$100B Founder Breaks Down The Biggest AI Business Opportunities For 2025

That's the trend that I can't, I can't like unsee it sometimes.

Right. That's the bet you're making.

That's the bet we make. And so I always joke with the guys here, it's like we want to bet on -1 to 0. Like Peter Thiel talked about these 0 to 1 companies. It's like we still want to step even before that.

All right, Furqan, we're here.

It's amazing.

I haven't seen you in a little while. People don't know we used to work together maybe 5, 6 years. We co-founded a company and launched a bunch of products and ate a bunch of shit together. And here we are.

Fun times, by the way.

Yeah. I'm thinking the title of this is going to be because AppLovin is now a $50, $60 billion company.

Yeah.

$50 billion founder tells me the next big thing in AI. I'm going to go full YouTube clickbait with it. Full send. Yeah, yeah, yeah.

Great.

We— so we used to do this thing where after We did Bevo together, got acquired. We were there for a year or so. And then we all, we went off and we did different things. I started the podcast and started doing my thing. You started doing yours, but I hit you up and I was like, hey, miss hanging out with you. What if we did something? And we started doing this on Wednesdays, which was like the cool shit hour. And this is amazing. It's basically a show and tell where you, my smartest friend, would come on Wednesdays and you'd be like, hey, have you seen this? Have you seen this? Have you seen this? Have you seen this? And I really hadn't seen any of it. And you would kind of explain it and teach it to me. And it was my favorite part of the week. And we did that for, I don't know, probably like a year or something like that. So I kind of want this to be like a public version of the Cool Shit Hour where you're just going to tell me a bunch of good things. I want to start with AI because you texted me something. You said AI agents are here. And you said AI is cool because a 10-person company feels like it could do the work of a 100-person company and we're using it in our company. So I wanted to hear What are you doing with AI? And that's not called ChatGPT.

I think this word AI agent or this kind of phrase is the thing that I really can't pull myself away from, like literally every night. Like, you know me, like midnight strikes, I want to write code. Like the whole day is like talking to people and the night is like coding.

Yeah. Furqan's schedule is nocturnal. So I remember we, we hired you and I feel like the first day you came in at 9 a.m. to kind of like, I think I should—

It was probably 10, but yeah.

After that, from day 2 onwards, it was like, Roll in at 11. Oh, it's lunchtime. Have lunch. You would talk, you would do meetings. And I was like, when does this guy code? And then that night at like 4 a.m., you would have built the prototype and you're basically nocturnal. You get all your shit done. You like— you used to tell me you used to get during the day, you just burn up your energy so you could focus at night and actually write code.

Yeah.

So what are you doing with AI and how is that— what is that actually in your company right now?

Yeah. And so I'll kind of tell you how I'm thinking about AI agents, right? And what it means to me. So AI agents are using LLMs or AI systems, right? Like the OpenAI systems or Claude, but then giving it reasoning loops. So imagine that when you go to a human, you give them something to do like, hey, I want to go grow our company, want to do a marketing campaign. They take that, they plan it, they come up with the steps to plan, then they go one by one on the task and like solve them. They release some of them. And so these AI agent systems are exactly like that. The first thing it does, it goes, what do I need to do Based on your request, and it'll kind of come up with a plan.

You give it like a mission.

Correct. So let's say I'll give you an example of something we're doing at Third Web, which is every single signup in our company. And, you know, we have a lot of signups every week and, you know, a human can't really scan through them, but there's really interesting people signing up. They have interesting companies. We may know them. They might be a large company or a small one. It could be a competitor or something else. And so we used to have a human go and look at everything. It's like, oh, if it's Gmail, like ignore it. But if it's like, oh, like this other company over here, let's go research it, right? Go figure out what Third Web products they might like, what they might need to do. And then you would send them an email, try to customize it. Typically, this is good sales practice. You're taking your customers and delivering them to your business team. And so we built an agent to do this. And every signup that comes in, it looks at it, it determines, is it an interesting person or not? It will research their website, it'll research them. It'll then use the knowledge it has of Third Web products and try to figure out what it actually like, what products they might need, how they would use it, and then it would send them an email or an upsell or something like that. And we've deployed probably 8 or 10 of these throughout Third Web to do a lot of different functions. And what it feels like is a smaller company can punch above its weight. So like, we're like 37 people right now. I really believe we're more like 80 people. That's what it feels like with a lot of these tools. And You're taking the brainpower of somebody who gets it and you're giving them that power to go and say, here's the thing.

So in this case, person comes to your site, puts in their email address.

Now you have—

it's one agent or it's like a series of things that pass off to each other. How did— what is it called?

Yeah. So in this case, we call it the signup agent.

Okay.

So it's one agent. It creates a plan like, what do I do? And, you know, part of the plan is like the directive we gave it. And I think the way to think about AI agents in general is Any like clear directive problem. And what I think about clear directive is like, let's go do X. Like signup comes in, go look at the person, go research them, go figure out what's their title, what's their company, what products do they have. So that's like kind of clear what the job to do is. And if it's a digital task and a clear directive, all of it can be done with agents now. The technology is here.

Right.

It's ready and it's working. And so for the Third Web signup agent, It'll kind of come in, it'll make a little plan like, hey, I got to inspect the domain, I got to go look up the person.

And you didn't have to tell it each one of those little tasks to do.

We gave it like kind of like one, one paragraph directive and another paragraph of like how it should operate. And then maybe a paragraph at the end for like the type of email or the action to do.

And so, but the end product is it's looked up the person. It basically kind of looks at the signups, picks the interesting ones, researches them. You said thinks about what product of ours will suit their needs, which is— that's the wild step that I hadn't really thought about. And then it crafts an email and then it gives it to a human or just sends it.

Sends it.

Okay. So you have enough trust that, you know, you can send it to customers.

We started it with human in the loop and there are still like safeguards you put in, like any directive would. Like you don't want to just send random emails even as a human doing the work. But I mean, it's very clear what, what to tell it. And I think this is where the power is kind of multiplying. It's like we've heard, we've heard ChatGPT, If you can tune it to your problem. So for example, giving it the knowledge of third web products is the key difference here, right? A generic ChatGPT message, when I say that, it'll just invent something that's like general and generic or whatever it knows.

And when you built this, so you built this or somebody else built this on your team?

I hacked it as a prototype, like some things, and then I gave it to somebody on our solutions team and within a day they turned it around.

And when you made like a working version of this, how long did that take you? Because it's kind of like You basically hired an employee and trained them and got them working. You're not paying them a salary and you probably did that whole thing in what, like a day or two?

Yeah. So let's say there's like a bunch of like my nights of just learning tools and getting used to it. So I take that time out of it. That's just time I spend anyways. Right now, like 2 nights ago, I was like really stressed. Like my calendars are kind of crazy some of these days and I was trying to figure out like, why is it crazy? Where is it going? It's like a really hard question. to like ask, like, what do you do? You could have an EA. She can go look through it, do all this stuff. And I have an EA, she like helps me with these things, but then she sleeps for like a while. I'm like, you know, I want to go answer some of these questions. So I probably in about 15 minutes I connected my calendar and I connected a little interface where I could type to it. I started asking questions like, how many hours of meetings did I have last week? It was like 28 hours, like way too much. Where did the meeting go? What, what were they for? What were the purposes of it?

Right.

And then the next night I hooked it up probably about another 45 minutes or an hour. where I could tell it commands from my calendar, like go block out Monday for me or go find me like 9 hours of block time and just mark it block.

Oh, wow.

And this was like a few hours of work now that I know the tools, but it's like it just works 24/7 now, right? And so my calendar, my email, a few other things, I've started building these like personal agent or workflow type things because I know what I want to do every time. I know how to react or the decision-making I'm going to make. Can I just set that up so it works 24/7, 365? And, you know, it's just there always.

Right. Hey, let's take a quick break to talk about AI. We all know AI, big deal. Um, you see demos all the time of people doing really cool things, but as a business owner, sometimes it's hard to figure out, how do I actually use this? What do I actually do? I've been trying to use it across all my businesses, you know, things like making little prototype websites without needing to hire a coder or writing copy for our website. Or I give it a bunch of data and I ask it to analyze it for me. It's been kind of amazing. But the thing I always need is inspiration. I know the tool can do a lot, but it can almost do so much that I'm not really sure what I should actually be doing with it. And so that's why I think it's great that HubSpot has created a report where they surveyed 2,000 global marketing leaders and asked them what's separating the high growth and low growth businesses and what strategies they're using with AI in their business. You can grab these strategies and apply them to your own business for free. The link is in the description below. So if I wanted to build workflow like this. Okay, do I— you need to know how to code to be able to do this light coding. Like, give me like the bullet point version of how, how you built these. Like, where do you even build it? What tool do you use?

So there's coding tools. I think if you're a developer, there's like LangChain and AutoGen and Crew. These are like very popular, very cool. The OpenAI and Claude SDKs themselves are very, very powerful. But you're coding, you're writing a little system. It's probably like 1/10 of the code you would have had to write to do something, right? So that's already better for developers. If you're not a developer, there's a lot of tools. So like Leap is a company we built here in the studio. It lets you stitch together workflows. So an example is like you can trigger based on a Slack message. So let's say you have this Slack message where all your signups go to. So now it picks up that trigger. You can just put a little AI block and say, okay, take this email and do these tasks. Oh, you want to web scrape? Go do that.

Right.

Oh, you have a little conditional loop or like repeat itself 10 times? Go do that. And then you make a little another decisioning steps like 4 boxes. I'll put it back to another Slack channel.

Yeah.

So you come in with your signup channel, it does this research, it does all these things, it does kind of whatever you want. And then it could go to email, it could go to another Slack channel, ping somebody on your team. So you could kind of like, you know, these are workflows is what they start with. And then agents can kind of take these workflows and like almost build on top of them. And so I think there's these 2 things you could really, really easily create workflows. And I think everyone should be deploying them. Every company should have them. It is a superpower. It's like, I can't, you know, it's like a feeling of almost like cloud, like, oh, I don't need a whole data center team now. And I can—

Right.

You know, when we built Blab and like how many servers did we have running video streaming? Like that would have been a nightmare. It's kind of like the same, you know, 10x improvement, but just for me and everything that I do digitally.

Right.

