He Turned Down Google's $100M Offer, Then Became Chairman of OpenClaw

Dear listener, I've put myself in the hot seat with somebody that knows everything. My boss. Twice. First time when I was 19 years old at Apple. The Apple Group was actually Group ID1, the first social media campaign of all time. We want $200 million. You need to replace the camera icon on every Android phone with the Path icon. The TechCrunch headline says, A hundred. We turned down that offer. Both of our phones started vibrating. All of the text messages said the same thing, which is that Steve Jobs just died. Within 24 hours, I had the same feeling that I had when I first installed Facebook. I felt like I could make anything. I made up a lie. Sorry, Dustin. But that's a classic Matt Van Horn move. Tomorrow was invented in the last 30 days. It's the Last 30 Days podcast with Matt Van Horn. Where the builders tell their stories of the future being born. What's real? What's next? Who's breaking through? The hottest minds in AI. I'm talking straight to you. It's the Last 30 Days podcast. Hit the tomorrow today. Matt Van Horn on the mic, leading the way. Ship, ship, ship. Learn, create. Tomorrow was invented in the last 30 days. Into last 30 days on the news. We're starting with Jev versus Klef. There's been a lot of people who love Jev, and now we're comparing it with Klef that just came out. These are two products in the same category, decision models that return type choices and probabilities instead of text. Jev from TypeSafe AI came on September 15th, and just a few days later, on October 1st, came Klef, which I have to say, I have not used Klef yet. I'm deep in Jev land, but I have not tried Klef yet. Jev, the community is positive and curious. Jev pitches itself as fast, structured AI decisions for software automation. I like this quote from Sam Raj summed up the mood: This type of intelligence is getting commodified very quickly. Klef was still much more expensive and not as smart as our testing, but I'm sure it's just a matter of time. So the strengths of Jev: huge mainstream reach for a brand new model class, very, very, very cheap, strong on judgment calls, and it's already being recommended in papers. Klef: excitement about OpenWeights, skepticism about price and real-world performance, says the community. Cloudflare pushed Klef as the smarter and faster than Jev. The audacity. This product didn't exist two weeks ago. Would love if you could show Trevin put together a nice video. Do you have that demo video? So he built this thing called KlefCam. It's looking for gestures out of his camera: a peace sign, a thumbs up, an open palm, and will it recognize the object? All right, got the peace sign matched. As someone who spent a decade building computer vision software to recognize most commonly cooked foods, the fact that Trevin was able to build this over the weekend in a few hours is mind-blowing to me, considering how hard it was to make the June oven recognize food. Can you imagine if this was around, Matt, now when you were building June oven? I mean, we'd just be able to do so much more. It would be— Incredible what we'd be able to build. Does it see Luigi? Yep, there we go. Now it found Luigi. And I asked Trav, like, did you train it on this mug? And it's like, no, it just knew what Luigi was. Jev, when head-to-head with Clef is a bit faster, a little bit cheaper, and can read tone. However, Clef wins when it comes to, obviously, looking at images or video, which you just showed, and that's amazing. And it's open, right? Open source. So that means that you could ideally download that if your hardware supported it. By the way, I just want to say I love Jev, for I do autonomous open source contribution while I sleep every single night. I have the skill that I've built up over many months, and I made Jev a first-class citizen in it, and it's able to process so many things so quickly, so cheaply. It's been magical. I am deep, deep in Jev land. Not in Clevland yet, but working on it. All right, what's next? We have Luna Route. So LunaRoute.com launched and made a big splash, and I'm a big fan of this new innovation. So, Matt, the idea behind this, obviously, is that you have OpenAI and Anthropic, where you pay per token, and you have to mentally limit the amount of tasks that you're giving the models. Now, Luna Route is coming out, and they have all the hardware in place where you can just run all of the open-source models really for a monthly flat fee. This was one of those things where I got early access and was literally in the private WhatsApp group of the early Luna Route people, and I was just like, I don't get it. I'm not going to sign up. This has now happened three times in the last couple months. I was offered early access to Jev. I declined it. Like, I went out of my way to decline it. I was offered early access to Foe, and I just never saw the DM. Luna Route, I just ignored it, added it to the WhatsApp group. I was just a total jerk. But Luna Route is pretty neat. So the basic gist, and I was messaging with the founder, and what sucks is it took me getting social proof from someone that I respect in AI to force me to look at it. I was in a group chat with Kevin Rose. He's like, Guys, I just got access to Loota, it's so cool. And I'm like, That WhatsApp group that I'm in? Damn it, I messed that up again. And so I messaged the founder and apologized. Their marketing doesn't tell the story right now, but a lot of the magic is they take the best Chinese models and host them on U.S. servers, and