#58: Greg Isenberg on AI Co-Founders, the Multi-Preneur, and Why Anyone Can Build a Startup Now

0:00So I believe that this like builder distributor combo is is probably the most lethal type of person in the AI age because they they understand distribution. They understand how to capture attention and how to convert that attention [music] into uh revenue but they also understand how they [music] can drive value from a product perspective. The product has to be good and the [music] bar for good products is increasing and increasing especially when it comes to digital products

0:42this episode is brought to you by Google for startups. Today we're in Aventura, [music] Florida with Greg Eisenberg, one of the sharpest voices helping builders understand the next great [music] shift in startups, AI, and the internet. Greg's career has been built on one rare skill, seeing where the internet is moving before the market catches [music] up. He created Five, which was acquired by Stumble Upon, and he built islands, which was acquired [music] by Weiwork. That pattern recognition took him from founder to trusted adviser where he's met with [music] some of the biggest and best companies in the world helping them on their strategies around AI. Today he

1:17leads late checkout portfolio of AI native businesses powered by agents and internet communities through idea browser Greg's letter and the startup idea podcast. [music] He shared over 2,000 startup ideas and built an audience of millions [music] of followers. Hope you enjoy our conversation. We're in July 2026. What are the biggest shifts you're seeing in AI right now at this moment? I couldn't be more excited with like what's happening right now? So, I just spent all morning playing with Grock 4.5 and I sort of I dismissed Grock. I was just like, "Okay, it's an open AI anthropic Google world." Like, I didn't I've never even thought about

2:02Grock. But I started playing with this model. I connected it to my Hermes agent which I mean we can talk about this sorts of stuff but I had this basic I gave it to my personal assistant and I had it basically running 247. I said here's a startup idea. I want you to go and, you know, build features for it, optimize based on analytics, do SEO, um, and do all these things and run while I was sleeping basically and in the morning this morning. And seeing how capable it was

2:36and how good it was. I mean, it's incredible. Like, I'm sitting here. We're at the 38th floor looking at the ocean and I have a essentially AI co-founder that's working. I'm paying it like the equivalent of like $2 an hour or something and it's super intelligent and it's creating a startup for me while we're hanging out.

3:04Mhm. I mean, it's hard not to be extremely excited about building companies right now with AI. If I'm trying to get started, a lot of people talk about the idea and then there's also going after a business model that's already working. Where do you like to start or where do you advise people to start? Do I start with coming up with the right idea or do I should I look at the world and see what business models are working and try to iterate off of something that's already out there? I mean, I'm a big trend guy. So, I like to see how things are changing in the world

3:41and based on that go and create something to feed that problem. The beauty about the world we live in today is all the trend data is out there. You can see what people are tweeting about. You can see what posts are going viral. you can see what the search volume is. You can see what people are putting into

3:58chat GPT and what the LLM visibility is. So, if you're able to harness that data and basically have your agents tell you, hey, there's an opportunity to go and build um you know, a software company for South Florida dentists and it probably looks like this and it has these features, then I'm going to go and do that because I've got a validated

4:25startup idea. Mhm. You know, when I was younger, I would like how I came up with startup ideas was would be like I don't know, you have a glass of wine, you go for a walk on the beach, and then and you kind of wait for this perfect moment and you're like, "Oh my god, I have this idea." And you write it down and you tell all your friends and and you fall in love with the startup idea. And as I've gotten older, I've realized that the businesses that have worked the best for me have been ones that have been validated based on data. Not validated based on like how

4:58I was feeling at 6:32 p.m. I had the wine was hitting me at a particular time. The sun was, you know, hitting me at a particular time and it just something hit like that. No, that's that's one way to do it. But if you are if your goal is to create a startup with the highest probability of success in the age of AI, I think the way to do it is to look at

5:21the trend data. So talk me through how you do that. So you're on you're on X. You're reading tweets post from the right people. You're what what else are you doing besides that? I mean, one way to do it is is exactly what you said, which is basically like you have a content diet.

5:41Yeah. You follow certain people on Instagram, you follow certain subreddits, um you you follow certain ex accounts, and based on what people are saying, you're able to deduce. You're kind of triangulating based on all these things. It's pretty manual, but you can do that, and it it works some sometimes. The second way is you connect a and this is you know we're getting into like a little technical but you connect an MCP to a codeex or a cloud code. Now what is an MCP? You know it's a fancy way of saying basically like an API or just plugging into another repository of data that is going to help you make that

6:25decision. For example, one MCP is the XMCP. So you can basically say you're using cloud code or codeex which are you know by codeex is by openai cloud code is by anthropic and they're basically this v this way to interact with those platforms. You can basically have it connect to X.com's MCP and you can say like you know every single day you know here's my niche give me trending you know tweets in these particular uh verticals and then you can have your LLM which is the super intelligence basically refine those ideas for you curate those ideas for you extract the pain points for those ideas for you give

7:11you new ideas based on on that and have it just send you a digest every single day. Um, and there's other MCPs you can use. You know, you can use I created my own MCP around startup ideas called idea browser. So, there's a bunch of MCPs that you people can just tap into that. Yeah, people can just tap into that. So, those are the two ways I would say like more manual first or more LLM first. when you talk and listen and what you're hearing people who are building trying to build a business who are trying to use AI what do you think most people are missing