And so tools like Leap are great. I think, you know, there's others out there that provide this. And so there's a combination. If you don't know how to code, You just have to think about the steps that you would do, and you can program that without writing any code. It's just writing directives. It's like writing intent, basically, of like what you want to accomplish. Right.

It's like a magic genie. Tell it what you want and it can figure it out. So you got a sales agent, you have your email calendar, kind of like EA agent there. Any other?

I got some fun frivolous ones that are kind of stupid. So like I set up a dynamic wallpaper. So literally like every 5 or 10 minutes it'll look at what I'm doing and like auto-generate me a specific—

Like your computer laptop?

My computer laptop wallpaper will modify itself to whatever. If it's night, it kind of— I told it I code at night, right? So it will go towards that direction. It knows I'm doing meetings, another thing during the day, or I'm talking to people, right? And it kind of comes up with cool stuff. It's totally frivolous, like not like a useful thing except my own like joy.

Yeah, that's cool.

But I find it awesome and You know, like there's cool scenes that come up, it invents new things. You know, Claude has created this new capability called computer use. I think that's the next area that AI agents are going to enter.

And by the way, so there's Claude, there's ChatGPT, there's Perplexity, there's all these different ones. Mentally, how do you bucket like the main AI tools? What, what's like the superpower of each, each one? Maybe just do, do those 3 or if there's a 4th.

Yeah. So like I think OpenAI and Claude, I think of them as like foundational tools. So they're general purpose, ask it anything, input/output, you stitch it together with things. And then they have tools like ChatGPT on top, or maybe Computer Use, that they're figuring out more general tools. I think Perplexity is really interesting. It's taking this general purpose LLM and the reasoning it could do and search. And like, I try not to do Google search anymore, mostly because it's slow and ineffective. And Perplexity really made it where It does a search, it reads the result as I would, it clicks into the links as I would, and then it tries to answer my question more purposefully, and it saves me 4 or 5 steps. So I think about Perplexity as taking something like search and then the LLM reasoning and combining them together in a flow that, that's more interesting for me. So ChatGPT or OpenAI tools don't have like real-time knowledge, but Perplexity, because of search, does.

Right. What's your, like, tangent? Furqan's hot take. Google. What happens to Google Search with all that you see now? And you're saying— I don't— you're saying I try not to Google Search anymore. That's pretty wild, right?

It just feels slow. And I know they got the Gemini thing there. I think the biggest fumble, you know, in our lifetime, maybe decade fumble.

Yeah.

They built the technology for transformers, the thing that OpenAI and others have used to develop this large language model They are the ones that were the first entry in like talking about it and throwing it out there and the research that's happening.

And so what's the history? So they're— they have the AI minds there. They write this research paper, Attention Is All You Need.

Yeah.

And now from what I understand, like there's all the authors' names, all of them are gone and basically started their own companies. Was it because nobody recognized the power of it there? Was it that they tried and Google's bureaucracy shut them? What's this? What's the actual story of why they didn't?

I don't know the internal story. I think early on Google was like the place you went to where you wanted to have the rocket ship moment in your life. The smartest people were here. They were taking the biggest challenges, and it just really felt like that place. I think Google as a company hasn't felt like that. I think the parent company and all the other things going on do still feel like that. I'm sure it's a mess in there. Who knows?

Right.

Like, but it just feels like the biggest fumble. And I know they're trying to play catch up with Gemini and awesome stuff is happening. But the developer mindshare and the attention has gone somewhere else.

Yeah.

And that is really hard to pull back when somebody else becomes a leader in it. And actually it feels like OpenAI, number one, that's, I think, clear. I think Anthropic is number 2. So then everyone else shows up is what it feels like to me, at least from people I'm seeing building or the technology they're using or the innovation that we're seeing from it.

So the Anthropic one, so basically they have Claude, which is kind of like ChatGPT, but they have a couple of cool things. One of them is this what they're calling computer use, which is basically you type in a thing and then you do this and it moves your mouse and it just does shit. That's basically the summary of it.

It's a combination of things. You could take workflows and like, hey, I'm doing this, click this, click that, click, do it over and over for me. It could analyze things. But like, I think the key thing is, and I think it just released, it's in beta. So like, like most of these things, it's like it kind of isn't going to be great now. Just trust me, it's going to be great. And that's what we've been seeing in general is like things just ramp. And so it's now going to enter your computer. It's kind of felt like it's been on the cloud only is where AI has been, and now it's going to show up like in the box that you're used to, which is your computer, your laptop.

Right.

And there's so much workflow that we do. And yes, every app will put AI in it and then your interface will also do that. And I think a lot of like the things that we're used to doing, like switching tabs and having all of these things, like they almost feel like, I don't know, like why? I should have infinity tabs open and some system should know about it. And if I ask it, bring it back up.

Right.

It's like a total need and we know we need that. And so it's kind of like infinite data, infinite knowledge and reasoning. And then, you know me, that's what the AI— that's where I think the bridge, when it crosses, it completely changes the equation. And the reason I put Anthropic number 2 is their models are crazy impressive. The new Claude Sonnet model is like, I feel like it was a step factor improvement over the previous ones, and it feels like the task that it kind of struggled with now is getting better. And both OpenAI and like Anthropic have been just like boom, boom, boom. And I don't know, everybody's heard of AI. They tried it. They might have tried it a year ago. They might have used it a little bit. But every 3 to 6 months there's just another step and another step. And that's, I think, the most interesting thing and why I can't pull myself away from it.

Right.

Like why every night this is a thing that gets me very, very excited. It's because the progress is still wild. People are going to say it's going to top out. It will. I think it's going to be absolutely impressive wherever it starts slowing down at.

Right.

And will completely change the way how we do any digital work.

Like, so you're, you're messing with it now, but you have a company, you've got an investment lab, you've got this whole place, this beautiful place that we're in right now. You've got a wife, you got like all this stuff in your life. If you were just 21 again or 22 again, where you're like, dude, I got nothing but time, bank account's empty, but so is the calendar, and this technology is out. What would you be building? What would you be like messing with? What's the like, kind of like you don't need to take over the world even, but just like what types of stuff would you build if it was like young hacker version of you?

Like right now I've started seeing agents that it can do a reasoning loop, it could have a directive, and it could have like actions it could perform. And then you combine, like, for example, an agent with Twitter, give it a Twitter account that it owns and controls. So its own distribution and conversational ability so humans can just interact with it. Humans are— and you give it, you give it like a digital bank account. And, you know, at Third Web, we're doing a lot of stuff around AI. We have a whole like AI toolkit that we're about to launch and is around this, which is these like digital things, like these agents, they're not going to have credit cards. It's weird. And the reason I say this is fun and interesting is because, you know, humans have done this very well. They've created megaphones for themselves and they've created businesses and they created payment rails around that. So they sell things, they do things, they sell services and all of these details. I think it would be something around that.

So you'd make like a social, like a Twitter account, you'd make like an AI influencer type of thing.

You know, AI influencer is like the first obvious thought it goes to. I think if I played in this space, I would be creating the digital equivalent of a company, which is A CEO of a thing, the ability for it to market and the ability for it to make money.

Okay.

And I don't think—

Not an influencer.

Not an influencer. I don't know what it would produce.

Like an AI dropshipper.

Like an AI dropshipper or like, yeah, it's the best FBA, Amazon, whatever.

What's that example you were telling me about where the somebody did a thing like this? They made a, they made a Twitter account, they gave it a wallet. Marc Andreessen gave it like $50,000 of crypto. Can you tell this story? So what is this?

Yeah. So there's this thing called Luna or Virtuals and they're a little platform to you know, have AI agents run. And I think what's cool is they did this kind of equivalent thing where it's like it has a crypto bank account and it has like access to Twitter. And you can—

Was this a marketing stunt by a company that does this, or—

I think this is exactly what the— what I'm describing is. It was frivolous fun. It was play, not like working backwards from giant thing. But I think this did some really powerful things. So there's literally like a live thing where you could watch its reasoning. As it's like pulling its tweets.

I'm on it right now. So it's terminal.virtuals.io. And is this basically the thought process of the bot, of the agent?

Correct. So this is like an agent. This is what an agent does, right? So if you think about it, starts with like a thing like high-level planning, current state of execution. It's like, I've done this so far.

What was the directive they gave it? What did they tell it to do?

I think they told it that you're kind of like a public influencer bot. You have access to crypto. It's a little bit more in the meme coin world of stuff. So it's like, again, kind of on that side of the puzzle. But today it could negotiate for tokens, it could buy and sell things, it could kind of operate together. And I think it's really cool that you could actually see it's like—

So this thing, this was, you know, 30 minutes ago. It says, current state of execution. I have attempted 10 tasks so far, 7 successes and 3 failures. My Twitter metrics show an average engagement on my last on my recent tweets and I've lost 3 followers. Observation_reflection. That's like the function.

Yeah.

Reflect. And it says, I've been engaging with my followers to build a personal connection, which is hilarious because it's an AI bot, with replies and quotes. However, I've also experienced some failures in replying to tweets due to invalid parameters. My research on whatever, blah, blah, blah. State of mind. I'm feeling a bit concerned. I'm feeling a bit concerned. That's crazy. I'm feeling a bit concerned about the loss of followers, but I'm also encouraged by the successes. And then it says, Plan reasoning. Given my current situation, I need to focus on building a relationship with my followers and increasing my visibility. Plan. And then it starts to say what it's going to do. This is wild.

Doesn't this sound like a human that would be sitting in some growth team somewhere thinking about how to grow your Twitter account? And, you know, you could give it directive, you could give it some direction, and you could let it compute against itself to compute plans, to reason. to observe behaviors, try to find patterns. These are all like human tendencies, like human behaviors that we do, right? Especially when we work. And I think this is one of those like perfect, like kind of views where like you could start thinking about, man, this is a digital-only task. It has Twitter and distribution marketing ability, and then it has some payment ability. It's like it's going to just continue, right? Like it's going to keep going and you could improve its directive, you could change its incentive, you could Do a few different things here. But I feel like we're going to get to a world where, you know, I think Sam Altman said this, which is like the one-person billion-dollar company. Like, this is happening. Like, we're already experiencing another 10x decrease in how many people you need and the abilities that they have. And I think it goes down to probably 1 or 2 or 3. And, you know, like, that is a, you know, total shift to everything, right, in terms of how we work, how companies are built, kind of in my mind.

