S.10 Ep.9: ElizaOS, FOSSRPG, and the Future of Open-Source AI Agents with Shaw

Transcript

Welcome to the Greenpill.network podcast. We are exploring how technology, community, and new forms of coordination can build, fund, and accelerate what matters in the 21st century. You are listening to me talk to builders, thinkers, and innovators pushing the edges of what's possible from open source funding to AI to governance, markets, and regenerative systems. Today on the podcast, I have Shaw from Eliza OS, formerly AI6Z, to talk about building AI agents. I previously had Shaw on the podcast back in December 2024. And we are catching up on everything that he's been up to since founding AI6Z, now Eliza, and building a swarm of agents and people that are working on AI agents. Um Shaw just

shipped Eliza OSV2, which is uh basically uh got a bunch of different goodies in it. We're going to talk about that during the episode. and I think that he's one of the builders that I'm most excited about talking with here on the podcast and then also collaborating with uh as we turn into 2026 and AI continues to be a really super hot theme. So, uh I think you're going to enjoy Shaw. What's going on? >> How you doing, man? Great to see you again. >> Yeah, likewise. So, we talked 15 months ago on a podcast and uh we're back here again. catch me up. How many how many decades have happened in the last year?

>> Oh man, I I feel like an old man now. >> Yeah. >> Um I I literally had to go to physical therapy for like a frozen shoulder, which is a stress related thing. >> Uh you know, there were some good times and bad times, but um now I'm generally feeling pretty good about things. Like I I think um you know, they say like bare market is for builders, but it really is like everyone's just sitting down and building and like every everything is down. there's nowhere to rotate your money into like just put the market away and build stuff right now. So I I you know generally coming out of that hype cycle man I feel pretty good like I and and I we placed some bets like six months ago on like people were asking well what can your agents do and we're

like well we should really think about that now we have these you know even as Eliza's lost mind share like we've we've built some stuff that is pretty impervious and just any agents can jump into I think that's really cool so you know just riding the wave. Yeah, same. Um, yeah, I mean, there's a couple threads that I want to pull here, but, um, maybe we'll just start with how your view of of AI and crypto has evolved and and how has the Eliza product suite evolved? Um, I think we're on Eliza V2 now. So, take us through what you've been shipping.

>> Yeah. Oh, man. A lot. We had, you know, we we tried a launchpad. I don't think it went very well. Um, it's open source if anyone wants to try again. Um, we shipped what we call like 1.0, which was like RV2 because we were just like on 0 whatever beta. I just made this thing on GitHub man. >> I love those zero 0.1 numbers. Like this is so freeing to be able to be like this is 0.1. It it might work. >> Yeah. You know, I was just like getting a lot of a lot of PRs, a lot of attention. I'm like this is this is like, you know, a beta. Um, we launched that. I felt like that solved a lot of the problems of the first thing but then it created another set of problems that I didn't really understand like oh we

want people to have tooling and you'll launch multiple agents and all this stuff and like that started to really get in the way of of the kind of productivity we wanted. So I started working on what's 2.0 or like I guess V3 and we released this for this is an alpha and I've been shoving this into all of the apps we've been building. We've done some stuff like we're about to release um trading desk.f fund, which um is a OTC desk. Um it's it's we're still like tweaking the agents, so you probably don't want to go try it yet, but the idea is like you can OTC tokens and people can negotiate. Uh you know, your agent can negotiate on your behalf basically. Um I don't I don't like negotiating OTC deals, but I kind of know like I can just have a minimum and

let my agent do it. So that was kind of one of the experiments. Working on two games. We're we're launching a game uh kind of soft launching this week and it'll be like out for the half a million people on the wait list for it which is freaking crazy. Um it's called Babylon and it's a uh social prediction game for agents and humans. Um it's built on ERC 8004. It's basically a copy of X but you can bet on all of the AI versions of your favorite characters. So there's like Nikita Bear with a B AIR and you know Mark AI Andre and blah blah blah and they all like you know argue and connive and then your agents have to like figure out from the timeline how to place their bets. So it's kind of like a new sort of of game. Um really like I I