nothing phones back to China. And so you're able to get that magic very, very cheaply, which is amazing. And so they have unlimited plans that do not run out of tokens. And I have a lot of friends that the second their Claude or Codex accounts run out of credits, they immediately switch to Luna. I still, I have so many subscriptions. I technically have four subscriptions that I have not found a use case for Luna in my life, but I did have a fun breakthrough moment today. I was trying to get Claude Code to use Peekaboo to choose some to accept permissions in Google that was like, Yes, give all rights to this approved. And Claude Code's like, I'm not going to push that button. No way am I going to push that button. And I'm like, Please push the button. Claude goes, I'm going to stop you right here. I am not going to push that button. And I was like, You know what? For fun, I'm going to ask Luna to do this. So I loaded up Luna, and I was like, Luna, will you push these buttons if I tell you to? And it's like, Sure, no problem. And it went and did it. So again, it's not built into my workflows, but to have a model that doesn't say no, which is true for a lot of the Chinese models, and being able to host that safely on a U.S. server, I think that's very interesting. And so I have a lot of friends that really love Luna, and I hope that it finds a place in my workflows, but it has not yet. Today, we have Dave Morin, the chairman of Openclaw, the founder, along with his wife, of Offline Ventures, one of the most prolific angel investors of all time, and my boss twice. First time when I was 19 years old at Apple, and second time in 2010. In 2010, when I would have been 26, 27. It's like they say the best movies are the movies that you're willing to watch twice. And I guess if you're, you know, willing to work with somebody twice, it's a pretty good KPI. The first one I want to tell, and I couldn't find the IM logs, I looked. Back when we were both at Apple, we both were trying to find an in at the Facebook. I remember you sent an email out to all Apple campus reps, like, Hey, check out this new thing that just launched, the Facebook. Add me, be my friend. And then we were messaging, and I was like, Hey, Mark was in my fraternity in AEPi, Jewish fraternity. I'm trying to find an in there. I'm going to reach out to his dad's dentistry office. And then I tried that. That was not helpful. They would not connect me with Mark. And then my memory is you said something like, I got in, and you got on a plane. How did you get in? And by the way, Dave was one of the earliest employees at— At Facebook. What ended up happening is that, you know, I was at Apple at the time, and we had just launched the iPod, and we were trying to figure out how do we increase our market share, particularly with college students. You know, we only had like 2% market share. If you guys remember the ads on TV at the time were, Dude, you're getting a Dell. So we were trying to figure out, like, how do we get, you know, much larger market share? We had this new device called the iPod. We had this new music store called iTunes Music Store, and you could buy songs with one click, put them on your iPod. We had this harebrained idea that, like, maybe if we could convince all of the colleges to record all of their lectures, that students would then want them on their iPod. And so we were trying to figure out ways to build community on college campuses. I was really obsessed with this idea of social networks and building online communities. And so I had been signing up for every single social network that had ever started. You know, there was like LinkedIn, and there was Friendster, and there was MySpace, and I was just really into this idea. I'd used AIM all through college, and when Facebook came along, I very quickly, I mean, within the first day of using it, I was like, This is going to change the world. And one of the reasons why was that I was transported back to my dorm room my freshman year, and I remembered being in my dorm room using AIM, trying to figure out who was on the internet that I could talk to that was like me, that had similar interests to me, that, you know, even just went to the same college or was in my class, right? And the first thing I noticed about Facebook was that the AIM screen name was above the fold on the profile, and you could click it, and you could instant message anybody on your campus, and you could browse around the directory of every profile that was on your campus. And so all the people who were nameless, faceless— list that you're walking by every day on campus, you could now send a message to and, you know, get to know. And you could see which were their favorite movies and their favorite books and what you had in common and which friends you had in common. And I was just like, this is game over. So I very quickly went to work trying to figure out, how do I work with these guys? I want to be their first partner. Ultimately ended up talking with Mark, talking with Dustin, talking with Sean Parker, talking with Kevin Colleran. And through this conversation, we came up with this idea of that Apple should have a group on Facebook. I kept saying to them, I don't want to buy ads. I want to have a community. Like, you guys have all of these profiles, and you have all of these interests on those profiles. I want to be able to, like, be inside of the Apple interest, and, like, I want to be able to give students free iTunes and free iPods, and I want to, like, cultivate our community