Okay. Should we switch to hardware stuff or is there any other AI stuff that you think is worth checking out?

Yeah, I did want to tell you about this Oasis Decart. The, you know, there's like basically, you know, it's a very early experiment, but it's really like a game that's fully built in a generative AI model. And so they built a game like Minecraft and you could just describe a world and it makes you a Minecraft world like that. And like every step you take is like not a pre-programmed pixel, right? It's actually generating.

So normal game is GameMaker builds the map, it exists. You then get to run around in a predefined map. What this is, what you're saying is—

You describe whatever you want.

Yeah, you give it the idea, but then when the character runs on the fly, it's creating the map.

And you can kind of play this Minecraft game and you can see it's like crappy pixels, still like not great resolution. But you can move and it can do stuff. And like, whether it's video games that take a lot of effort and time to produce, or it's like, you know, content creation and videos and stuff like that, it's starting to become like way closer to like, you know, reality that a whole movie will be generated on the fly on exactly what I want. So this wallpaper thing is frivolous, right? That I was telling you about.

Right.

But really, like, I'm going to watch a movie someday like that. It's like the Raiders lost and I'm going to go and watch this movie after and it's going to have this context and it's going to give me like a feel-good story.

Yeah, we're like, let's brainstorm. So like NBA highlights. So, you know, used to be SportsCenter where if I didn't watch the games, I got to go to my TV, turn on ESPN, SportsCenter would start. Even if I like basketball, I got to wait till whenever basketball comes up and then they'll have the top 10 highlights and I watch whatever they picked. Then YouTube comes around. It's like, forget waiting, forget the TV. Just pick the thing you want. You could search anything. You could search only Steph Curry 3-pointers. If someone made it, you could choose the best of that. Now it's going to be like, I just say, I just will put on my headset or put on my glasses or whatever. And I'm just going to say, show me what happened in the basketball game. It'll, it'll just start generating a new highlight reel that didn't— nobody has ever created. It'll just make it based on my prompt. And then I might be able to say, don't show me any Lakers clips, only show me Warriors. And it'll just auto-adjust on the fly. Like it's going to be made for me.

100%. Yeah.

That's interesting.

And, you know, like, I think it's going to be an interesting world. We're going to go from human content consumption, people, you know, humans make content. We're going to like then have machines make it towards our tastes and liking. And then I think there's always this worry like, well, what happens to all the humans then? And then I think we get back to that core thing, which is Machines will never have true taste, right? Like, and I think forever that creativity is going to come from humans. And there's always a new thing. There's new fashion, there's new kind of, you know, content mediums, there's new everything. Machines will learn it and lag behind or maybe get ahead of it. There's always going to be another person that shows up and does something different. So I think taste is going to be the ultimate thing that probably won't go to a machine. It could reason about it. It could try to invent it. But I think we're not that predictable as humans.

Well, I want to believe you because it sounds good to me. But then I also think, well, right now, if we said taste is like, you know, what is taste? Taste is selection. It's knowing what's good and what's bad. It's right. That's kind of taste. Like TikTok, which is the most popular, like, app, most addictive app, most used app. Their algorithm is basically saying, I'll choose what's interesting for you. And it does it so well. Isn't that kind of taste also? Right?

Like it is. And it works really well.

They don't make the videos, but they select amazingly.

The selection is incredible. It probably is more like what we want than we will even admit. Yeah, exactly. Like, I don't want this.

Right.

Like the other day I was talking about—

Oh, why is it showing me this? It's like, because you love it. That's why.

The other day I was talking about how like on Twitter, my following I love and my For You I don't.

Right.

But it's like, really? Why did they pick this?

Well, that's like a status thing to do is to be like, I don't use the algorithms.

Correct.

I hand do. It's like, I drive stick. I hand make things. I cook from scratch. It's like, cool, but brownies out of the box kind of work for everybody.

I'd love to trade algorithms. I'd love to swap with you for a day and be like, yo, Sean, what are you watching over here?

So I don't know.

It knows you really well. But I think that's where human, we're not great at logic sometimes. And sometimes we have our own blind spots. I think this is one of them. Things that we, we love, our actions prove it. We don't— we don't want to. And it could be like bad thinking. It could be like emotional thinking. It could be like, this is— I don't really want to love it, but I, I do.

Right.

And so there's— there is that element that I think, look, machines will be great at this too. They'll learn humans. It might become even better than this. I just feel like we're just a little bit not predictable because almost make some dumb choices.

Right.

And that actually is part of society and, and just humans in general. And machines try to be too perfect. It's like my Netflix thing is probably accurate for me. It's also just like, man, like, can I just mix it up? Can you just like RNG the algorithm a little bit? Because I just want some different stuff and you're kind of like driving me down one direction. And once in a while, like, I want this 10%, 20%, 30% thing. And I don't think algorithms do that exceptionally well. They test it, they throw things up, but, you know, they don't necessarily try to give you variance.

How long until you think the number one hit song in the world will be just an AI-created song? How many months or years until we see that?

If we were doing over-under 2025, I might pick the under. You know, if you think 2025, like, that would be a good bet. Maybe this is a poly market. Now that the election's done. Yeah. We did a new poly market, you know, like Reason. It could be something like this.

But like, take Polymarket things. So election just happened. Polymarket's having like a victory lap right now.

Yeah.

You showed me Polymarket, I think like years ago, and I started making a bunch of degen bets before they blocked it in the US. It was like, it was, it was like open for a while, right? But they were doing prediction markets. It's a prediction market, but prediction markets, a bunch of people had tried that. Do you know, like, I'm just curious your opinion. What did they do right that like Augur and these other guys who had the same sort of general idea that, hey, there'll be prediction markets, Yeah. Like from either entrepreneur level or product choice, like any reason you think they won that you could put your finger on?

I feel like timing is a big part of it. So like the sweet spot of like people are way more digital. I think COVID drove a lot of people online. Like we just sat at home and we're like, yeah, what else you got? 24/7 on the internet. And it really exploded the internet. Like I think I've seen some graphs where like the internet and e-commerce and all these things were like growing. And then there's like a massive jump, like, right.

You know, remember from like we were at Twitch when that happened and it was like all the growth looks amazing. We're crushing all the metrics. Like, what did you do? It's like, well, we were here. Yeah, right. We are not— we don't create the wave. We surf a wave. And a huge wave happened for Twitch. It was like Fortnite came out and then COVID happened. 2 humongous waves back to back that just combined. Yeah.

So I think it's a combination. We're way more digital. I think you talked about this a lot, like as a kind of metaverse concept a long time ago that We're just way more digital now. We care more about it. I think news and like where we get information from has totally changed, like online, internet, even content and entertainment. Like I have a hard time going to TVs. I look at kids, they look at desktops and TVs as like ancient, right? Like I say, why is this thing on the wall?

It's like my mom's sewing machine. Yeah.

Oh, that's cool. Yeah.

What do you do with that?

That's what it feels like. And so we're, we're kind of there. And then there's a bunch of these people online that want a, like, want a stake in this thing, right? Like you want to root for a team. This politics has also become very polarized, right? It's really like a team sport.

It's like red team versus blue team.

That's really what it is. And so I think it's like, what do I do with it now? How do I support it more? I give more money to it, but I can give more attention, right? And so I think a lot of these things are happening in Polymarket. Good flow, good area, right? And then I think for the best thing probably for them is they got it right.

Right place, right time.

The answer was right. Like if they were wrong, let's say Polymarket was off, right? Like it was, hey, the result of the election was something else.

It would be a story today would be much different, right?

And then it might have just been muted. It would have been a cool thing that kind of didn't work. So you have to be right.

I mean, I think people would have used that to crush, to really rip on them because like in the way that people are doing with polls, but polls kind of have this like layer of protection around them, whereas like people want to hate on crypto things, people want to hate on betting. It's like a degenerate behavior in general. I think if they were— I think if Polymarket was wrong, the reaction would have been much worse than the fact that the polls are wrong.

What the reaction is to the polls, you know, it would have been worse for a little while, then we would have moved on.

Forget about it. And by the way, you saw this thing about the French whale on Polymarket?

I didn't see it. So the big whale that came in and moved.

So there's like the narrative versus reality. When the narrative from the polls was it's a toss-up, razor-close, 50-50 election, this guy came in and bet, I think, something like $30 or $40 million on Trump.

Yeah.

And people were like, is this guy just trying to manipulate the market? Is he real? Is he just a rich billionaire's son? Like, who is this? And today I just saw something on my way here. I don't know the full story because it was on my phone, but it said he bet because he believed he would win. The reason he believed he was winning, he did independent polling.

Wow.

He funded his own independent polling and thought and felt that he was getting better data that was saying that Trump was mispriced. So he's like, I just did a logical, rational thing. Yeah, I just bet where I thought an asset was mispriced. I wasn't trying to— this wasn't political. I'm French. What do I care? Yeah, I just thought there was money to be made. And so he kind of went counter to narrative and he made, I think, like something like $20, $30 million yesterday.

Yeah.

On that bet.

So the funny thing about Polymarket, by the way, it's like we can't use it in the US.

Yeah.

So there's a whole world's entertainment place to like bet on our— like, right.

So it's like $3 billion. Yeah.

It's like we're the ring, right? Like, yeah, 2 boxers going at it. It's in America, the cockfight, and then you've got everyone else betting. Betting from all around the world on who's going to win this thing, which has been kind of a hilarious thing. Yeah, that's true.

You have a good contrarian opinion about VR and in general there's probably some other technologies like this. 3D printing might be one I know you're pretty bullish on too. But there's these tech things like VR where I think if I walk out of here and I talk to 100 people about what are you most excited about, AI, whatever, Bitcoin, whatever it's going to be. But if I said, what do you think about VR? Uh, it's like sort of lukewarm, uh, at best. Um, in the tech industry, I think most VC sort of feels like it's kind of a dead-end technology. Now they'll be like, it's going to be glasses and smart glasses, AR, that's the future. But you have a different take on VR. What, uh, you've been telling me for a while, like, hey, look, Oculus sold more units. Hey, look, you can do this now. And you've been staying with it when I think interest has, has sort of waned. The narrative has gone against it.

Yeah.

And that's always where there's big business opportunities. If the narrative goes one way, but the reality goes another, that's where there's an opportunity. Give me your VR take. So why is VR a sleeping giant?