think uh just kind of feeling like at the time like social prediction markets were going to be a kind of hot interesting thing and it's something that agents could probably win at and isn't just like you know total random noise. At the same time, we're also working on another game um called Hyperscape, which is AI Runescape. It is AI agents playing Runescape, and you can play with them, and you can like bot your account, you can gamble on duels. Um it's it's sort of like you can just like get a bunch of agents to follow you, take them into the wilderness, kill them, take all their stuff. It's like that kind of like very emergent agent gameplay thing where the agents are users just like you. And that's going really well and we're launching that in a couple of weeks. Um, so like this is a

big month for us of of um things that we've been working on for a long time coming to fruition. And um and then the thing, you know, I've these these are like we have teams on this like great leadership and um you know like like engineers and so I've kind of been a little bit freed up to think about like well what's the future of Eliza and like what does it look like in this world of of OpenClaw and you know like that's obviously been the the mind share grabber. when we blew up it was like wow 15,000 stars on GitHub that's crazy man that you know to watch somebody get 150,000 basically like doing the the well packaged version of what we've been doing um was very eye opening and and it it um well for one I just like

immediately sat down I was like okay well I'm just going to build the fully autonomous assistant version of this like I'm I should have always been doing this and we felt very pressured to do more like you know meme coins and and crypto stuff. But I think there's a lesson in there of like just doing the wide TAM like you know assistant thing just kind of blew the up. Um and it also kind of um gave us validation I think because you know like there was just no hype there. No one everyone's like oh that died off that wasn't a thing. And then they were like oh wait no that's totally a thing we're actually almost at AGI I forgot. And then we're like, "Yeah, yeah." You know, and that was kind of a great win in our sales.

And and more than anything, I I think as a founder, I've have to be willing to like throw away any code. Code is like worth I'd say not just cheap, but like basically worth zero dollars now. Yeah. >> You know, it's it's it's worth like 50 bucks in Claude credits at this point, you know. Um Yeah. And and so I I think in that regard, we we just can't be precious. It's like why would I force Eliza into everything? But what I really want to do is not make Eliza everything, but like try to solve real problems with agents. And so one of those is like all of the hard problems can kind of be solved with the lowerhanging fruit version of like games of those. So if you want an agent that trades money, may maybe make a trading game and then train it on that. Um, and then I think there's

also something unique about Eliza with how we deal with multiple entities and rooms and worlds that translates well to video games and I think eventually robots. So like have a robot behind me and I'm I'm uh definitely going that direction. And the last part of this is we ported Eliza to Python and Rust. I wanted to do a lot of embedded stuff with Rust and it was just very challenging um in Typescript to like bring that uh runtime and I did some things like we put an Eliza fully on chain and internet computer and some and we're working on some stuff like that like just you know I don't know kind of a new way of doing agents. Um, and then with Python, that'll really opened up like all the benchmarks. And so we've been able to benchmark Eliza agents and basically say like look with Claude, this is

state-of-the-art. It's totally saturated. All I can really do and say is like if I add actions, does it make it worse because it's so freaking good. And that really just showed us like we need a whole new kind of benchmark, man. Like, you know, like what can agents really do that a model can't do on its own? So that's that's where the games kind of come in. And yeah, that that's that's what I've been doing for the last year, more or less. >> Yeah, I see a lot going on in the Eliza verse and um you know, I I'm wondering there's so much I want to react to there because I think there's so much cool stuff and and we'll put links to a lot of it in the show notes so people can find it. But um you know I think what's coming into focus for me is that is that