7:46with the opportunity with AI I think a lot of people I mean it's July 2026 we've talked about it like if you're listening to this you've probably used AI sure but you probably have gotten some bad results from AI so the truth is AI I is only as good as the context you feed it. So what I think I think there's a lot of people that will have listened to this and be like, "Yeah, I tried AI to build a landing page and it looked ugly." Or, "I tried AI to come up with a startup idea and it was dumb." Or, "I tried AI to write a post and content and

8:27it sounded like AI." And that is because number one, you didn't give it the right context. You didn't connect it to the right MCP. You didn't say, "Hey, this is how I like to write and here's my framework for thinking about how I write or how I come up with startup ideas." And number two, you haven't shown them what does good

8:50look like. Mhm. So, when you hire a new employee, let's say it's a junior employee and maybe they're a social media manager, and their job is to write tweets and reply to tweets. If you're hiring this person, they're fresh out of school, for example, you might want to tell them, "Hey, this is what a super good tweet looks like.

9:14Here's what a good reply looks like." Yet, we we don't say that with LLMs for the most part. So the people who crush it with how to build with AI understand this context game and they understand this what does good look like game which in technical terms is called an eval which is means an evaluation. An eval basically says this is what good looks like and before you give me a result check or

9:41evaluate based on what I consider good. Yeah. such that uh like don't hand me something if it's going to be bad. Basically, this episode is brought to you by Google for startups. If you're building applications with generative media or considering it, Google just released a brand new technical guide on how to do it right. It's no longer just about generating a cool video. This is real guidance on how to build, orchestrate, and scale dynamic media systems in production using deep mind models like Gemini Omni and Veil. You can download the guide at go.gle/gen

10:20media guide. So recently I set up my whole workflow in co-work. I've got multiple inboxes. I've got all these social media accounts flowing in all of you know all this context around how I like to send emails and do emails. I've got all of my meetings flowing in from Zoom and other places. The whole thing took me like six hours to get set up and it it wasn't the most technical process, but it was still kind of a technical process. And I think when you say people aren't giving this their AI systems the context in some ways we just expect these things to happen because it seems

11:01like a some of it seems kind of obvious like don't send me a bad email. It's like you're exactly right like what is a bad email? You need to define that. But we kind of we kind of know what at least is a really bad email and sometimes, you know, it's it's spitting out really bad results. Do you think how far are we from that sort of s I mean slightly technical process being totally evaporated and just like the system's doing it or do you think like this is the way that if you're going to want to use AI properly in two, four, five years from now, you're still

11:34going to have to do all these kind of manual setups to like dial in your system. Is this just some way some thing you're just going to have to learn how to do and be great at and continue to refine over time or is it just going to figure it out? There's two types of work that exists. Yeah. Regular work and creative work. When it comes to regular work, for example, let's say SEO, you know, in the past, people have hired SEO agencies to do SEO. And what that meant was uh you had to create blog posts and you had to optimize the blog posts and you

12:10had to do a certain tasks certain set of tasks that were pretty binary because the result meant either you went up in Google, you went down in Google or you stayed the same. those tasks, those that work, those we're probably 12 to 18 months away from that being like pretty autonomous and is just going to do its own thing. The creative task like what we were talking about with tweets and maybe creating video content or branding or copywriting, things like that, there's some creative element that exists. when creativity is a part of it, even if it's just 1%. Um, there needs to be some human intervention. So, I think that

13:01from creative task perspective, we're never going to get to the place where this super intelligent thing is going reading your mind, reading your mind. There's always going to be place for art in society. And you can the counterargument to what I'm saying is is this. You know, you could say, well, if it's plugged into the human API, aka X and Instagram, like if you're getting all these posts from, let's say, MCPs and you're able to deduce what human beings want, can't it create creative on the fly? Um, and the thing about humans is we're we're kind of at least creative humans, we're kind of like punk rock and it's hard to pro

13:46it's hard to program that like punk rock nature to it. So I believe that and I'm a AI pled as it gets. Um, and I I advise the world's largest brands on the planet around, you know, getting their businesses to be AI native. So, I would be the first to tell you, hey, this is all going to be autonomous and creatives aren't going to exist. I see these people tweet, you know, you need to escape the permanent underclass and it's like I joke, I laugh at it cuz it's like there will always be space for creatives. Human beings will always be a part of it. That being said, the binary

14:29stuff will be autonomous for the most part. when you're talking to an established company or a team who's been working in a certain way for 5, 10, 20, 50 years and now all of a sudden it's like this reinvention of work. What do you tell people to start to really make a significant shift in that way? Not just, hey, sign up for Chat GBT, but like change the way you work in an environment, in a culture that has been doing it a different way for a long time. It always starts with like, what would you do if you unlocked five or 10 or 15 hours of your week back? You know,

15:08like you've got all these jobs to be done that you've been doing and what if you can automate 60% of it? What are the things that you would wish you can do that you don't have time to be doing today? And that's you you actually see people's face light up when when they hear that because they're like, you know what, I always wish I can do this thing, but I just find Yeah. Oh my god, I can't believe it's Friday already. I guess I'll do it next week and then it's