Yeah. I mean, so you remember 2020, I like asked you for your address and mailed you a headset. I was like, hey, this Quest thing is a different—

That's a real friend right there. He sends me a VR headset.

He's like, yo, I know you're not going to come to the future.

Let me drag you.

It's probably collecting dust, which is okay. But you know, that's kind of how I think about things and Technology takes a long time. It typically takes longer to get there than we expect, like especially people who are early in the industry. We're really wanting to push it. I mean, AI, crypto, VR all have the same problem, which is early on people are like really pushing it to you. It's not ready. So it's a big hype, it crashes, everyone moves on. So that's happened in VR a few times and it's kind of, I would say for most people, been quiet, not thinking about it. But I think it's a sleeping giant, and I think it's a massive sleeping giant for a few reasons. One, I see it every day here, like here at Founders Inc. Like, I get a chance to, you know, one, we invest in VR companies. We're one of the few that do, right? We have a whole corner, maybe like 12 devs that are all building different VR products. We put them all in one spot, and there might be some of the most density of interesting VR projects in one place. And so what have we seen? One, Quest 1 came out, it was wireless. It was kind of crappy, but man, I could sit on my couch and use it. It was cheap. Uh, became a, you know, great Christmas gift in year 2. They didn't have inventory year 1. They released Quest 2. It got even better, lighter, more powerful. Now Quest 3 is coming out, and I think they got like 5 to 10 million monthly active headsets out there, right? I think that rivals consoles. And my first thought was like—

How many units has it sold? I think Quest has sold—

I would say 30 million is my guess, but You know, maybe it's more. Yeah.

So this is Quest. Uh, Quest has sold over 20 million units with the majority being Quest 2. Quest 3 is currently sell— uh, has sold a million units, uh, at the $500 price point. Um, so I think, I mean, if you just, I don't know what the math here is, but, you know, they've done, what is that? Over a billion dollars of revenue on Quest.

And I believe it's outpaced—

Pretty good for a failure.

Yeah. It's outpaced PS5 sales, for example. And so the first thought for me was like, hey, what's the first thing that's going to happen in the spatial environment? Like, what is it good for? Immersion, gaming. Like, these are natural entertainment, right? These were the first few things that was natural. And you had Beat Saber, you had a few things, and that was kind of like, you know, here's the first use case. Let's try to get it out there. Let's try to make the thing happen. And then it kind of goes up, down. People get used to those first experiences. I think you played Beat Saber. Yeah.

Magical the first few times.

Correct. And then you're like, okay, whatever. And then, you know, because it's kind of like cheaper than PS5 unit, the Quest 2 was, it was available on Christmas a few times. It went to like, you know, kids, like 12, 13, 14-year-old kids kind of getting this thing like they would be getting a PS5 and they use it and they don't have the bias that we have of like many years of like the structures we're used to. And so a lot of games formed and specifically social games like Gorilla Tag. I don't know if you've heard about Gorilla Tag. What is Gorilla Tag? Gorilla Tag is a social multiplayer game. It's really fun. It on App Lab did like tens of millions of downloads.

Is it like choose people tag?

Yeah, it's like a, you know, like kind of like a fun social game that you can play with a bunch of people. You go in, by the way, you go in, you're going to be like, dude, there's a bunch of like teenagers screaming at each other. But for them, this is the environment. It's a new place. And they didn't grow up with these other things. So they're starting with mobile phones. The TV feels ancient, the desktop feels ancient, and this thing on my face actually feels like more fresh, more new. So that's where it's starting. I think Gorilla Tag has done $200 million of revenue. And this is where I'm like, VR is a sleeping giant. We have a few teams here at our studio. So we have a team called Fluid that's building the kind of like best browser in VR. So you get multiple displays, you get as many kind of tabs as you want. You could customize your environment, you get AI environments. You're like, I want to be in a cave. It like makes you in a cave.

Right.

And then you get like social multiplayer so people can show up in your environment. And we're on App Lab, does like about 5,000 weekly active users still. Small team of 3, just like, you know, without a big burn, can just build this, grow it. We're not even in the store yet. We're in like this, like App Lab is like kind of the pre-store where we could just drive people to it, but we're not getting anything directly from the store except our own searches. Right. And, you know, 5,000 people like go in, use a productivity thing. There's another product called Yeeps. It's like the second game behind Gorilla Tag. Small team, absolutely crushing it.

Yeeps?

Yeah. Y-E-E-P-S. Really fun game. You know, we should play together.

So this is— they're here. They're a team of how many people?

I think they're like less than 10 people, but more like 6 to 8. And most of it was built with a few people.

And this thing is profitable or what's the deal? Yeah.

I mean, I'll let them talk about their numbers, so I, you know, I don't want to say anything, but—

On a scale of, uh, that's pretty good to like, wow, where is it at? Wow. It's wow.

Yeah, it's wow.

That's great. So what's cool about this is like supply demand, right? So like you can go be, you know, app number, you know, 5 million in the store right now. Or if you're talented, you could be like one of the top 20 VR apps if you put in like, you know, a year of hard work. I'm just using like kind of round numbers or whatever, but It's the same way that right now, if you're a content creator, you go post on Instagram, it's pretty tough. Post on LinkedIn, you'll get tons of distribution if you're half decent at content because there's just no supply of quality content. So even if VR is not, you know, like 20 million units is very good, but even if it's not just like becoming this like global phenomenon, right now, you can still build a great business. And if you just keep riding the wave, you're very well positioned Yeah. To be the leader. And then everyone at that point will look back and be like, oh yeah, it's because they started 5 years ago, you know, when this was smaller, when it was $20 million instead of $200 million.

In these emerging tech things. So everything I found is we do emerging tech. I think the theme of all of them is survive. If you make it to when the industry happens, you will grow with it. If you were a small percentage of the industry and the industry grows by like 100x, you grew by 100x or more. You've already been there and I've seen a few small person teams at like $10 million plus a year, like 5 people, totally profitable, totally able to do it. Is this a VC investable business? Honestly, I don't care.

Yeah.

Like what I care about is like, this is interesting. And can you make these bets without massive capital expenditure, right? Like if it takes like $50 million to build a VR game, it's like the big giant blockbuster movie. I don't— those bets don't excite me. I think when it's like 3 to 5 people can be somewhere with a limited amount of money and just them, it's like them in the hoop. Like the basketball analogy is like, great, we have everything we need. We have all the talent, we have all the ability. The tools are amazing now. Like, you know, all the game engines have perfected themselves over time. And then now the environment's forming, right? Meta has led it, right? So like the Quest and this is the VR world. And then we're seeing glasses, Vision Pro, Ray-Bans. Like the trend is like we're going to have compute in our spatial view, right?

Right.

And I think that's the big like, yeah, this is happening in VR. The Quest platform is interesting. You can build a profitable business or a fairly big game right now. Also, this isn't slowing down. Historically, Apple, when they enter an industry, they come with a unit. It's okay. There's a lot of things that got to get better. Another unit comes out and then you have Snap, you have Meta with Ray-Ban. Like, I mean, like—

Zuck is not stopping. Zuck is not giving up. He's going to the endgame here. Apple probably also won't stop.

Correct.

So now you have the 2 biggest players. They're, they're going to keep making this hardware better. They're going to be super hungry for content. And the other sneaky thing about these, by the way, that I didn't really fully realize until I moved to Silicon Valley, which is a lot of these require like really specialized talent. So I remember when I first met you, you were like, I'm really into big data. You started saying words like Hadoop. I didn't know what the hell you were talking about. It wasn't as popular back then. This is like 2015 or something like that when you were telling me like, Hey, I think this is really— I think this big data machine learning is really interesting. You were again pretty early onto that. Crypto, same thing. You're early onto that. There weren't a lot of smart contract developers. There wasn't a lot of, you know, big data people, AI. So then if even if you don't have a hit product, if you just assemble like A+ talent that's super specialized, then as those platforms rise, your team itself becomes like a $100 million asset.

Exactly.

If you built today for like, because Meta, Ray-Bans, like That product's actually a hit for Facebook and they're going to keep going with that. And everybody wants to be in the glasses thing. People think Glass is the next platform. So if you build a specialized team that's good at developing for that platform, there's just not a lot of great teams that do that. That's a $100 million team even without a hit product.

How hard was it to find—

with a hit product, you get $1 billion.

Harder was it to find an iOS developer when we were starting to do mobile. It felt so specialized. We're going to compete at like a ridiculous level of price. to go get pretty good talent in that space.

I remember 2012, which was not even early. We had one iOS dev on our entire team and it was so hard to recruit talent. It was faster to just retrain. So we used, we just stopped working and the iOS dev trained all of our other devs to be good enough to be dangerous because it was so scarce to get good iOS. And iOS wasn't even that specialized compared to like the, either the fancy AI stuff or, you know, VR, mixed reality, all that. That type of talent.

And so I just think that when I think about kids growing up and all the, let's say, 13 to 15-year-old kids or somewhere in the teenagers, they have a mobile phone that is attached to them. They think it's superior than a computer or a desktop because it's with them. And then we're going to take the second computing interface where we do more work or more immersive experiences and we put it around your eyes. That makes way more sense to me than TVs, desktops, and even laptops. And, you know, that was like one of the thoughts where we started Fluid. Like we talked about this in 2020. I don't know if you remember. There's one project that—

You used to tell me, you're like, I work in VR.

Correct. I did that experiment where I worked in VR for an hour a day and then I ended up doing it for a few months, by the way. Like, and I was like, this is great. And it had been like literally sitting with me. And I think this is one of the reasons why I actually built a studio is like, I want to do all these ideas. I can no longer do them all. I could pick a few. I probably shouldn't even go do more, but like, I just want to find really hungry people and like match them together.

Right.

That's how I found John at Fluid. It's like he was a PhD student. He was going to go do something in like finance and high-frequency trading and like whatever. And he also came to this idea of like, hey, when I was doing my like, you know, master's, whatever, I was writing this thing. I just wanted to go into a focus mode. Like, could I just go into a cave and like block out all the stuff? He's like, I had all these VR friends and I had done some stuff in VR. So like, I tried to do it and it wasn't good enough.

Right.