inside of Facebook. And so that was what started—that was sort of this, turned into this conversation and the creation of Facebook groups, and how the Apple group was actually group ID one in the first Facebook group. And I think you can kind of say today now it was actually the first social media campaign of all time. Like, there were no pages, there was no following, there was none of that kind of stuff back then. And I think it was like us and Paramount, maybe, whoever was doing—there were a bunch of movie releases around then, and I think it was SpongeBob SquarePants that was the other one that did this, and then Victoria's Secret, I believe, was the third one. And so that was kind of the genesis of the original conversation and how it all began. I remember in 2007, I was at the first F8 conference, and I met your wife for the first time. I think it was your girlfriend. I think she was still an undergrad at the time too. And I was there with Logan, and we were working on Zimride slash Carpool. Yeah, I mean, we tried to—I think we had 100 exactly launch partners for Facebook Platform after we had spent, I think, around nine months building it. And we tried to have really interesting partners, interesting use cases. And I remember you and Logan were launch partners. Jack Dorsey and Evan Williams and Biz Stone were launch partners with their new app called Twitter, which was a way to update your Facebook status using text messages. Aaron Levy had this idea called Box, which was a profile box that you could put a file in on your Facebook profile. What was funny is we didn't have a single— Launch partner that was a game. And within a year, everything on Facebook platform was games. You start Path with Sean Fanning and Dustin Moskovitz, two of my favorite people in the world. You invited me to interview at Path, and Dustin—how do I describe Dustin? So he's, you know, top three most talented designers in the world. When he likes you, he's like an amazing human and funny and social, but he does not like being social the first time. I don't know how else to describe it. I walk in the office—it's like an apartment—and there's like no one there. 40th floor of the Infinity Tower. It's Sean Fanning's apartment. Sean is not there. And I see on the couch in the distance having a meeting is Dave— Technically, Matt, actually, technically it was Path's apartment that Sean Fanning then lived in. So anyway, I walk in for my interview, and I see Dave meeting with Adrian Grenier, the, like, Vinny Chase from Entourage, is just, like, chilling on the couch. I don't know who Dustin is, and he's like—I'm like, Hi, like, here for, like, Path or something, like supposed to meet with Dave. Like, Is that Vinny Chase? And Dustin, What? I forgot about this. Is that Vinny Chase? I don't know. And I'm like, okay, sorry to bother you, sir. Like Dustin was clearly wired in and did not want to be my friend that day. He's my friend now, for like, I finally broken through. And then Dave and I met for a while, and he's like, Awesome. Well, like, yeah, I'd love to work with you, Matt. This would be great. We're gonna take down Facebook. We're gonna take over the world. And then I'm like, okay, cool. He's like, Oh, but you have to convince Dustin that I should hire you. I'm like, the guy that didn't want to talk to me? And then I'm like, crap, what do I do? And so I made up a lie. Sorry, Dustin. I tapped him on the shoulder again, and I was like, Hey, Dave says we need to go for a walk and have a meeting right now. You did not say that, for the record. You just told me you couldn't just hire me. But that's a classic Matt Van Horn move. And he goes, Uh... By the way, I messed up. I did not pull a Matt Van Horn right now. I was literally eating sushi downstairs very fast, on the phone with Lauren, like trying to shovel it in my mouth. We've got the pod, and freaking Steve Ballmer just walked by me right now, like 30 minutes ago, and I'm like trying to explain who this is to Lauren. I'm like, and she's like, What are you doing? Go chase him down. And I was too slow. There's no happy ending to the story. So maybe I am losing my touch, Dave. Oh, I don't know. Steve Ballmer. You got a busy life. Well, we met with Steve many times at Path, so you've already done that meeting. I was actually on my honeymoon with the Steve Steve story. Oh, want to tell that one, Dave? So I was on my... On, Dave. So I was on my honeymoon when Steve Ballmer came by Path and was literally in a meeting with Dave. I was not there. I would have been. Thank you for always inviting me to cool meetings, but I was on my honeymoon. What happened, Dave? Many amazing interactions with Steve Ballmer over the years. You know, he was a big partner of ours at Facebook and, you know, was very helpful during the Path years. And I think that they were investors in Path. I can't remember. I'd have to go back and look at the cap table. It's been too long. But he was, or maybe he was there and we were talking about it. And we were sitting in the boardroom table, and both of our phones started vibrating and kind of didn't stop. And you know how sometimes you're like, Oh, that's annoying. There's just like one call or something like that. But it was strange that both phones were going off. I turned over my phone. Both of our faces kind of, I think, went white because all of the text messages said the same thing, which is that Steve Jobs just died. Yeah, we ended up sharing stories for the rest of the meeting, actually, about Steve. And, you know, I was lucky