15:35Friday again and then it's Friday again. They never get to do it. So, I think what we're going to I think what we're there's this narrative that like all jobs are going away. Uh AI is eating our jobs. I think we're actually moving into a world where people are going to be more creative. And AI is going to take jobs away from the the stuff that's going to be extremely binary. So for example, I studied computer science in in university. So I dropped out, but I'm a trained engineer. Engineers now have this identity crisis because they're not programming anymore. I saw that Boris the guy who created claude code

16:15literally says that he hasn't like de touched he hasn't he hasn't created a line of code in like six months. So you have one of the most well-known engineers of our time not writing code. So what do you do when your identity is engineering and you can't write code? You sh you shift into a more creative aspect of engineering. Now he's reviewing code. Now he's designing systems. Now he's going on podcasts and he's talking about engineering and he's talking about cloud code. So going back to your question, um what do I what do I tell people who you know who have human workflows that are now moving to AI? You're about to do

17:01you're probably going to do a bunch more creative work. And if you're if you love creative work, you're going to love your day-to-day more. So in an age where in an age of abundance where so many more things and skills are accessible to so many more people what becomes the competitive advantage your creativity. Yeah. The most creative people win we're talk we're talking generally. Of course some people unfortunately are going to lose their job and aren't going to be creative like they're they're not going to be Boris in terms of like being

17:39Yeah. Not everyone can go on and talk. We can't just have engineering podcast from every engineer. No, no. But I do think that the ones who make the shift to being creative and learn to be creative are are going to love their their day-to-days bec I I think because a lot of people don't like a lot of the minutia that they do on a day-to-day basis. And that's sort of the unspoken truth about white collar work in general. like a lot of people are pushing buttons that if they were like, you know what, if I didn't have to push these buttons, I think I would actually

18:11enjoy it. Do you think that for somebody that wants to work for themselves, is the future going to be where they have all these agents building lots of different tools? Like, am I going to have 10 different ways that I'm making money and it's maybe $1,000 from each or am I like we've heard for so long in Silicon Valley, it's like focus, focus, focus. Put all your energy into one thing and let that compound. Is it going to be all the agents working in one thing or or is it going to be like, hey, have lots of ideas, have lots of things making a little bit of money for you and

18:49that's going to kind of build over time. I think that's the other part about this big shift around AI, which is this is going to be the golden era for entrepreneurship. If you look back, you know, hundreds of years ago, uh before the industrial revolution and things like that, uh people actually were more often entrepreneurs than not. This idea of you go and work for, you know, factory Henry Ford and and stuff like that is actually a relatively new concept. entrepreneurship, I think, because it it really does celebrate freedom and this idea of like going out there, forging your own path is a just such an incredible concept. Um, and I

19:30think that we're kind of like all that's new is old type thing. We're kind of going back to this era that we're going to have millions of people who are going to embrace AI. They might they might have been forcibly have to embrace it in the sense that maybe they get laid off from their job. Maybe they, you know, find an idea that they really like and they just can't stop thinking about it and they go do it. and that they're going to have multiple uh streams of income in the same way that if you look at the average creator who's making money uh today I often see

20:04like Tik Toks are short for uh reals where it's like let me tell you how I make $75,000 a month uh as a creator and like hooks you in. I'm like how does this person make $75,000? Then I go and it's like well I made $800 a month in brand in uh in affiliates. I made $6,000 a month in uh X uh revenue share. I made

20:28I think that's 60,000 and only fans. Yeah, 60,000 and only fan. No, it's usually 60,000 of brand sponsorships. Okay. I believe that we're entering what I call the age of the multipreneur. So a multipreneur is someone who owns multiple businesses, multiple lines of revenue. I will say I think that you know like creators the multipreneur is going to have one product one service one business unit that you know drives the lion share of revenue but I do think that they're going to have multiple businesses and if you can have agents running it and you can create a high margin business and a whole system

21:05around it um I think that's pretty cool. You talk a lot about boring ideas and uh trying to go after things other people aren't thinking about. How how do I go about finding and identifying a boring idea worth spending a lot of time on fixing? I like boring ideas because I don't like so much competition. And the context here is I'm not talking about you know you're a Palo Alto guy now you know Silicon Valley you're thinking big. So, I'm talking about the ideas that maybe you can do a million a year, $5 million a year, $25 million a year, which for a lot of people is

21:46incredible. Yeah. Um especially if you make it a AI native businesses and you end up having a $20 million a year business spin off $5 million a year and you can sell that business for $20 million very easily, you know, at a 4x multiple. So, um I like boring businesses. I think there's not that much competition in it. Um that being said, um there's been increasingly amount of competition in boring businesses because people like me have been like talking about it,

22:24talking about it. Um I think though where there's always going to be opportunity is a boring business in your geography. So I think geograph you know there might be a lot of uh competition in South Florida or in the Bay Area or New York or whatever but I think in you know where I'm from in Canada or in you know other geographies where you can take these concepts localize the language that's another big

22:50piece of it. Um you know there's this there's a story um that I really like. So, I'm from Quebec, and in in Quebec, it is one one of the places where people drink the most amount of Pepsi on the planet. So, why do Quebecers drink Pepsi instead of Coke? And they did some research to understand why. And it has to do with marketing in a local language to people in a way that, you know, really resonates with them. So in the 1960s or something when KO came to Canada, they basically did all the ads across Canada or or when KO started marketing with television in Canada,

