored really high, while some people believe, "Well, AI is doing all the coding now—what does that mean?" He's basically saying that because AI transforms the work but the demand for software is going to grow as each developer becomes more productive, the score does not account for demand elasticity, latent demand, regulatory barriers, or social preferences for human workers—that'd be like kindergarten teachers, for example. Many high-exposure jobs will be reshaped and not replaced. So "red" doesn't mean replaced; it means reshaped. He did another thing that was pretty crazy: he created an auto AI researcher that runs all night. He explicitly programmed it not to look for human feedback, so it's a recursive loop: generate a hypothesis, do research, create an output, evaluate your own output, give yourself the next direction, and go again. It just runs in the background all night doing its own research. Now it's called "Auto Researcher," which is really how AI is going to advance—the big leaps from AI are likely to come when AI programs the next leap.
39:24	Sam Parr	Have you become the person where...? So my Friday and Saturday nights I'm like, "I want to learn this stuff. This is so exciting. I'm so locked in — I love this." It could be small stuff. I don't need to use Tableau anymore. I just make my own dashboards in Notion, and I created a script to auto-update it. All these things make me feel like I have superpowers.
39:47	Shaan Puri	"Yeah, I feel like I have superpowers. I'm like the early—you know, in Spider-Man when he realizes he can shoot webs. But really, all he's doing is... shoot. He's like summoning the Coke can to him. It's like, 'I could just get up, or I could just web it to myself.' I didn't have to get up from my bed. That's kind of what I'm doing. I'm doing a lot of stuff that feels cool to me; it feels like a superpower. But does any of it matter right now? No. Still, all the useful stuff is just chat with Chattypetty, chat with Grog, chat with Claude."
40:16	Sam Parr	Dude, I've done really simple stuff where I'll say, "Make 20 landing pages in HubSpot for when people search X, Y, and Z that are related to my company." That's a very monotonous, boring thing that I normally would have had to do by hand, and it just made it all in five minutes. This is for a lot of people listening: the agency business model just got really interesting. Historically, agencies are a pain in the ass, but they're actually pretty amazing because you don't need to test if there's demand for an agency. Right? Everyone needs an agency—every business hires a lot of agencies. You go through a lot because you don't particularly like them. But running an agency has always been a huge pain in the ass, regardless of how great the agency is. Now it's significantly different. I think that if you own an agency, you could take the path of "my business is over," or you could take the path of "my multiples are gonna go up a significant amount."
41:11	Shaan Puri	my buddy romine sent this blog post to just me and ben like a maybe a year or two years ago and he wrote it was called service as a software so instead of software as a service service as a software and he basically had this thesis he's like service businesses have historically had lower margins and lower multiples because two things one because it required so many people to do the gross margins were worse and two because it required nice skilled people you couldn't scale it the way you can a piece of code that can just keep running anywhere you in the world whether one person is using that code 10,000,000 usually doesn't really matter and he was like ai seems to have changed this and he called this early and he was like now services businesses are gonna have high gross margins like 75% gross margins instead of like fifty forty% gross margins and he's like the reason why is because obviously ai is gonna make the service company more efficient with one person you could do what you know four five six seven people were doing before and then the second thing is that the more people have software service is gonna be become more rare it's gonna become more like it's gonna become the harder thing to get and in addition to that you have so not only does it have the higher multiple but also it'll scale better right because if you're using the more the more internal software you have the more clients you might be able to serve and so you you get the scalability and you get the margins and so he had this idea and this is exactly what we're seeing we were talking to a bunch of pe folks and they've all basically shifted budget away from buying saas to buying service companies and they value the service companies like they used to value software companies which is kinda crazy as a shift to make for somebody who was like ingrained in my mind that like you know software is the thing and services are bad you know and he was he was a 100% right let me show you this thing that I