enough to work with him and have a relationship. And in a lot of ways, I look at my life pre and post that moment because a lot of things changed. And, you know, my current business partner—I have two business partners, my wife, Britt, the other one is James Higa, who worked with Steve Jobs for 30 years. And I don't think I would have met him if it weren't for that moment. Now onto a different story that I don't know if you ever talked about ever. Dun dun dun. Dave's look, he's like, This jerk Matt knows me way too well. Here we go. Yeah, here we go. Dear listener, I've put myself in the hot seat with somebody that knows everything, so we'll see where this goes. I work at Dig, okay? And I'm bored. And it's my final days there. I don't know that I've—I don't think I'd officially been recruited to join Path yet, but, like, kind of my final time there, it was kind of happening, and I was a little bored. And I just decided, I was like, you know, I want to meet interesting people in the Bay Area. And so I just started cold guessing people's email addresses. Classic Matt Van Horn move. You actually need to do a podcast on how to do Matt Van Horn things, because I think that many of the new founders coming up, like, you have—you are an encyclopedia. Like, you could write a book on the kind of— Ways to do, like, really entrepreneurial things that are extremely creative, like this, like, you know, just guessing email addresses for people that you want to develop a business relationship with. Like, you've just, you've got to library these, so maybe you should just change this podcast to Matt Van Horn's library. So I reached out to, like, Michael Moritz at Sequoia, who was not an investor in Dig or Later Path, and I was just like, Hey, Mike, would love to discuss the Dig strategy that could be valuable to your portfolio companies. We're about to launch something new. And he was like, Sure. And then I kept doing this, met with Ben Horowitz, met with kind of a slew of interesting people, and everyone said yes and responded to my cold emails, including Andy Rubin, founder of Android, and who was at Google. And then with the Andy one, I didn't think I was cool enough, so I used my boss at the time, Kevin Rose, as bait. I was like, Kevin and I want to discuss our Android strategy with you. We'd love to come down to Google. He's like, Great, let's do it. And so it gets scheduled. Kevin and I actually get in the car, we're going down there. And by the way, just as, like, a side point, just what a different era that was. Like, I actually think that people are so overwhelmed now because of agents and social media and all this, and the calendar is just so full that the Valley is not as much—it's not as porous anymore. Like, it's very hard to, I think, get those types of meetings anymore, which is maybe kind of a bummer for people coming up. But that era, you could really do that. You know, whether you guessed the email or not, or you kind of asked somebody for an email address or an intro, like the sort of casual coffee, like, let's get to know each other and see how we can help each other, maybe. It's an artifact of the past. So back to Andy Rubin, though. So we're in the car, we're driving down there. So Kevin and I are in the car, and we get the email, Sorry, the meeting has to be canceled. Andy got called into something. Thing. And so, and that was my last week at Dig. And the meeting gets postponed another four or five weeks. And instead of being like, Hey, I'm switching companies. Is it okay if I bring my new boss, Dave? I'm just like, I'm keeping this meeting. I am not modifying this. So I forgot about this. I distinctly remember going to Google with the giant—they had like the giant green robot in front of the building. I have this, like, very distinct memory of going to this meeting with you, but I don't remember this context. I'm like, Hey, Andy, so good to see you. Really excited to talk about Path. And at the time, Path didn't exist. We had a beta with maybe like 20 people on it, 50 people on it. Was it that early? We had no users. Yeah, I think we had no users, no revenue. We were in Sean Fanning's apartment. We definitely hadn't launched it yet to the world. I'm 100% positive. Okay. And then we go in there. Andy has no idea that I bait and switched the meeting, that I brought Dave instead of Kevin. Busy man. No idea. And no, it was totally fine. But he, at this time, right, everyone was desperate to beat Facebook because Facebook was eating everyone's lunch, and Path was the best bet in the world on that. And Google hadn't announced their social product yet. And so Andy's like, Oh my goodness, like, have you talked with Larry? Have you talked with Vic? Have you talked with Bradley? We're like, literally, this was a Digg meeting, and I just brought Dave. And we're like—and we didn't even, we were so—we thought we were so cool, we didn't even show the product to him, even though we had the beta on our phone. Like, no, we did not show him the product. We kept it vague and secret of what we were working on. Fast forward, we're having secret hotel meetings. We're like picking up keys in lobbies, meeting in secret teams. Next thing I know, it's a couple weeks later, and we're in another room. It's me, you, Larry Page. And David Lowy, next thing we know, Dave, you finally folded. You'd put out your phone, and you handed your phone to Larry and let him see it. And he's like—and he kind of like shrugged. He's like, Cool. Like, he didn't really seem to have much of an opinion. He was talking about carpooling. It wasn't that good. Oh yeah, no, he was talking about—I remember this distinctly. He was talking about Waymo. It wasn't called Waymo, but he kept talking about how, wouldn't it be cool if there were robot cars driving around Palo Alto, and that you could just, like, get in them and drive around places, and that that is definitely how the future is going to be. And I was like, Yeah, that sounds awesome, but I have absolutely no idea. I remember him saying, like, how cool it'd be if you could, like, drive somewhere in a robot car and have a conversation with a Stanford professor. And by the way, that was what year? That was 2011, maybe? 