23:34they did all the ads in Canada and they were all the same. And when they went to Quebec, they speak French in Quebec. They just translated the the ads into into French. And it's a different culture. And what works in Toronto might not work in Montreal. So what Pepsi did was different. They hired a local agency to do it and they basically said we're gonna, you know, we're just going to do something specific to Quebec that's really going to resonate with them. And it's almost 65 years later, people to

24:08this day still drink Pepsi. So going back to in the age of AI, boring businesses. I believe that you know local localization marketing towards different geographies but being AI native there's a lot of opportunity there and localization is one of the first killer AI uh use cases that it solved. I felt like like we within my startup we had spent hundreds of thousands of dollars on building localization tools pre AAI and we just rebuilt them and in a matter of like 3 weeks and it is like in and it's all like translated in real time like it's incredible and I mean we had really smart people that were working on

24:54it before. It's like this is something that the the tools solve really well. It's a really hard problem that the tools like can really automate. You've talked about this idea of a builder distributor. What is that? Oh, well in the hai anyone can build, right? Like all if you can you don't even need I was going to say if you can speak English, you can build. But you don't even need to speak English. You can speak any language and you just prompt to build. How do you compete in that world? Well, the way to compete is if

25:23you understand distribution quite well. So if you understand how to create scroll stopping Instagram posts or videos that people like or um you understand how to use cling which is a video model to create like really interesting commercials um and then use that to uh connect that with meta ads and and start you know pumping ads then you have an advantage. So I believe that this like builder distributor combo is is probably the most lethal type of person in the AI age because they they understand distribution. They understand how to capture attention and how to convert that attention into uh revenue but they also understand how they can

26:13drive value from a product perspective. The product has to be good and the bar for good products is increasing and increasing especially when it comes to digital products. So you have to be able to take feedback, optimize, create features, have that loop. And so I'm a big believer in that type of person. Are the AI tools to the point where they can help you build the product and then if you feed feedback into them, it can sort of create the product and the features that your customers and users are asking for and then like deploy them and then get more feedback. like is that loop

26:51solved yet or is it still coming? This is a bit like in the minutia but we're going to go there. There's a term that actually uh Andre Carpathy and Peter Stein uh I think his name's uh Peter Stein Brener I think uh from OpenClaw the guy who started OpenClaw they've been talking about this concept called loop engineering and it's just basically this way to tell your LLM to just keep going and building on a loop and Twitter just went like completely completely uh just love this concept. To me that concept is it's interesting but it's way more interesting for business use cases not engineering use cases. So loops are

27:39actually really interesting in the context to what you were talking about. Say you're able to say you started a company you have a product and every day you can speak to 10 customers. Well, in the AI world, you can actually speak to these 10 customers, have meeting notes that get synced to, let's say, uh, Google Docs or Google Drive folder, and then you can set up a loop that says, look at this, look at every day, look at this docs folder, then uh, that customer feedback docs folder. see what people like about the product, see what people don't like about the product, and every day pick the most interesting

28:18problem to solve and create a feature so that I can review it and it'll go and do that on a loop every single day. Now, in 5 years or 3 years or two years, we might get to a point where you don't even need a human in the loop. You might have this autonomous company builder type of experience where you basically just you basically say like your goal is to generate revenue uh or your goal is to create retention or your goal is to uh drive virality. Go look at customer conversations every single day and go pick one feature to go and build these people. By the way,

28:58isn't that crazy? It's crazy. Like, isn't that isn't all of this crazy? Like, we we're all so numb to all of this. But like, I'm basically telling you that you can essentially press a button that every single day you're going to be able to drive value for customers. And all you're going to have to do is interface with customers, help customers, get them to tell you their problems, and then on a loop, you're going to have the world's greatest super intelligence at a fraction of a price that you would pay a team. We're talking about a dollar an hour, $2 an hour, even

29:38less. And you can have it every single day drive value for customers. And all you're going to have to do is we're sit there from your high chair in the 38th floor looking at Miami Beach and you're just going to be like, you know what, that's a good feature. Good job, Hermes agent. Yeah. Go and deploy this. It deploys it and then you drive value to customers lives every single day. Where in the past it used to take you millions of dollars that you had to raise venture capital for. You had to hire a team. There was just as many bugs if not more bugs. And

30:11now you're driving value every single day and making the lives of your customers happy and healthier. And hopefully you can even give them a cheaper product cuz now your cost structure is a 20th of what it used to be. How how amazing is that? I a few months ago I sat down with my son who works at a basketball card store and they have no inventory and so like

30:40management they have no software for it. So we're like, "Oh, let's let's just play around with one of the vibe coding tools and and see what we can build." And in a matter of three hours, I just like leaned back in my chair and I was just I like I was like my mouth was like gaping open. I I couldn't believe it and my you know my teenage son is like what's the deal? And I was just like when I 15 years ago when I first quit my job and started to build a company this

31:10exact same process took me six months. It took me three months to find people to build this with me and then it took them two months to get a first version out and then it took another month for me to give them feedback and then get another version out. And so like and that's that's where we are in three hours and it was like 200 bucks in tokens or something. That's the other people that's the other thing that people don't talk about this AI age which I think is so important which is in the past you needed to have graduated Stanford went to Harvard lived in