made alright so I think I've told you this before but my brother-in-law who's come on the podcast before his name's sanjeev he he is one of the biggest owners of retail shopping centers in the country he owns I don't know like some I don't know what the number is but like maybe like you know a couple million square feet of real estate of retail shopping centers in the country probably one of the top five to 10 operators but he you know it's like a blue collar type of business as in like you don't you know his job is he does a lot of construction he does a lot of leasing with brick and mortar businesses so he's not hugely exposed to ai so he he joined like tiger twenty one or something like that was at a meeting and he and they were he's like dude was all about ai he's like I feel like I gotta like know more about ai he's like seems like you know I'm I kinda know about it but I gotta know more you know like seems like I'm probably under investing in this so last night I just did I did a little demo for him so I basically went into claude and just go I'm a retail shopping center owner this is my business tell me 10 ways I should be using ai so it gives me 10 ways so it's like here's 10 things you could do a retail expansion and contraction tracker okay so like you know your tenants are these retailers you could scrape earnings calls 10 k's press releases trade publications and flag which stores are opening more locations which ones are closing locations and that would create a hit list for your team to go call those tenants and basically be like top of mind for them as they look for new locations lease expiration you have all these leases right if you have let's call it 50 centers that you own and each center has 30 tenants now you have a 150 tenants that you would have each one has a unique lease on it a unique timetable you need to know who it's expiring and when just upload all of it to a folder and it'll tell you what is expiring when it'll give you notifications and it'll cross reference that with their public announcements so you know hey they had a hot quarter you know go and look for an expansion of that lease they're contracting that's gonna be at risk you should start thinking about possible new tenants to put in and he gave a list of like know ten ten different things and then I just said cool give me the prompt to to build this to build number one that like expansion tracker so check this out so then I go and I basically take that prompt I throw it into the new I wanted to try the new perplexity computer I don't know if you've played with
45:14	Sam Parr	This at all, but it's the computer — just OpenClaw, yeah.
45:18	Shaan Puri	Like OpenClaw, but in their sandbox. Less technical work to start, but it's in their environment, so it built it. So, check this out — it's basically like, "check this out."
45:29	Sam Parr	Is this just makeup?
45:31	Shaan Puri	No. This is like an active, live website now that he can use. He built this.
45:35	Sam Parr	His data—or this is his data.</FormattedResponse>
45:37	Shaan Puri	This is not his data. This—I just said, "Hey, here's seven states that I operate a lot in: Texas, Nevada, New Mexico, Utah." </FormattedResponse>
45:43	Sam Parr	Oh, got.
45:43	Shaan Puri	So I just— I didn't tell it anything else. We went and said, "Okay, Dollar General is a retailer that's expanding; know their account." They're going to fit 450 new stores. Then you click in and it says they've planned this many new stores for 2026. They're going to do 4,000 remodels and 20 relocations. They currently have 20,000 stores. Here's where they're likely to be expanding. "Here's the quote from the source, from the CEO, talking about what they're going to do." Right — crazy. Let's go to contracting. Walgreens: they're planning to close 100 stores. They've already closed 500 since 2024. Then you have things like a Macy's one, and other different retailers and how they're doing. Then you're like, "Okay, cool — now I basically have a hit list of retailers that I can have my team focus more attention on (and less attention on) in terms of getting new leases done, forecasting and anticipating, and being proactive about potential vacancies I might have in my centers in the future."
46:46	Sam Parr	What did he say when he saw that?
46:48	Shaan Puri	He was like, "Hey, can we—can we do a call today? I need to, like, kinda... what the hell just happened. I just need to talk to somebody. I need to talk to somebody. I need to see what the—" He was like, "I'm on ChatGPT. It seems like I should be using some other things. Can you tell me what these are now?" And I'm like, "Yeah." You know, because I've invested millions of dollars with him. So I'm like, "Yeah, I can just show you a couple of things that I think are gonna..." Today it's all manual, right? Everything is manual—everything is gut, everything is experience. But you can productize a lot of that and turn it into, you know, competitive advantage for yourself.