2010? I mean, just to— Larry Page is a real visionary. I literally was like, We want $200 million. You need to replace the camera icon on every Android phone with the Path icon, and you need to open up—bring up to code our balcony at our office because we had just signed that lease, and the Leif Faning's apartment, and the balcony was not going to be able to be opened. And yeah, they laughed at me, but they gave a real offer. Do you remember what the offer was? I don't remember. I think it was 150, maybe. Yeah, I don't remember either. The TechCrunch headline says 100. We turned down that offer. Well, I think, you know, we had not launched the product yet. We had a really strong idea of where we wanted to go. You know, we were really focused on this idea that I'd call it the theory of clubs within clubs. As a club gets bigger and more and more people join it— There's a natural human tendency to want to make a smaller club inside the club so that you can have more intimate relationships, closer—you know, you want a group that you understand everybody that's in the group. And every social network up until that point had chased growth and really tried to just maximize growth, like growth, growth, growth, bigger and bigger groups. I think Facebook had, in that two-year time span, the average number of friends had moved from like 150 to 300-something. You know, Dunbar's number suggests that the human brain cannot maintain more than 150 stable human relationships. And we had a strong point of view that more intimate, personal social networking was a really powerful idea, and we didn't think that it was going to work on top of a bigger system. We were really conflicted. We knew it was a lot of money for a unlaunched startup. And still to this day, now that I've invested in thousands of companies, and I've now—you know, I have the real benefit of hindsight telling this story now, Matt, because I know now what I did not know then, which is how rare a huge win of that magnitude is. Like, it just—it still doesn't happen very often. There's sort of a capital reality at play, which is that, you know, now with my investor hat on, I understand that the investors that were involved were looking at this like, wow, like, you know, you guys started this thing in January. It's like October, and you don't even have a product launched, and you've got 100, 150 million on the table. Like, that's— I think we had raised like maybe five million at that point. And I think, if I remember actually correctly, there was a Series A that had started, that there was already a Series A process underway. And so the choice was not between roadmap, vision, and taking the offer. The choice was actually between roadmap, vision, and then how to best execute that. And it was either take the offer from Google or take the offer from Index Ventures and Kleiner Perkins to do the Series A and launch the product and go on the journey. And I've always been a bigger fan of go on the journey, I guess. I still am. It was not easy, though. One of the things I learned actually as a founder was that I was really torn up about this, and I kept asking for advice from people. And every time I would ask for advice, you know, it would be a day later and a bunch of things would have happened, and I'd have to re-explain the context to everybody I was asking advice for. I'll never forget, actually, Mike Volpi, who became one of our board members, called me and said, Dave, it doesn't matter what you decide. Only you know. Only your intuition knows. And you can't keep trying to explain all of the context to every person that's an advisor to you over and over again, because only you know what the right choice is. And so you're going to have to make the decision. And that was a pivotal moment where I was like, okay, my intuition is that we should keep going. We have a lot to give to the world. We have a, I think, very unique point of view. And I think we have not just an idea and a roadmap here that's interesting, but I think we've got a group of people with a very distinct point of view. That deserves to be expressed at much larger scale independently, and not be influenced by a, you know, large outside energy of some kind. And so we chose to turn it down and go it alone. I don't think I've ever said thank you for doing that, but I'm about to. So even though it would have been great financially at the time, and it felt insane that you made that decision at the time, the thank you I have is I never would have met Nikhil, my co-founder of June, and co-founder of my new company, because he hadn't joined yet. He joined a couple weeks later. He would not have been excited about joining Google from Apple. That wouldn't have made sense. So thank you for turning that down. We were able to create a lot more economic value just in June than had that happened. Yeah, and I think this is one of the things