31:46Silicon Valley in order to create a startup tech startup that can scale. Today you can be anywhere. You can be anyone. You can be 15 years old or 85 years old. And that to me is awesome, right? The fact that like when I wanted to start a tech startup, I had to move to San Francisco. I had to leave my family and friends and move to a different country across the continent to a place I knew no one. And I'm not, this isn't a boohoo story, but I to be

32:23honest would have rather stayed there. Yeah. Like, you know, but I had to go and do that and it was a fun experience and stuff like that, but at the same time, it's so cool that people are able just to wake up today and the limitation that people have is just their ideas. Yeah. Whereas in the past, there was all these societal limitations that were were placed. You needed a VC to basically be like, you know what? I like this. I like the school you went to. I like this product. I'm going to give you a million dollars. And you just needed the right VC at the right time to tell you that

32:59thing. And what kind of world is that, right? Like I hate that world because like to me that's a non-inclusive world. Yeah. They're they're gatekeepers. They're basically we're gatekeepers to uh to making dreams and ideas possible. Yeah. and and the gatekeepers now with AI, it's like democratized like who can do what because the

33:20barriers of entry are so much lower. Exactly. And I'm not saying like building a ventureback company is bad. I'm not saying like VCs are bad at all. You know, some of my best friends are actually VCs and and I've run ventureback companies and sold ventureback companies. So, I'm not saying uh it's a bad thing. What I am saying is it's no longer the only model for building these kind of companies. It is amazing that there's now this whole other set of opportunities that are now presented and you you know some people like pizza and some people like pasta you know and now the people who like

33:55pizza can go and eat some pizza and they can be whoever. Well, and the reality is the venturebacked model works for like less than 1% of all companies that are created. And of the companies that VCs, investors back, only like a very small percentage of the ones they back, and they're only backing like a small percentage of the ones they look at. So, we're talking about a very, very small group of people. Only a small percentage of the ones that they back actually work and return the money. And they know that their whole model's baked into that. So the vast majority of of anything anyone is building should not be that model

34:32like it because that model only works for this tiny fraction of people and companies and is having raised a bunch of money on a bunch of different companies learned the hard way that some of the things we've raised for should not have been venture back like we should have done this other model but like one of my companies the first time I raised was because this exact problem we're talking about like I had six engineers and I They had 300,000 in recurring revenue and the engineers were let's call them 100,000 each. So it's like the engineers are costing 600 grand. The recurring revenue is 300 grand. Like doesn't take a genius to

35:06figure out like the math doesn't work. It was like this golden anchor that was going to sink our entire ship. And in the middle of the night I had to be like I had to email anybody I could find or I'd ever talked to and say, "Hey, I think I got a cool concept. Could you back this?" But it probably was never going to be a billion-dollar outcome for anybody. In which case, if I could have bootstrapped it, I would have. But it was like, you know, just like the cost didn't make sense. Now today, if we'd had that same insight and we'd had maybe two or three of us

35:37engineering and building the product with these other things, supplementing it, like we could take it so much further and and maybe we could have, you know, you could you could bootstrap the whole thing. And that's how you build a real business, right? It's like it makes money like it it provides service customers like enough to keep paying for it or to pay more for it. They tell their friends, you know, that creates more momentum. You know, no VC is going

35:58to, you know, create customer traction. And you see, I thought when the social network movie came out in 2012, I was like, this is peak startups. Like I remember going to the movie and I walked out in the parking lot and I was like, that's it. Like this is like jump the shark. And then startups just exploded. But actually the the value of startups exploded and the the market size exploded but actually the number of startups getting created didn't explode but that is now that was like 2012 I think or 2011 now it's 15 years later and actually we see the like the app store um app submissions is like

36:34exploding people are like creating apps like crazy um you see the stripe numbers of new businesses that are creating and the number of businesses that are creating is exploding and I I mean, what else do you point it to other than like this age of abundance and opportunity that now has been unlocked with AI? Exactly. Like when I see the Stripe charts and stuff like that and the app store charts, it's exactly what I'm talking about around the age of the multi-reneur. It's like the cost to try an idea is now so small that if you're listening to this, you should probably try an idea. You know, like you should

37:11even if it's hey the goal of it of this endeavor is to get really good at cla code and that's my KPI or have a hundred just happy customers. Like you don't even need to do it to make money but just to like learn or you know it's sort of like playing an instrument. Like it's fun to learn how to play an instrument cuz once you master the instrument it's

37:34awesome. Yeah. And like I'll be honest, when you start playing an instrument, it sucks. It's you you're you're let's say you're trying to learn how to play guitar. You're listening to it and you're like, "This is horrible." Like I learned how to play guitar. It's like I used to hear myself playing guitar and I was like, "This is making my ears absolutely bleed." And you sort of bleed through how bad you are. And eventually you get to a point where you're playing like stairway to heaven and it's this magical moment. And that's how you're going to feel when you build an idea with AI. And that's kind of also what