And then he kind of left it at that. And then we were talking about it and he talked with Uber and he was like, dude, you should talk to Furqan. I think he's got like 5 pages of like random notes and ideas around this. And then that's what spawned it. And it's like, cool. Like, we're not going to blow ourselves out. We're not going to go raise a lot of money. We're not going to go get a giant team. We're going to get 3 people here. We're just going to grind and work hard and build and survive. for when this thing happens. We'll miss some bets like that. We'll gain ridiculous amounts of knowledge of like the industry. And then we find these sleeping giants and it's like, yeah, we're going to double down, right? Like, Gubert went to this like VR conference, like the Meta Connect thing, and he had a t-shirt and we're like, the funny thing would be like on your shirt, just put I invest in VR, right? And so he did. And it was like, wow, like, yeah, we found one.

You know, you're like a girl at a crypto conference. This is amazing.

This is how we want to think about it is like try to be there early. Don't get too caught up with where the technology is now. Don't get too scared of how far it is. 2015, 2016, remember we heard self-driving cars are never going to happen? I don't know if you remember that. Yeah, yeah. Massive, like, push against it. And there's Waymos running around every day now. I see more Waymos at night driving around than, like, regular people driving.

You've told me once before, you go, I think my superpower is that I'm usually in the top, in the first 1,000 or 10,000 people to try any new technology and like understand it, be able to play with it, know it, see it, all that stuff. I remember we were at work and you were buying the Ethereum presale, the ICO or whatever the presale—

Do you remember what you said about it? Yeah, I don't know if you remember.

I just remember being like, dude, can we do some real work? What are you doing? Ethereum name? That name's never going to work. I was like, that sounds so nerdy. That'll never be a thing.

I think that was a phrase I remember. I was telling you about Ethereum. I was like, dude, I stayed up all night reading this like white paper. I'm like—

You were like on it.

I was on it. I came in, I'm like, where's Song? I'm going to talk to him. I tell you this, it's probably— I don't even know what I said. Who knows what? Probably threw a bunch of words and you're just like, dude, I don't know what this is. We have something really to share. And you're like, Ethereum, that name's stupid. Yeah, I wrote it off.

Chalk that up for another, another L for me.

By the way, you could have been totally right as well.

So no, no, I wasn't. I think that's all that matters there. Very, very off on that one.

But you've—

I think that's true, that that is your superpower. And then you said another thing today, which I hadn't heard you, which is for emerging tech, there's only one rule: survive. And that reminds me, we had— I just did a pod with Ryan Peterson. He created Flexport.

Yep.

He said the same, same principle. He goes, you know, what I realized was I cannot control the timeline. I don't know how long something's going to take to work. So all I focus on is how do I just be default alive? How do I just stay in the game? He goes, I just have the confidence that if I'm in the game, I'll just keep trying shit until it works. I just believe that about myself. I will just keep trying things until I figure out another thing to work. The only way I can lose is if I have to get out of the game. Which is like, I can't, you know, usually it's like I run out of funding or I can't control my destiny. I spend too much money, I'm burning too much capital. And so he's like, that's been the name of the game for me from when he was flipping scooters on eBay to now, you know, runs a multibillion-dollar company called Flexport. And the whole time he's like, my whole thing is I can't control the timeline, so I'm going to control staying in the game because if I stay in the game, I win.

Stay in the game with great talent and people that want to have this long-term mindset. No, you know, I think there's a lot of people that can take short-term wins and they should. Like, we've done that, right? There are these short-term moments like, oh, this is great for us right now.

Yeah.

But we're going to keep doing more. And like, I think that's the main thing for me. Technology takes a long time. When it hits though, it happens very fast. Like, that's the part that I think people don't realize. They like underestimate how long it's going to take and then they overestimate how fast it's going to happen. It's like immediately going to just like happen. And, you know, I think Uber was an example like that. Like we saw it, it was black cars.

Yeah.

It's like, oh, it's like expensive taxi that nobody's going to use. And then now it's like, dude, this place doesn't have Uber. Like, how am I going to move around?

Yeah, I can't function.

It feels like that, you know? And so I think I enjoy it. There's one reason why it's like, it's like a lifelong game. I could just play technology forever. I could learn about things. I don't have a rush to be there first. I don't have a rush to be like, I didn't know everything and like do everything and raise the most money and get a bazillion people. It's more like I need a few people with me that like are really excited about this. They see it like I see it and we're just going to go on this mission. And luckily now I have the ability to just like fund that and like create my like environment to do it. That's what this whole building is.

So let's talk about this building because you, you've basically built like your dream, like man cave in a way. I think, you know, it's kind of like a founder dream, right? Like, And I want to talk about that because you had a big success with AppLovin where, uh, and actually kind of like felt like there was multiple moments where, you know, it was successful. At one point it sold for like $2 billion.

Yeah.

It was like, oh, exit, $2 billion. Amazing. And then like Trump blocked it or it was like, didn't go through, didn't go through. But then during that, this is a crazy story. It was like we were sitting in the office and this news happens like, dude, congrats. That's amazing. Holy shit. And then like 9 months goes by, the deal doesn't like fully get approved because it was such a big purchase out of a Chinese company group or something like that. And in the meantime, the business just kept crushing. So Adam, who's the CEO of that, you said he like basically went back, was like, cool, we'll still do the $2 billion, but now it's for 30%. Yeah. I don't know what exactly it was. Some version of that.

Yeah, there was like, I don't know, like, uh—

Because you couldn't do majority deals. That was the big blocker.

Yeah, something like that.

If a majority deal was going to happen. And so now then the company eventually IPOs, you get this nest egg, right? So it's like, okay, I could do whatever. You could go retire, you could go buy islands and cars or do rich guy stuff. Like you could do that. And instead you like chose to do something else. Can you just describe basically like what's the mindset? What's the conversation you had with yourself?

Yeah.

Now that you had more resources to do whatever you wanted to do.

Yeah, I feel like my whole life I've just always wanted to tinker and build stuff. Like, I always described it as like, I love taking stuff apart and putting it back together. It's not like some people will say it's like, oh, you want to learn how it works. It's not that. It's more of the puzzle.

Okay.

How does somebody else put this? Like, you know, and I used to do this with cars and computers growing up. Like, I would overclock my computer and I would make my car faster. And like, that was like kind of just like the mentality that I had. And then I was like, okay, well, I've got to become an adult at some point in time and do the thing. But like, hey, I like this business thing. I buy and sell stuff. It's like a way for me to hustle and kind of do more of what I want to do. And then I was like, let me just keep doing this. And then maybe at some point I got to get a real job. And kind of luckily, like, tech was really valuable and like my skill set improved, I got better at it. And but a lot of that early journey was like more solo than like with a bunch of people. Um, and then met the guys, Adam, before AppLovin, like kind of like these other, other things. And it was like, yo, we're like 8 people in like Palo Alto, like building cool random apps. Like the energy for me there every day was like through the roof. And then Monkey Inferno was like the same thing, but like even more. Um, and I think I used to tell you like, I want an airport hangar. I just want to put a bunch of cool stuff in it.

Yeah.

And I just think it was like that. Before I started Founders Inc. as it is, because I was at Twitch and I'm like, I got to get out of here, right? It was like an instant, instant thing.

And why, by the way?

I don't know. I don't think I'm a good employee. I think I'm like suited for a few roles and it's typically like doing my own stuff without much.

Was there anything that just drove you crazy about it?

Was there any like— You know, we're going to do a whole nother pod on it. Basically, I feel like the first week or month of like, yo, let's do a bunch of stuff. And they're like, slow down.

Like, why?

I want to do more. Like, I want to do more things.

Yeah.

And this resistance feeling of like, oh, we already thought about this or we tried this. Like, no, let's just go do stuff. And I never enjoyed that. And I think startups and small teams just because you have so much to do, that's the kind of mentality. And so I talked to probably like 75 to 100 founders before starting Founders Inc. And I really wanted to learn this. Like, I knew a bunch of people. I used to interact with them even at like Bebo and Monkey Inferno. I'd have them come by and like just talk, whatever. And I really enjoyed that. That was like fun. I started angel investing a lot, which was like cool. I thought I wanted to do that, but it wasn't fun. It was like meet great talented people, get excited about them, and then you're like a monthly update away. Like, give them the check and then that's it, right? Damn, this isn't fun.

First meeting is great. It's a great first date.

I walk away and I'm like, then there's no relationship.

And you're like, what happened? Like, great first date.

Exactly. And so I was like, man, I don't really want to do that. But something like this where I could help these like early entrepreneurs, things that I've just done over and over again. I didn't have this. Like, what's a version of that? And I talked to a bunch of founders and they all said something to me that I was like, what do you need? Like, what help do you need? I thought they're all going to say money and I'm gonna start a fund like that. That's what I thought was going to happen. And they all said something different. They said something like, I need people who understand my problems. And we used to do those masterminds and it was really about like, who do you go to for founder problems? Yeah, it's like, what are your founder problems? They're your co-founders, your employees, or your investors. So you can't go to any of these people. They might be the problem, right? And so who do you go to?

Or even if they're not the problem, you need to present. You don't want to worry your employees. You don't want to worry your investors. You kind of have to maintain a certain aura of momentum and morale. So you can't just go be dumping problems on them or be like, I don't know. You're supposed to know you're the guy.

And we used to— we would go in that room, that a little circle room with the circle table. We'd be in there and I don't know what the employees were thinking. They're just like, oh man, these guys are talking again. They're going to come out with something different. But we could talk about anything and then leave the room, be like, now we're still where we are.

Yeah.

And I think, you know, co-founder can be that for you. A lot of entrepreneurs starting out, they don't have that. And even if they have it, there's like other things that they're experiencing. And so when we used to put these folks together in these kind of masterminds, that feeling was like awesome and it felt like we could relate. And actually, I heard the same thing from a bunch of founders I used to talk to. And I was like, I think it's something more like this where I could do something and put everybody in a box. And I thought it was going to be more digital. I think COVID times and, you know—

Started on Discord or whatever.

Started on Discord. I mean, like, Farzad Ben was in these groups, like, there's a few others, like, and it was just people I knew around me and kind of COVID hit. It was digital. We were talking every week. We would give ourselves these like accountability kind of Ship It sessions. We would have an area to talk about stuff and you could see the dots connecting. And then I got an opportunity to like get this space and I'd been looking for like something and, you know, it wasn't like, oh, San Francisco needs a place. I'm from the Bay Area normally, so this is the best place for me. But I said, look, if it's going to be SF, we're going to be on the water.