that it gets talked about here and there, but not everything is about money and— Power and distribution. Sometimes doing art and expressing yourself and getting a group of people together with a distinct point of view that's different matters. And the gratitude goes both ways, Matt. I mean, I think one of the things people don't know about you is that you're one of the best recruiters alive. And, you know, I think that one of the things that you and I did so well together is that we really worked, I think, beautifully well together on recruiting and building a team filled with individuals that each had really interesting points of view. You know, we hired a lot of amazing people. People ask me a lot, like, do you miss Path? I can't tell you, like, every week, every day almost, I get on a Zoom and people are so kind and they're so wonderful about Path. People have—I mean, I was at a dinner two nights ago in San Francisco, and I had a guy who, he runs a venture fund that today is much more successful than any that I've built. And he said to me, like, I gave a—he came up next to me and he said, I just wanted to say thank you. And I said, Why? And he said, Well, I was in college when you were building Path, and you gave a talk about social networks and Dunbar's number, and I used that talk. It was very inspiring, and I put that into, like, my thesis, and it was part of my—you know, I gave a presentation about what you talked about, and I don't know if I'd be in tech if it weren't for that. And I think he said that was 2013. What I say to people is it's a bummer, sure, that the product's not around, but I actually miss the people more, and the creativity and the point of view and the design, the message, the engineering, everything that we did from the moment that that— You know, that decision was made until 2015 when we had to sell the company. It was a beautiful thing. Like, we created a really awesome band, and it was you, Matt. Like, you were a huge part of building that team that I really view that as a defining creative period afterwards. Let's talk. I feel like we're on, like, what is this, the fifth stage of Dave, right? There's the Apple, the Facebook, the founder, the investor, and what the hell is Dave going to do next, right? He seems like he's got a good gig. Out of nowhere, you become—you're the person that convinced me to install OpenClaw, and this was, it was called ClaudeBot. And I remember I sent you a text message, and I'm like, this is in January. This is long before you were chairman of OpenClaw. And I'm like, this thing's interesting. Like, what the hell are you using it for? And then neither of us actually said anything interesting that night, for the record, but we knew it was special. I remember you, yeah, asking me something like, What's, like, a use case that even matters? or something. I still do that. I had a friend a few weeks ago was like, Hey, do you want early access to Dev? And I'm like, Give me a use case that matters. I don't get it. And he gave a pitch. It was Hiten Shah, and I was like, I don't get it. I'm going to pass. I'm a late adopter, I guess, and now I'm obsessed with Dev, obviously. So anyway, I need to listen, shut up in my own head and adopt the things, even though I'm still a pretty decent early adopter. Okay, so my question for you, Dave. So you're chairman of OpenClaw. I want to hear about that, but I want to hear more the how the hell did you get in there, right? If I think about, like, you talk about the old Silicon Valley where you could cold email someone and they respond. Peter Steinberger, the hero of open source. How the hell did you convince him to be your friend and you become the consigliere to the most important person in this crazy agentic era? And I get it, you're awesome, but I'm still not connecting the dots of how this happened. I was at home over the holidays. I'm from Montana, and I was up in Montana visiting my family, and I'm in a few group chats, actually from the go back to, like, the Sean Fanning days in the '90s. Like, I'm in a few group chats with friends that I've been kind of— Hackers that I've known and just people, like curious tinkerers. One of those guys pinged me and was like, Hey, have you seen this Claude bot thing? You got to check it out. And I had been having this running conversation with a few people in this group for two or three years, actually, since like the beginning of ChatGPT. And some of this goes back actually to the Path years, Matt. Like, you know, if you remember at the tail end of Path, we actually built Path Agents, and we had a product called Path Talk, which had what I back then called artificial artificial intelligence, which used humans in the loop to produce an artificial intelligence-like agent experience. And zooming back forward to December, I'm like, It's holidays. I'm tired. Like, I'm just chilling. I don't know. I'll check it out a couple days. And when I installed it, within 24 hours, I had the same feeling that I had when I first installed Facebook. I felt like I could make anything. That was the really powerful feeling to me, like this idea that it's mine, that I can turn it into anything, I can create anything with it. You know, my entire arc has always been about kind of personal technology. Like, that's why I liked working at Apple. Like, I love this idea. I love the feeling of having my own computer, right? That I have my own. And OpenAI was the first time in AI that I felt like this was mine, and I could configure it however I want. I can put it on my own