38:09which what fires me up about people building not just to start a business but also just to learn. I have a friend in Dubai and he's really talented entrepreneur. He's raised some money for a startup, but he had this engineering team. Then he ran out of money and he couldn't raise more money and so he had to let his whole team go. And it's just him. This is like eight or nine months ago. And I spoke to him recently. I

38:35haven't talked to him in a few months. And he was like, he's like, "Man, eight or nine months ago, my whole team, I had to let them go and it was the worst thing ever." And I didn't know what I was going to do. So I just started learning how to claude code and he's like I've completely reinvented the whole thing. He's he's now this is like you know this is like you and you can do this too. He he he's like in the process of his company getting acquired for millions tens of millions of dollars. I mean this is this is says

39:08much more about him than anybody else. like I couldn't have done what he's done, but just that this is even possible. And he's like I'm like well when you get to this new organization like you can probably have a team. He's like I don't I don't want a team. Like I I'm doing everything. He's like okay like even everything is part of the sales process now. He's like hey all the docs for it they're all in this tool

39:30that I built. You can just go in here. He's like build his own Carter thing, you know? It's like everything now he's like so he's so self-sufficient. He's so self-reliant and like he can literally do anything. Nothing is stopping him. No one is blocking him. And anytime something comes up, he's like, "I'll fix that." Myself. And six or nine months

39:50ago, like he couldn't have said that. And I probably not everyone is as resilient as him. I certainly am not. But can we be 10 20% of that? Like yeah, you can you can be 10 or 20% of that. But like you said, like you just have to get out and do it. And I've interviewed people. One of my favorite questions asked now I was like tell me what tools you're using. Um, I had an engineer recently who asked for a big raise and

40:14wanting to be called in an AI engineer. And I said, "Screen share your cloud credits. Let me see how much you're using. Like, what are you doing?" And it wasn't super positive thing for him to show, which was not to make him feel bad, but it was to then say, "Hey, like let's improve. Let's change. What can we do? Let's talk in two weeks, let's talk in four weeks." Immediately, like that day, like he started using AI more. they started shipping more. Um, but I do still interview people and they say like I say like what tools are you using and it is shocking to me but they'll say

40:49nothing. Like I'm not I'm not using anything. I'm like you're not using anything. You haven't tried? No, I haven't tried anything. Well, what are you interesting about? What are you reading about? I haven't read I I've literally had this in interviews like I haven't read any I haven't seen anything that I'm interested. I'm like well then you're in a bad spot professionally because you've got your head in the sand. Meanwhile, there's all these other people like my friend Simon and others who are like changing the way they were completely reinventing the way they're they're building products and services and making money and like they're going to win. those people were clearly gonna

41:23win. And and so it's honestly it's not it's not even surprising to me that most people are still not really diving into the tools because the truth is it requires I'm not going to say 10,000 hours but like a thousand hours really that much. I think there's the, you know, thousand Eisenberg's thousandhour rule, you know. Yeah. Uh, you need to like really put yourself through stuff not working. Um, pushing stuff into the world, seeing how it works, you know, playing, reading about it, studying, go watching other people, all that stuff. It just takes time. And it's sort of like, you know, going back to like the guitar analogy, like how many

42:10people like music? Everyone. Everyone listens to music. How many people play an instrument like a guitar? I don't know 1% of people, 2% of people. Playing the instrument is difficult cuz like I like I was saying, you have to go through this trough of despair for a really long time before you end up feeling good about it. There's instant gratification when you use chat GBT. if you ask it a question and you get something back like you you sort of see the value and it's it's obvious but there isn't the same instant gratification that you get for becoming AI native which to me is you have a set

42:50of jobs to be done and set of workflows that you do as a human being and you've now basically said hey in a world where I'm managing 100 agents what does that look like if those agents are fulfilling the work and that actually is somewhat complex, somewhat tiresome to go through, somewhat difficult to really like fine-tune to a place where it's doing good work. Put yourself in this scenario for a minute. You're fired from your job. You're told you're not employable anymore. Your skills just aren't what they used to be. You've got 90 days to create something for yourself. It might not be totally sustaining you ever every

43:40every aspect of your life, but you've got to start generating some income in the next 3 months and you're starting from scratch today. Where would you begin? The first thing I would do is I I say this uh often which is you know marry mar mar mar mar mar mar mar mar mar mar mar mar mar mar mar mar mar mar mar mar marry the niche date the product. So I would pick a niche that I love and I feel like I have some unfair advantage in. Now someone listening to this is might be like well I don't have an unfair advantage in anything and it's

44:16like everyone has some advantage to someone else in the sense of you're from a specific neighborhood. You went to a specific school. You know these people you you um you're you're good at guitar. you're bad at, you know, you're there's something unique about you that that you your life experiences have led you to this point and you have that that so I would figure out what my step one would be figure out what my niche

44:47is. Step two is I have this framework I call the ACP framework, audience, community, product. And I believe that if you're building a business in the AI age, you want to build a business that follows that framework. Every business I starts as a X account. It starts as an Instagram account. It starts as a Tik Tok account. It starts as an audience first. Because in the AI age when you can build anything, the mode is in a lot of ways the distribution. So what I would do for the first 90 days is I would make it a goal to figure out what is the format that is