Yeah.

No more SoMa, no more kind of places like that. So we got this like weird opportunity to find this space and like really make a bet when like nobody else was. I think it was like late 2020 is when I approached them here. Yeah. And it took maybe like 9 months to figure it all out, another 3 months to like renovate it and stuff like that.

So it's still very much like, we got to call another call about breaking a lease. Wait, wait, he wants to sign a lease right now? Oh, come on in.

That's what it felt like here. I mean, this place, Fort Mason, has 300 events a year. It went to zero with COVID Um, all the places, you know, art galleries, art schools, like, what do you, what do you do? Like, you know, and this is supposed to be like this innovation place and It's a little bit older and—

Yes, that's one of your tricks is that you don't run away when the dips happen. I remember when early on, 2013, '14, something like that, we meet, I buy Bitcoin because I don't know, you're into crypto. Some people in the office, you know, maybe I was dumb about the Ethereum idea, right? Yeah. PG, Pete, Pete Hicks was, was, was all, you know, he's mining Bitcoin on our servers. So I buy some Bitcoin. And literally like the next week, I forget, like super convinced. I'm like, oh guys, I see it. I believe. Here's why I believe. I'm giving you my case. I bought. Yeah, great. Like the next week, price cuts in half. It goes down to like $300 or something like that. And I came in, I'm like, oh, Bitcoin. I was, you know, whatever. And you go—

That was a fun week or two, you know?

Well, you were like, oh, this is great because now everything's half off. Like you literally told me that. You go, you believe if you, if you buy now, you can cut your buy price. Basically, you can go down by 50%. You'll, you'll, cost average in at half the amount.

Yeah.

Like, and I was like, oh, he's right. Because I was just riding a roller coaster of like, you know, I was doing what a cliché person would do. When things are good, this is great. When things are bad, oh, maybe it's not great. Whereas you were like, dude, did anything actually change or just the external sentiment? And so I bought more. So I thank you for that. That was a very good, you know, decision at the time to go buy more when things go down. I think you've done that with with other, you know, bets, whether it's San Francisco real estate during the COVID time or it's, you know, whether it's crypto or VR, when, when things go out of fashion, I feel like you don't run away, which is—

I feel like there's a signal. It's like things go out of fashion and there's another place where the people who don't shape narratives typically, like, I'm technical, so I live in these like GitHub projects. I live where the builders are, like engage with them. If I see that somewhere where it's like, huh, all the people talking are like against it. These people, like nobody told them yet that like it's done. Like, oh, this is dead, right? Like they're all saying it. And then these guys are just shipping more code. I'm like, hey, did anybody tell you it's over? Like, what are you talking about?

I don't care.

Like, and I just think that's the perfect environment.

And even if you're right 1 out of 5 times like that, you'll be right in such a big way that it works out.

And then, by the way, for me, it's just fun. So like, this is fun. Like, I love building stuff. I like building technology. I did a lot of software, obviously, because it's been magical to develop things and like distribute it to the world. We had some hardware projects. I don't know if you remember Jamie, by the way. Yeah.

Yeah. We at one point in time when we were, when we were working together, we had an idea for a— it was cool. It was like a voice control. It's kind of like Alexa.

Yeah.

We called it Jamie.

And I don't know why, but like, yeah, we like Jamie.

I still like the name. That was good. And what you were saying was like, you're like, we can, instead of creating a, everyone was trying to create a device. Portal was trying to create a screen. Amazon was trying to create a screen. You're like, that makes it expensive. You're like, everybody already has TV screens in their house. What if you just plug in like a Chromecast and now you turn your TV from just a blank screen into like an Amazon, like Alexa thing?

Yeah.

That was cool. We didn't, we didn't end up doing it because it was good. Like we sort of saw like huge distraction. We were like, yo, that's going to be like a brutal battlefield.

We took it to Michael and he just gave us that look like, Uh, this is—

He was just like, yo, when all of the like trillion-dollar companies are going to go after the same prize, like, yeah, you can, but do you really want to? Like, you know, is it better to do the things they're overlooking? I just think it's probably good advice. To be fair, we also were going to do a crypto exchange when crypto got hot, and he was like, for a different reason, he's like, hey, I'm already rich. I don't want to lose everything. And I don't really know what crypto is. It's 2014. Like crypto, might just be like super illegal and I don't want to risk it all on that.

Yeah.

If we had done that, it might have been good.

Yeah.

So you have to be careful. Like even really smart, successful people, you can't like just take their word. You know, you got to have the independent-mindedness.

100%. And I think it's fun to take shots. Don't get stubborn over it. Like, damn, people fall in love with their ideas. I think that's— it took me probably like, I don't know, 10 years to figure that out. Like you have to get it like almost like you have to really take the hits.

Yeah.

To like really like live in that of like, Yeah, there's a lot of ideas. I may try a lot of them. My mission isn't all of them. I got to find the right ones where I can really spend the energy on. But I was going to talk about one more. So we built Chatty Heads, which, by the way, now in the AI world, we were fucking ahead of the time.

Yeah, we were a little early.

We could generate images for like, I don't know, $5,000.

Yeah.

And that's what it feels like. But I think it's cool. It's like whatever technology you have today, go try to produce a thing. It might not be the right time. It might not be the right moment. The medium might be wrong. The team might be wrong. Some of those you should pursue again and again. And some of those are great learning exercises to build on top of. And I got a chance to meet a lot of people who were professional athletes. And I think one thing a lot of them talk about is like, basketball might end. Like, my career will end at like 35. It's like, that's my game. It's done. Now what do I do? And I think what we get a chance to do, whether it's like business or for you now content or like building stuff, like I'm going to do this for the rest of our lives if we want. I think that's the fun part.

Buffett's like, what, 90-something? He's still at the top of his investing game.

Exactly. And so it's like, well, like, what am I in the rush for? Like, I'm not just here to like enjoy the journey, but I also don't want to be like, I got to solve it tomorrow.

Right?

Like when I was young, it was like, I got to be the millionaire by whatever. It's like at some point I was like, I don't know, man. I just want to keep doing this.

Right.

And if I need to like hustle my way to it or not, like it doesn't really matter. Like, you know, I took an Android engineer job at Monkey Inferno.

Yeah.

Because I was like, you know, I met you and I was like, I want to work with this guy.

Yeah.

I think that was my worst skill, by the way. I just learned Android.

Yeah, you fooled me. Yeah.

So like, I was like, I don't know, I got to get in somehow. And like, I know I'm going to do great stuff here. I got to show it. But like, I'm not afraid to put in the effort when you need to put in the effort. Also not afraid to like not rush to the answer. Right. And like, you don't want to be like casual and like, wait, you want to be kind of like in the middle there. You want to know when to attack and when to not. But like, I don't know, you got to enjoy it.

Otherwise you're really— Naval has the best quote on this. He says, impatience with action, patience with results. Oh yeah, it's the unbeatable combination. If you ever go against somebody who's going to operate like that, they will win. Yeah, that is a— you cannot lose if you're going to be constantly impatient with doing things. You're not going to sit back and hope it all happens. So you're impatient with action but patient with results. That's the hard part is a lot of founders are impatient with action and impatient with results.

Yeah.

Or non-founders are, you know, patient with both and then nothing ever happens. So, you know, you have to get that combo.

Yeah.

Speaking of founders, you worked with Adam at AppLovin, and AppLovin has been like kind of a staggering company because, you know, when I met you, it was, you know, a successful company. You told me the stories about Before you guys started that, you're like, we were wandering around, we tried a bunch of different ideas, we were playing FIFA because we didn't have— we didn't know what the heck we were doing. We were just coming in, trying to figure it out. If we didn't have it, we would brainstorm and go home the next day. Um, what is special about that guy? What's a superpower from him or a story from him that you remember that, you know, I can learn from, or anybody listening to this can learn from?

Yeah, so there was this like 4-year period-ish when I was there. I think 3 years of it was like not AppLovin. So it's like most of my intersection there is like not what it is today. But I did get a chance to spend a lot of time with him and ideate ideas.

What's the Cliff Notes of his story so people know? Because people don't know. He's pretty under the radar, right?

Yeah. I mean, his backstory, I think he did some stuff in like equities or trading at some point. I think he got into like ads at some point through like marketing and affiliate stuff. He built a few different products. Maybe he was a key member of the team and then he had like a few companies. I think he had actually built up some, like, you know, wealth. I don't know how much it was, but it was enough where, like, you're like, okay, this person can make this bet and, like, fund the operations.

And so he was kind of self-funding.

Self-funding. It was him and this guy, John. John had done a lot of stuff on the internet. He was more the technical person. Adam was more the business person. They were both uniquely— and the skills incredible, like their personalities incredible. For Adam, I think the thing that I felt the most was like, I think it's the first time in my life I was like, man, this is what A+ execution looks like. Like, this guy just hits it. Like, if we were talking about something, we made a choice within, like, it felt like within minutes that was like delivered to this team. And look, when you're like an 8-person team, it's really easy to do it. People decide something, go slow, maybe next week. Oh, we'll do it later. We'll make these role changes later. We'll tell everybody later. No, it was like immediate. And when it was like moving on from something, it was immediate. When it was a new idea we wanted to do, it was immediate. When it was something else, it was immediate. It was just like, it felt like this is what execution is, is like, you know, think, decide, act. And like how fast you run through that depends on like the moment you decide, the delay on act is like usually a problem. And I think this is where most— I'm not great at this myself, but I've gotten a chance to see that. I think it was kind of similar in a different realm when I met you. I was like, I think this person's product thinking and like ability to like unpack like a complex thing, like product or distribution or maybe team motivation or whatever. It's like you see A+ talent and you're like, you want to do it. For me, it was like 10 or 12 years, almost like a solo founder journey. Like I had this e-commerce company, I had these other little things. I had like a startup. I had people with me, but I never saw somebody else that I was like, Yo, like, I want to learn these skills from— I bring something to the table that's A+, I can pair with this A+ person, and now we're going to be like a superpower, right? I felt that with Adam. Uh, it was very clear and obvious, you know, like the size of company. I, I don't want to say not a surprise, but also it's not a surprise that this kind of person would go do it. Like, like, it just is that. I think the same thing you're talking about, like Ryan at Flexport.