47:22	Sam Parr	The problem is turning it into a competitive advantage. It's about getting someone like him—smart and successful but not well versed—and showing them, "Oh my God, everyone feels the same amount of doom and overwhelm." They're like, "I know this is important, but I don't know the first place to turn to in order to get help." Now, you and I have Twitter; we could just tweet out, "Who knows how to do [blank]?" or "Is doing [blank] worth it?" But it doesn't matter how successful you are: if you're just a guy and you're not locked into this industry or surrounded by nerds, I don't even know where you would start.
47:55	Shaan Puri	Well, yeah. I think you have to figure out some proportional investment of education and time to work on this—whether it's once a week, three weeks, you hire a consultant, or whatever you're going to do. It's one of the reasons I think the biggest opportunity right now—the obvious no‑brainer thing that anybody who hasn't made it but wants to make it in life needs to do today—is to basically go sell AI transformation. What do you do? You go find business owners and operators. You become an AI expert in your nights and weekends, basically. Then you go to them and you sell a "no‑risk offer" to them, which is: I will come and I will find ways—I will come and do an audit and basically figure out how I can use AI to improve your business. Then you do the audit, try to figure out how to improve their business, and you sell them the solution: the implementation of it. "Hey, this is what I think AI could do for you. I'll do it for you if you pay me this amount, and here's the value it's going to drive for you." You do it once—with one dentist, let's say—then you go do it with 50 more dentists, and you can literally make millions of dollars doing just that blueprint. That is the blueprint. If you remember when social media marketing agencies became a thing, there was this acronym—SMMA—and it became this like "business in a box" that anybody could open up. Social media is a big deal; businesses today don't really understand how to use it. They're using it in their personal life, but they're not using it for their business. Anybody in any locale, any niche, can create a social media marketing agency. You could go to these business owners and say, "Hey, I'll run your social media for you," because they don't know how to do it—they know they should do it—and you can bridge that gap. That became the business in a box. Well, what is the business in a box today? It's going and selling, "Hey, I'm gonna use AI to improve your business," and you go do that. You should be doing that all around the country.
49:39	Sam Parr	Over the next couple weeks, when it's just you and I, I want to do an episode on how to develop good taste. What we're seeing online is people saying that building things is not important anymore because you have a machine that can build them for you. The next most important thing now is how to have good taste. I've found that piece of advice incredibly unhelpful. What is the definition of "good taste," and how do you develop it? I've been working really hard on defining it. I think there's an interesting opportunity to teach people how to develop good taste, because I believe you can. I want to do a whole episode on defining what "good taste" means, because I believe there's an objective definition: knowing and speaking the desired language of the people or thing you want to speak about — and then learning that language.
50:39	Shaan Puri	I'm all in. I love that. I think that would be super fun to do. And I think, like you said, it's timely and necessary. I also think most people would agree with the premise that taste matters. It's pretty hard to disagree that taste ever didn't matter — now it matters more, right? If AI can produce more stuff and makes it easier than ever to create things, well, taste matters even more. Taste is the discernment; it's the decision between good and bad. So yeah — obviously it matters more. But I think it's these other soft skills that people don't even think of as skills. Meaning, you just sort of assume you have a certain amount of it and that it can't really grow like any other skill. And it's like, "Well, no — if it's a skill, it means it can be practiced, it can be learned, it can be improved." And so, if it's a skill, then that means the only question is...
51:25	Sam Parr	I have that answer, and so I won't reveal it today. I've been thinking a lot about this — coincidentally, for about two years — about how to learn good taste, because I don't think I particularly have good taste, but I've tried to develop it.
51:40	Shaan Puri	Alright—down. Alright. That's it. That's the AI news we wanted to react to today. Sam, send them off.</FormattedResponse>
51:46	Sam Parr	Peace. That's it. That's the pod.

Discover

Listen

Catalog

Guides

Connect

This guy cured his dog’s cancer with ChatGPT + 4 other crazy AI stories

Transcript