computer. Like, what a powerful idea Peter had, this notion of putting the AI inside the Mac and, like, letting it just have access to all the tools. Like, that's such a cool idea. Within a few days, I was like, I got to talk to whoever created this because I don't— It was, it was, I wanted to talk to the person out of pure gratitude. And so that's what happened is I reached out to Peter, and we started talking. I want to help you protect this and get it to everybody in the world. And we just started working together. Why? Why were you able to break through? Like, you explained how, but if Peter were here, what would he say? Why did he trust you? I trust you completely. You're one of the most genuine, incredible people in the world. Why do you think you worked out so well for Peter to trust you? We both had a good sense of each other's point of view and values pretty quickly. I think, you know, Peter also went through some very difficult entrepreneurial experiences. Choosing to do open source is kind of entrepreneurship on expert mode. We both knew that going in. True passion does not come through unless you are authentically interested in it. I think I was and still am deeply interested in this idea of personal AI. And by that I mean not an AI that's running in someone else's cloud, not that's designed by some other company, not that you have no control over but is called personal AI. I mean, like running on your own computer. You have total control. Like, I think that this is a really important idea in the AI age. And I think people deserve to have free, open source software that they can control, that they have total agency if they want it in the AI era. That may not be for everybody, but I think that it's an incredibly important thing that needs to exist when everybody's scared that this and that and the other thing is going to happen from AI. Like, I think that this mission that we're on is very aligning if it's something that you truly believe in. So I think that's the other thing is that I actually believed in this straight out of the gate. Dave, you're an investor, you're chairman of OpenClaw. You live and breathe and see this world every day. Right now, it feels like on X, everyone is just launching something and everyone else copies it, and then everyone has the same product. The roadmap is just like what people are posting on X. Like, what do you believe that the industry doesn't seem to believe that is not being posted on X today by everyone else? What we just did in the last 30 days, I think it's maybe it's a little bit more than 30 days now, but like we launched OpenClaw 2.0, and this took us over two months, which in AI agentic engineering land is an extremely long time and a lot of tokens to build. And the reason why is that we have spent a lot of time talking to the biggest companies in the world about what teams need in their workflows. Most of what's going on in the ecosystem right now is Individual personal agents that I use one-on-one. What we are working on at OpenAI, and we have been now for the last two or three months, is multiplayer agents. And this is an incredibly difficult problem. There's, like, really serious problems around multiplayer stuff in how AI works architecturally, which, Matt, I know you know, and a lot of your listeners will know. We've been focused on enterprise and security and making OpenAI the most secure agent in the world. So multiplayer, I think, is the thing that happened in the last 30 days that nobody's paying attention to. So, like the text I sent you, which I pulled up, by the way, because LMs, you can pull up text messages now. I asked you, OpenAI, what are your top three use cases for Claude Bot? So, Dave, what are your top three use cases for multiplayer? We have this amazing thing happening now where people are producing, you know, running a agentic engineering session, working on an idea, and now design can jump into that thread and add value immediately in flight while the session is playing out. I think it makes the design-engineering collaboration, like, completely different now. It makes it a true collaboration. Like, we are both in the same session, working on the same feature and improving it before it even goes into a PR. That, to me, is, like, my number one maybe favorite use case. The second use case is we have seen there used to be this thing in companies and teams where, in order to get a dashboard, you know, like an internal app or something that improves the collaboration between the team, you would have to have, like, a meeting for it, convince some engineers to work on it. Now we've got this dashboard feature where you can just generate a new dashboard. It shows it. It puts it on the left sidebar. Anyone can look at it. We've got all these cool dashboards that different people have made for, you know, viewing the data, and everybody can have their own now. And the ones that are the best stick around. If you want to improve it, you can just jump into the chat related to that dashboard and improve it. It makes collaboration around a giant software project just, like, way easier. Obviously, like, there needs to be a lot of iteration. I think there's still a huge uncanny valley out there in getting this type of working to, like, everybody in the world. But at the frontier, if you're doing, you know, agentic engineering and design and building software systems, like, I think what we're doing with what's currently on main in OpenClaw is really, really cool. All right, last question for you. What's a