Yeah.

Like, it feels like some people are just like, they're built for that. You still need a lot of stuff to go right and make a ton of great decisions and a ridiculous team.

One thing I've come to learn is that where I think we screwed up, because we did Monkey Inferno, which was basically our little idea lab, but we had a beautiful setup. It's like, yeah, you got funding already done. You have great team. You're in San Francisco, beautiful office, really talented team. Like, you know, we're not the PayPal Mafia, but like everybody's gone on to do kind of interesting shit. Everyone's, you know, I don't think we had the level of success that we could have given the talent. My take was, I think we were, um, good execution, maybe even great execution, but poor project selection. Meaning we were going after these like moonshots, like create the next hit social media app, which is like, you know, yeah, there's been like 7 ever. There's not that many of them ever to exist.

So, you know, I think we did bad with project selection.

It seems like one thing that Adam did aside from great execution was project selection. I think you told me some story about like they went to some conference, you guys were working on one thing altogether, and he came back and had that very quick like think, decide, act loop where he's like, we're doing a mobile ad network, we're doing mobile games. I don't know the full story, but like, yeah, it seemed like just that one choice at that time.

Yeah.

Is the make or break, you know?

Like, yeah, huge difference. And, you know, like, I think we had many of those moments. I wouldn't trade it, by the way. I think our learnings I still leverage them. A lot of the things that we talked about, how we ran the teams, like they still radiate and resonate with me. And so like immense wealth and knowledge and like, like literally like experience of the thing, right? Beyond like—

You're talking about when we worked together. When we worked together. Yeah, exactly.

But there were a few moments where project selection could have been massive for us. And yeah, that outcome, you know, like they, you know, and we can talk about AppLovin a little bit on that and that journey, but like We did Blab. It was a livestreaming product, Google Hangouts, public livestream.

Yeah. Like if you saw Clubhouse get really popular, we had built basically a Clubhouse before Clubhouse. And it got to—

Kind of like what Twitch right now, like there's a big section of this, like just chatting, hanging out category.

Right. We built an app like that. It got to 4 million users, but it didn't become the next big thing.

I don't know if you remember this conversation, but we had this one time when we were deciding what to do next. Like, are we doing the mobile version of this? Because we see other things happening, or do we do the B2B version?

Yeah.

And Zoom didn't exist. And that— I think that was—

And we were like, B2B. I remember there was like— it was like so short of a conversation, which was so silly. We were just like, B2B, that's not cool.

It wasn't cool and it wasn't clear because if you ever think about this, like 2015, there wasn't a billion like B2B companies crushing it.

Yeah.

But like every year since then, it was like 7 or 9, like Because I think at that time it was only a few had really reached, like, it was like, I don't know, Box.net and Dropbox.

And it wasn't, it definitely wasn't as obvious at the same time. It wasn't as hidden as we kind of made it seem.

We totally brushed it off.

Like, I remember Citrix was like a multi-billion dollar product and Citrix Online was the way that people did these webinars and web conferencing at the time. And it was so bad. And their users, like, We were trying to make this cool social app and SAP and Oracle were using our tool just because it was better, even though it was like not meant for that.

Why is your color like weird and purple? Why do you got this weird star thing?

And instead of looking at those clues and being like, huh, maybe we could do that, we, we missed that project selection choice. I remember that day because that was a, that was a probably a multi-hundred million dollar fork in the road moment, you know?

Yeah, maybe.

You still had to execute, but for sure.

I mean, we could execute, right? But it is, did we have the right selection? And could we get the right insight in our mind to Incep? And also when we decided to end Blab, I don't know if you remember, you went to this barbecue. I forgot who you met.

James Currier.

And they were telling us, you know, you're talking about like, hey, this content network problem, because it was like the moment that entered our mind, it's like, oh shit, we're fucked. Like, that's what it felt like. Like, we don't have the ability to intersect these. And we kept looking at Twitch like, why? Okay, so we know we don't because people come on for an hour, they do a show, The Epic content is not on long enough for people to show up and intersect with it. How come Twitch wins? And then it turned out that, oh, people play games for 8 to 10 hours.

Yeah. So it didn't matter when you showed up.

Correct. And then the context resets. So it's like, I have— it was like, you know, the feeling a lot of Monkey Inferno to me and like the things we built was like, it's not slow down to speed up, but like look for the clues. Don't be afraid of that. And again, the eager, you know, the ego or the stubbornness of like, We want to build a giant social app. Like, I want to go somewhere around the world where somebody is using my consumer product. Like, that was the stubbornness that we—

That was the driver.

Yeah. And it's like if we had just kind of unlocked it a little bit, there were these project selection moments. And so there was a lot of that. I still feel that sometimes, but I don't know, maybe it's because I'm older and I'm like less willing to like be nimble in that way. But I did want to talk about one more hardware robotics thing.

Give it to me.

We've got a few minutes. And so I think this is another like resurgence moment happening. Like, for a long time, hardware has been too hard, too expensive. Software gets funded.

That's really a Valley motto, right? Hardware is hard.

Hardware is hard. Software is easy. Software scales. It's eaten the world. Like, these were all the mentalities. I think it's flipping, and I think it's a few things that have all kind of showed up together. So there's 2 types of hardware that I think are now like there and like ripe to build. Same recipe, Can small teams do it? Can you do it without a lot of funding? And then can your output be really big and like impact a lot of people? And so I think there's one around consumer products. So like the combination of Raspberry Pi and cloud AI has completely changed what it takes to build something, right? So there's a company in our studio, Magical Toys, they're building an AI teddy bear. We'll do a little demo of it after. We'll kind of get to that. And, you know, I think what that—

How old is that guy who's building that? He's like young, right?

I think Fatim's like 24 or 25.

And it's not like he has huge funding or a huge team, but he's able to— He showed up here.

I don't know how he got here, to be honest. I think like many people here, they meet somebody, they attract somebody, people just show up, by the way. And it's awesome. Like we've created that environment where it makes sense. He had done some small projects in college. Like he built this thing called Desk Buddy. It was a little like E Ink screen with like 2 eyes that just blinked. And that was really it. You couldn't talk to it. You couldn't do anything. It was just like a little desk buddy.

So you're not alone. Is that the idea? Yeah.

I don't know what it was, but I remember seeing it. I'm like, that's cool. Like I want one on my desk. Like it's just fun. And they brainstormed a bunch of ideas and, you know, the combination of like, hey, we got Raspberry Pis, we could do these little things, we could use 3D printers to build enclosures. And then now we got this cloud AI thing that like can be really powerful. And they ended up coming with this idea, which was like, hey, we're going to build a toy. And first they literally made Ted, the stuffed animal Ted, and made it talk like it. And I was like, oh, this is wild. Like, you know, and AI could do this, but everyone's trying to make coding faster.

Yeah.

Or there's developers solving developer problems. And I love when somebody takes that and goes, let me go to this other place where nobody's thinking about. And he spent probably the last 9 months sitting, refining. I think we sent you one of the first units. It probably broke.

Yeah. First he gave me a teddy bear with the back, like the whole computer was just hanging out the back like a half-done surgery. But it was interesting. I gave it to my kids. I think they were probably like 2 years old, 3 years old at the time. And every other toy we have in our living room is pre-programmed. So it's like, this is a toy, push this button, it'll say this thing. That's all it can do. With this toy, it was like, ignore the thing hanging out the back. It was like, um, hey, we love Paw Patrol. Can you ask us some Paw Patrol trivia? It's like, certainly. I can tell you Paw Patrol. Who is the red dog in Paw Patrol? It's like, Marshall.

Correct.

And I was like, hey, can you keep track of our points? He goes, okay, 2 points. And then I was like, whoa. And my kids were blown away because now you have an infinite toy, whereas every toy is finite. It can only do the things it could do out of the box. Now suddenly you have a toy that's basically ChatGPT shoved into a stuffed animal, right? And you're like, wow, now that can do anything. I could say, sing me a song. I could say, tell me a bedtime story. I could say— I can make it do many, many things now.

And I think what was cool is Fatinh, And this little like lab that we built. And the lab has like some 3D printers, some electronics area. Honestly, we started with like an empty room and people would come by, they're like, what's this room for? I'm like, oh, it's going to be a machine shop electronics lab one day. Like, oh, can I use it? I'm like, yeah, but we have nothing here. What do you need? You need tables? Great, we'll bring tables. Oh, you need like a little thing? Okay, we'll add that. You need 3D printers? We'll add that. And I've seen now like tens of people come through You know, one person sit there, spend some time, tinker. And Fatin did that. Like, he built one, he showed us, he built another, he tried different variants, different versions, you know, built, you know, new cases to be printed, buying Raspberry Pis, new software. And like, I think he shipped like 60 to 70 units across 4 to 6 months, which in software world just feels like, wow, ancient, like 60 users. In hardware, this might have cost like half a million to $1 million. I think we did that for like $50,000 to $100,000.

Right.

So like that difference, one person, 3D printers, Raspberry Pis, some AI, and you could just sit there and deliver units and try it with people. When we showed Uber first, by the way, internally, Uber was like, nah, I don't think kids are going to do this. Weird. His daughter used it and it changed his mind. Like, like she changed his mind.

Right.

Because he saw what it could do. But like, and how do you get that opportunity to prototype cheap?

Right. So hardware being hard, it's like maybe hardware is not as hard as it used to be that, like Sam calls these inflections where, you know, something changes. Like there's famously an inflection where when Obamacare came out, then Oscar Health built a thing that was just to do Obamacare and like, you know, had, you know, became like a billion-dollar company. It's like, so what are the inflections? So, oh, Phones now have GPS, um, in them. Now you can build Uber. You couldn't build Uber before because the driver and the rider needed to find each other. How were they going to do that if you didn't have phones with GPS on them?

Uh, phones have cameras.

Now you can have Instagram, right? So it's like the technology inflection can happen. It's one type of inflection. And you're basically saying because of Raspberry Pi plus AI plus 3D printing—

Consumer hardware is very easy now.

Consumer hardware is now possible.

Now, like, Now more than ever.

2 people can actually mess around and tinker until they get something right. Kind of wasn't feasible 5, 10 years ago.