45:26going to get my niche to follow me. So on my Instagram for example, it's like follow Greg Eisberg for startup ideas in the AI age. That's what like my bio says. So like that's the that's my niche, right? What is what is your niche? what you know that's the that that's what I would be thinking about and then a community which which is once you have your audience you know how do you convert a small amount of them into an IRL or digital community so that would be my second step and my third step is once I understand this audience once I I've gotten a few

46:07hundred followers a few thousand followers a few people in my community a WhatsApp group you know a local meetup then it's what is the product or service I build for them. And that's not going to take long. Like the building is only going to take a day or something, right? And then it's just, you know, for me, it's prioritizing what's a business model that makes sense that I think people really need that's a a painkiller, not a vitamin that I think I can sell in a relatively uh short cycle. And that's what I would do. And I would wrap that around how would I come up with the idea? It would

46:43be a mix of uh what we talked about around asking customers what they want. It would be a mix around bringing in MCPs like the XMCP and the idea browser MCP and just trying a bunch of different products to see what what could work. What failure or apparent failure led to a big success for you? The first thing that comes to mind honestly was I basically I lived I lived in this little neighborhood. I think I must have been like I don't know 7 years old and I was really into baseball and uh I was really good at baseball honestly especially cuz I was in Canada so like

47:28no one's good at baseball you know not that many people are good it's not like in the states you know like people people played hockey right yeah so anyways I was pretty good at baseball um I tried out for the team and I played amazing like the try out I just like I think Hey, you know, I don't know. It was like I was at bat four times. I think I got like three triples and a single, you know, and I had a smile on my face the whole time. I was It was good teamwork. I, you know, I left the try out being like, "Crushed

47:56it. Crushed it. This was amazing. I'm so excited." And it was like the first year where we get our own jerseys. And I just like pictured myself in these jerseys. Walk into school the next day just like with a little bit of swagger in my step. Then I get a call that night from the coach and he was like, "Sorry, you

48:18didn't make the team." And when I found out, I was like, "I didn't make the team. Like, who took my spot?" Turns out it was the coach's son made the team for the shortstop spot. And he was like a year younger and also just like a very he was a year younger. when you're six and seven, that's like a

48:42big deal. Sure. I learned a very valuable lesson that day, which was the world is unfair. Um, sometimes you could hit really well, do really well in life, do all the right things, but sometimes just the way the cookie crumbles is some gatekeeper somewhere does something, some VC doesn't want to fund it even though they said they'd fund it. You know, there's a million things that could happen in life and that are out of your control. And that was such a monumental formative experience for me. What I learned from it was was that was basically just like being okay with the fact that not everything is in your control. What's a

49:31lesson in life that's been hardest for you to learn? Not everything in life is is in my control. I mean, I think as human beings, we want to control things, right? Like what I love about using a computer is that I can control it in the sense I can tell it to do it certain things and part of that makes me in the flow state of like I feel relaxed when I when I do it. I think human beings want to feel relaxed. We want to there's a certain amount of pressure that exists when we are in a volatile situation and I think we all want to be

50:08away from that. So you know I always say it's kind of like in some ways you want to live your life you want to live a light life and in that sense like you don't want to feel like you have all these like dependencies and external pressures and stuff like that. So that's been hard hard to deal with. At the same time, you also want to fulfill your potential. You know, you don't want to do you don't want to not do things because you're like, you know what, that's going to invoke a whole bunch of external dependencies and like I don't want to deal with like if that goes wrong. So

50:44dealing with those two things has been something that I I spend a lot of my time thinking about. What's a limiting belief that you still wrestle with today? In my mind, I have like two sets of inspirations. So, I have like the the Silicon Valley greats, like the people who built Silicon Valley, the people the biggest billionaires on the planet, the biggest founders on the planet, the people doing big things, the Jensen Huangs, the Mark Zuckerbergs, those types of people. And the other end of the spectrum, I have like my, you know, my grandfather who had like, you know, a small business that, but he he he he led his life on um

51:24his own terms. I you know I don't know if it's a limiting belief being like I want to own a bunch you know a bunch of small businesses but like I I I just what I deal with and I'm not really answering your question but I feel like the struggle I have is deciding where on the spectrum I want to

51:43be. So is that a limiting belief? Perhaps you could say that my limiting belief is I'm not thinking big enough. Um, that being said, I live a life that I love and that I I think Virgil Abau uh used to say like I live a life that makes your 16-year-old proud. Like my 16-year-old self is I think would be extremely proud of what I've accomplished and what I what I do on a dayto-day. So, it's kind of like I don't know where that's where do you feel

52:15fulfilled? Yeah. Yeah. There's a study that said that like vast majority of people don't live within their means. They don't live within even rich people like really really rich people who don't like a large percentage of really wealthy people still don't live within this incredible life they created because they're always trying to get more.