toy, but, like, what's, like, a new thing, like a shiny object in your life? It could be software. It could be something physical. It could be a new pair of headphones. It could be a hat. It could be your Mac SE behind you that's blurred. Like, what's something new? Because I feel like you always have cool new gadgets, toys, things that you're just like, yes, this thing is cool. I'm in love with these hats from Varsity. Of all the things that I collect, I'm a, I guess I'm, my wife makes fun of me, I'm a serious hat collector. Whatever your, you know, style of hat may be, if you like a sport, you know, structured hat, or if you like the, you know, this one I'm wearing is a soft, unstructured dad hat style. They have amazing materials, construction quality. They have different head sizes. The other thing I'm kind of really into lately is the iPhone Air. I just, like, have really—I got one this summer because a lot of my friends that have good taste, really, they love it. And I was initially—I'm a big photographer, and so I don't like not having a great camera. But now that I have one, I understand why people love it so much. I'm probably going to switch to the Duo when it comes out, but this is basically the design. This is the Duo. It's just two of them folded in half, and it's smaller. It's like if you chopped the top off. What do you think Steve Jobs would think of the iPhone Duo? It's amazing in person. I tweeted this the same day. I think it's easy to hate from afar when you haven't played with it. It feels like actual paper, and when you unfold it, it, like, click—it's like, it's so high quality that you're kind of like—your brain is like, okay, Apple did it. I've tried the Pixels and all the other folding phones, and they're terrible. They all have creases in them, and it's just like, ah. And Apple did it. They made a folding phone, and they even redid the OS from the ground up. Like, they had to create a new interaction model because you can't just have an OS that is the same thing, which is what Android did with their folding phones, and it's got all these problems that break. And they actually created this new interaction model where the right rail is now the navigation. And one of the coolest things that I noticed about it, you'll see when you use one, like, when you have it open in horizontal mode, you can flip with your thumb. You can, like, have the screen be split into two, and then you can go, like—you can, like, flick an app, and it will, like, flip it to the other side. Or it'll flip it into the top right corner, and it's, like, so good. You're like, oh, this is awesome. Like, this is gonna be, like, really, really good. Like, everybody, I don't know if you guys do this, but, like, I do this thing where, like, I enjoy, like, watching a great show, like Silo, you know, like high-quality TV, and, like, I like talking to Matt Van Horn or my friends while I'm in bed or relaxing at the end of the day. And it's gonna be amazing for, like, watching a show and, like, texting. It's gonna be, like, psycho good at this. And so, you know, I don't know. We'll see. Is it gonna be annoying to carry in your pocket? I don't know, but I think it's definitely gonna replace the iPad. And Apple also just rolled out the ability to transfer SIMs between phones seamlessly. And so, like, my theory is actually they want you to have a day phone and a night phone. I was gonna let you go, but you said day phone, night phone. Dave, what does that mean? Well, I got slayed in, you know, the early 2010s for—back then, I carried two phones, actually, to test. It was actually before the Watch came out, and we were trying to create the Watch experience for Path. And so I was carrying two phones because I wanted to have one phone that had nothing on it and that I could pretend was a watch, and then one that had everything on it. And I jokingly called it my day phone and night phones. And I got, like, massively canceled by Silicon Valley, and people still say it to me to this day on Twitter. But now, like, so many people carry multiple phones that I'm just, like, was 10 years too early, clearly, just like Path. You got in trouble, canceled for that article. You didn't do anything wrong other than, like— You were a rich guy with cool technology. Like, that's it. Like, that's it. Like, you did nothing wrong there. You didn't say anything wrong. You just, like, looked like someone that was, like, had a little too much money relative to the common man. Millennials, day phone, night phone. No, dude, like, millennials, we're real assholes to each other on the internet. Like, I don't know what it is about millennials, but we're really—we're, like, true assholes to each other on the internet. I actually admire Gen Z a lot because they just have a lot more fun with each other on the internet. Like, they troll in a way that's, like, hilarious, not in a way that they're, like, truly assholes, like they want to hurt the other person. Whereas millennials, like, we were, like, really into hurting each other in the 2010s on the internet. We kind of continue to. We're kind of like, you know how, like, if you're a millennial, you're like, Ah, Gen X, they're, like, so angry. It's like millennials were just, like, really mean to each other on the internet. Like, I, like, prefer hanging out in the Gen Z internet because it's just, like, more fun. Dave, been so fun. I miss you, even though we still text all the time. I don't see you enough. Yeah, man. I'm happy you're doing a podcast. I'm excited to see where your podcast goes.