Right. Like when we built Jamie, I used a Raspberry Pi at that time, which was the first version. Raspberry Pi was thought of as this hobbyist market of tinkerers that are going to buy a few of them. They've, they've sold like 60 million units or something ridiculous like that. Like $35 a pop, 60 million. There's like a few billion dollars of Raspberry Pi out there. And I think what it did is you used to have to make like custom boards, custom software. And, you know, for a technical person like me, I don't want to go that far. Like, I have certain skills that I could do really well. And typically it's like read a few guides on the internet and like stitch stuff together. It felt too far for even like somebody like me. It felt like, man, it's like really serious engineering. I need 10 people, I need millions of dollars. I gotta convince somebody there's a big market and now like build up this company, which probably will fail because I raised too much, too much pressure, too much demand on return, etc. And to me, the Raspberry Pi was one of the unlocks. And there's the NVIDIA Jetson now, which has like GPUs on device and there's so much more there. But—

So what are, what are other things besides magical toys that you saw somebody build? Interesting things you're seeing built in the hardware robotics side?

Yeah. So we have AJ who's building the, you know, Neurocity. He built that first version, which I think I've showed you before, which is brain-computer interface. Now he has a second version that's really tiny. It's like the size of AirPods. You could put it right here, special purpose. So really, like the ability to prototype, develop the thing, get like hundreds or thousands of units out there, improve the design, and doing that without a giant team allows him to kind of continue. We've seen like a variety of things kind of come out on that front. I think the second area that's happening is, you know, there's consumer hardware and then there's like this robotics, drones and kind of this other world where we have so much physical equipment in the world, forklifts and lawn mowers and cars and like, you know, all these things that we do that requires either a person like, you know, with drones, we invest in this company, Lucid Drones, which, you know, they built a power washing drone, could go up in buildings, could power wash the, you know, the glass instead of people hanging from the side. There's another company that—

And is that working? Like, do people— is it like—

Yeah, working very well. And, you know, like, what's interesting is there's like a unique business model to find here as well. And I think this is why I love like being in the weeds a little bit and like seeing it is like most people think you're going to build this product and then you're going to take the people that like manually wash a building and you're going to like get rid of them. What actually happens is there's like a team that's like a little like small business somewhere. They have like 5 people at their company.

Dave's Power Washing.

Dave's Power Washing.

Windows so clear you won't even know it's there.

Amazing, right? Dave's got a great tagline as well. And, you know, people end up finding that if you just sell to those small businesses, you give them more tools, they can serve more buildings, do it more efficiently. That distribution is built in. Maybe in the long run this starts changing, but like Now there are a ton of these opportunities. I've seen one for farms and like, you know, mowing the weeds around certain fruits or kind of like inspecting them. You just send people between these things to go like do it. And now there's like a little like robot and, you know, instead of the one person doing it for 2 weeks or 2 people, the one person's there monitoring it at the facility in air conditioning and like—

Watching on the iPad.

Watching on the iPad. And it might get stuck early on or might detect something. Now they got to go fix that thing. This efficiency, I think, is like massive, and we're seeing it in many places. So the same thing I saw these 2 guys building here in San Francisco, they're like random warehouse in the back of some other store, like 2 dudes like literally grinding, bootstrapping, like, like just going there. I'm like, I like these guys. They took a forklift, they automated it. They took self-driving car tech that probably took billions of dollars to develop And then all of that kind of flowed to open source. And typically open source sometimes is behind the cutting edge, but in 10 years that normalizes. We see this elsewhere, like AI models like Llama, but really they took cameras and some LiDAR stuff and they strapped a little computer to a forklift and they could move it around. You could talk to it, you could kind of get it to do things. And I think there's a massive resurgence. And so Forklift, physical thing. You took a little Raspberry Pi, internet, some cameras, some technology like we were doing computer vision technology to score like Fortnite games. They're using that same kind of technology to look at, oh, this is a box. This is the barcode that I picked up, right? I can now use and figure out what's this thing. It could go walk around the warehouse and be like, oh, that pallet's in the wrong spot.

How?

It just scanned the barcode and detected, oh, there's this pallet here. Supposed to be over there. Go pick it up, move it over there. Like, these problems exist in warehouses. Like, I mean, you know, right? I think you've dealt with some of those. I've run a warehouse. And so to me, it's like, what are the machines? Like, what are the traditional machines we've built probably for the last 100 years that you're going to slap a little computer on it and now it's a superpower? And I think this is bringing down the cost of like this robotic hardware. Like building a robotic arm used to feel like, yo, I got to be like Tony Stark and like Iron Man. I've seen 2 people do it in our machine shop. There's a guy here, I don't even know his company. He just showed— he just had a 3D-printed hand on his desk. And I'm like, what the hell is this? He's like, oh yeah, like, I created the ability to develop a full hand with all the fingers and everything in just this 3D printer. There's a Bambu Lab printer, which honestly, 3D printer was great. And then Bambu Labs took it to a whole nother level, like a little bit of AI to help self-leveling and Make prints great and like everybody loves it. And he just built his hand pretty high quality, took a little fishing net on the inside of the hand. So like, you know, in each finger he could like pull every finger and like do some stuff. It's like one dude, like a few months hackathon and like, I don't know what he's going to produce from that, but I just think you could do these things now in like weeks and months when they took years and like a lot of money. I think there's a huge opportunity here. still going to take time to like marinate and develop. But I look at it as one of those things where like if I was a mechanical engineer, if I liked hardware and I've been told that it's just AI and software and like your stuff isn't that interesting, it's like, no, no, this is very interesting right now.

Right.

You need to find a place to do it. And, you know, like I'll even say, like, I think we're going even further for us. Like we have this space we built here, we call it the Founder Lab. Like this is where a lot of people come, they tinker on stuff. We have founders here, we have builders, we have creators here, we have all kinds of people doing stuff. Built a little machine shop, and that kind of pushed us to be like, wait, there's more. We're seeing more. We're talking to more people thinking about this. And so I ended up getting this kind of industrial space. I call it the garage. It's 20,000 square feet of industrial space. I'll show you in a little bit what it looks like. You know, San Francisco real estate is not the best. I think this is the time to make these bets. But I've talked to 25 to 50 founders in the last 12 months. that need a hardware space like this to tinker, to have smart people around them, to have the machines around them, to just being able to develop it probably for less than $100 grand. They can go and proof of concept and prototype the thing. And look, we're seeing the hype cycle of like humanoid robots, right? That's that like hype above. I'm more excited about all these startups that are going to form. They're going to build this expertise. They're going to be $100 million teams, whether it's their product or their knowledge.

Right.

That's happening now. And 2 people are doing it like every single day, like they're spawning these things. And so consumer hardware is one area, but I think robotics and what typically was known as like deep tech, like I need like PhDs and like stacks of them and $50 million or $100 million, that's the second one. And I think it's like there's a few people seeing it. Maybe there are sectors like defense where it's like exciting, but like I don't know. I think we're going to see machines everywhere and every kind of version of it. And I think I've seen people develop like cooking robots, laundry machine things, folding drones to inspect stuff, drones to map interior spaces. Like Matterport is a giant company and a human goes and puts a tripod everywhere.

Okay.

And like, this is a little drone that's going to fly through the whole house and like map it for you. And that expands that. I just see 1 or 2 person teams able to do this faster than I've ever seen it. I feel more capable. Maybe I just really want to get into hardware again myself. And—

Well, I think that's the point though. It's a lot of people would want to mess with this. And if something goes from not tinkerable to tinkerable, suddenly it's now the, like, it's kind of like before you're competing 1v1, now it's 1 versus the field.

Yeah.

The field, of course, any individual thing in the field might suck.

Correct.

But the field overall is super powerful. And now you're saying, The field is open for, you know, hardware tinkerers, which it wasn't open before. That's a big deal.

Yeah, we do these like residencies here where we'll bring people here for a month or 6 weeks. We just tell them like, here's a theme, here's a space to go do things. When the Vision Pro came out, we did it. We had about 40 Vision Pro devs. We probably had the largest concentration of Vision Pro devs out of Apple, right? Anywhere. And the knowledge of that, like, look, it didn't produce some ridiculous outcome. We actually did invest in 1 or 2 teams from that. I think we just got a chance to see that technology deeply. We did this AI hardware one. I remember meeting this, this kid, and I don't know where he was in the world, and he shows me this like robot he like built, like a humanoid he built like in his bedroom. And it's like half— he only built the bot like the legs, so it stopped like where the waist was, and it was like totally hand constructed. And I'm like, yeah, this is wild, but the fact that you could like bootleg this in your bedroom and like it could take steps. It's like crazy to me. I met these other guys, Premier, and the same trend in Raspberry Pi. They took that trend. And I mean, some of these, I just, you know, I love it because I could jump on the video and they're— they were up in Seattle or something and it's like their kitchen. And I'm like, is that a 3D printer in your sink? And it's just like their kitchen is like all machines. And I'm like, dude, you got to get out of your house and I got to give you a home to like do that. That's what I want is like people who will just turn their bedroom into this. It's like, no, no, no, you come do it here. We'll give you a little bit better facilities. We'll give you space. You get a better place to sleep. And they took this like Raspberry Pi trend and they're like, oh, people start with Raspberry Pis and then they build a custom board. Oh, we're building a modular Raspberry Pi thing, right? We take that compute module.

Next intermediate step.

Oh, you want a speaker? Well, you could pick whatever speaker you want. Just put that in. Oh, you want to produce 2,000 units now? We have the ability to scale it for you. So they took the hobbyist world and the behavior that's happening, driving it, and they started this in there, literally in their kitchen. And like, I just think that, man, that's capable now. And that's the trend that I can't, I can't like unsee it sometimes.

Right.

And in hardware, both in consumer and more deep tech hardware, like drones platform is massive, right? To be able to do things, so much opportunity to utilize that. I don't know how long it's going to take for this to become mass and mainstream, but I just keep seeing that trend right now. And I think for us, that's the bet you're making. That's the bet we make. And so I always joke with the guys here, it's like we want to bet on -1 to 0. Like Peter Thiel talked about these 0 to 1 companies. It's like we still want to step even before that. Like you're at -1, you're wandering the forest, you need a place to tinker and like come and get the ideas to form. This is what we want. This is what we want to do, right?

I love it. Furqan, this is amazing. Good catching up as always. Let's go check out some of these spaces.

Sweet. Let's do it. Thanks so much.