52:35They're always trying to go for more. And I I mean I struggle with this too. And I mean I look I live in Northern California. So you know what is enough? And then what what am I like what's right for me? And that's that's only a question that I can answer. and what's right for Mark Zuckerberg who I used to live down the street from and I was renting but he he owned he owns like eight homes there. Um but I mean that was like a very real every day would drive home and I would drive past his house and his security guard and be like

53:08we live on the same street. I we sleep we both sleep on a bed like we're both breathing the same air but you know is it enough or do I have to go be that? And I think that's that's one of these existential questions of life is just coming to terms with who you are, who you uniquely are and then like being

53:27comfortable with that. Yeah. And coming to terms with that. Yeah. And if it's Grant Cardone, it's like it's always like, you know, in our interview, we spent a lot of time talking about the yacht. He wants he has he's like gunned for the question is like could you just rent I mean you he could afford to rent the yacht but he's like really got to own it and like well he's at the other end of the spectrum right he's he's trying to do big things and a part of that is like I don't know him or or know him well but my my what I've gathered from people who do

54:03those big things is they there's a certain lifestyle attached to it. It's like I think it's like that's what's right for him and actually like the answer is like you could say that's the wrong thing for him to go after but actually I think for him it's the right thing for him to go after because that's

54:18how he's wired. It's not really how maybe I'm wired or you're wired maybe. And you know it's sort of like no one knows except Grant. Right. Right. So if he's if he's being if he's being honest I mean I guess if he's being authentic to what he's saying like like and I think he is I think he's very authentic to who he is like he is the guy that you see and I've you've watched and that I've listened to his books like I don't know I spent two hours with him he seems to be that guy I can only speak to myself and what I've learned is the

54:51most honest version of myself is the version that you see today when I raised venture capital and I was doing the Silicon Valley thing. I was doing it because I thought that that's what I needed to be in order to be successful. And then I realized that I don't I don't need I can be whoever I think is interesting, cool to myself, even if it's uncool, you know, I could just do things that I think are interesting to me and make a living doing that. And and that's what

55:23makes me happy. When someone says your name in 50 years, what do you hope they say next? I would say he helped me. That's it. You know, if I can help people, I want to serve people. Like I always say like I live to serve. And you know, on my YouTube channel or on my podcast, like when people comment stuff and they're always like, "Thank you so much. Like I never knew this like tool existed or this idea like changed my life. I found an idea off your

55:49podcast. Now I'm doing 100k a month." And I usually just respond like my default is I live to serve. I think that if human beings live to serve, you know, we live in such a polarizing negative place that in some ways that if we can live to serve, I think the world hopefully would be a little a little bit better. Outside of family, who's the person you most

56:13admire? I don't know if I admire anyone. I have respect for a lot of people and I'm inspired by a lot of people. admire like that feels like like it's a muse, you know what I mean? Like I have someone on a pedestal. Yeah. Um I I I wouldn't I wouldn't say I have

56:31anyone. If you could implant one idea into everyone in the world, what would it be? Live to serve. Live to serve. If you can if you can find something some like we talked about, you know, marry the niche, date the product. If you can find a niche group of people that you really care about and you can serve them and add value to their lives,

56:55to me that's a life worth living. Hardest thing you've ever accomplished? The hardest thing I ever accomplished was is just being truthful to myself. Like not getting We live in such a a world where there's so much tension over here, tension over there. Come here, do this, do that. And I find it difficult to like put blockers on. Like you you I'm sure you've seen pictures or seen horses that have like blinders on and they just like go run. Like to me that's been the hardest in life to basically be like,

57:29"Okay, I have these three goals in life. Be an incredible father and husband. uh say um you know serve 10 million people, 10 10 million entrepreneurs by giving them ideas and tactics and stuff like that and give back to my community um from like a philan philanthropic perspective. I just want to do those things, you know what I mean? And sometimes life has this way of being like you should go over here, you should go over here and I'm trying my best to stay focused. Best advice for overcoming hard things. the people that love you the most. And it's interesting because the people that love you the most, you

58:09go to them the least for advice, for example, like your parents, your aunts and uncles, your cousins, the people you grow up with, you know, I think I wouldn't discount those people's advice to help you get you to where you want to go. It's really smart. Actually, a close friend of mine told me that one time that my dad had told him that he didn't think I asked for his advice very much. And I think it wasn't something naturally like just I done on purpose. I just hadn't thought of it. And so I've started doing that a

58:46lot more since they made that comment. And I've it's been really helpful that I mean it's my answer is the nonviral, non-clippable answer that I think is truthful. The answer that I could have given you that would like go viral would be like the people you grew up with just because you live there, you know, they they don't love you and you know your parents choose your own family and burn your bridges and and it's like picking where you are on the spectrum, who you truly are. Like the people you grew up with are underrated and that's my take on that. And that doesn't mean you can't have people

59:25uh you know you can't create a spectrum of people you get advice from, people who love you the most and people who you admire or look up to. But I think that we tend to model our lives after like the aspirational people. And we don't tend to model our lives around the people who love us the most. If you die and you meet God, what do you hope he

59:46says to you? You did a good job. And uh and how do you measure your life? Probably by what my family says about me, you know, if if I mean, I'm I'm a I'm a father now. I'm a year and a half old and and assume to have another one uh and a mom. So yeah, you know, as you know, being a dad like completely changes your how you think about things and you really just care about like what your your children think about you and your your partner thinks about you and what your family thinks about you and that becomes so important. Your last dying [music]

1:00:28thought won't be, man, wonder what that Sand Hill Road BC thought of me in that pitch. Well, my last dying thought, I'll have agents running a loop anyways. So, I'll be, you know, hope I'll be making money when I'm dead. [laughter] Hey, appreciate you, man. Thanks for having me.

1:00:53Give it