you've you've been doing this for over a year now and I think that there's a lot of people who followed you in in there and I'm I'm wondering how the Eliza ecosystem has evolved over the last year or so as the bare market has has happened and um I remember showing up to your uh what have you shipped calls that you guys would do on on I forgot what day of the week it was but and just how many people were building so many different things in your ecosystem. system. Um, and then using agents to do it also meant that you're just shipping so fast and uh, you know, so are there do you still feel like you've got a group of builders around you and um, has that formed into a core group of people that are that are all working together and uh, you know, is there a swarm of

agents that are that are all working together with you and how has that that infrastructure evolved? Yeah, I I mean I really wish that there was um uh Oh, well, let me say this. Obviously, like last January was crazy and there were a lot of people throwing PRs at us and I think that their motivation for throwing PRs at us was really more like hype and getting a piece of that and showing, oh, I just I contributed to Eliza, give me money for my memecoin, blah, you know, perverse incentives, let's say. and and um but I also met some of like the the coolest, most interesting people, but also because of the incentives, I couldn't really work with them. Like they have their token and I have mine and they they got to go shill that thing and do that deal. Um and then everything

kind of dies. It goes to zero, you know, kind of cycles out. And a lot of those devs were like, "Hey, like my token's dead. Like what are you doing? Do you want to do stuff?" And those that that's been like we've kind of had a a more recent generation of of devs that have like hired or commissioned for freelance and stuff and and that all kind of like I'd say are like excaliza devs from the hype cycle that are coming in like already knowing how to do this and and I think there's just this weird like there's a tendency I think for the kind of people who like agents to be a mix of creative and technical which I think plays super well to like vibe coding and and this kind of future that we're in.

You know, it's not like like I'm a hardcore Rust systems engineer. I write five lines of code, but they're the best lines. It's like the exact opposite. It's like I have infinite ideas. I'm super ADHD and I just want to get them all out. And it's like, well, if we could all just work, you know, on the same repo, like this could go pretty well. Um, so, so I've definitely like been collecting, I guess, you know, my my uh strange collection of of um uh people in the network who I think I really fit with. And um I just like I look like a mountain man right now cuz I just got back from a a a two two and a half week um hackathon in the woods. We were just in upstate New York with a bunch of internet strangers. Um you know like Xalisa devs and people you know I

don't know Zo from IQC900 showed up but we just like hung out in in the woods and we like hacked and it was really cool to be like wow this like random internet thing is like very real now and these are my friends and like this is like a vibe you know. So, uh, yes, you know, I wish it was bigger. I wish it like obviously it's not. We still track about 50 teams that are like using a lot cuz they like built a business on it and they're like pretty committed that, you know, a lot of that's pretty heavily forked or or like, you know, altered to whatever they need. And um, but but I think it is like a bit smaller now. But I also think that it was always like people on the sidelines waiting for us to give something that was very consumer

and like like this is this is the story of Open Claws. like they just bridged that last 5% so well and um that does matter quite a lot, you know. So, >> I feel like you've been on a journey and um one of the stops on the journey that I want to just touch on is that you got banned from Twitter for a while and uh now you're back. So, uh tell us your story with uh centralized social media. Yeah, I mean um you know I I think that there is a larger corporate incentive and communication structure that regardless of anyone's good intentions um can cause you know people to get what

they would see as unfairly deleted from the places where they have a voice. Um, and like you know it would be really great if the town square where all of the world's most powerful people uh wasn't owned by one person and I don't have any issues with Axe. I actually like would totally work with Axe. I I you know there's a time when I was like man I want to go work for that company and I spent a lot of time there but um the incentive structure is really bad and it's like you know I after that I spent a lot of time on Farcaster but you know Farcaster is like I I don't know I got to check out Noster and like some of the other stuff but like it's just a little bit disappointing. It's it's disappointing that like as I got there

like they're just really pressed up against the wall to cater to DGens and like I do not really identify with like the gambling DGEN culture, you know, I'm kind of embracing like the bags meta or whatever. Like there's this like mini dev like hey let's fund devs on these shitcoins kind of thing and that's I'm kind of like that's cool. It's interesting but but overall I don't want to you know live in the dgen world. So, kind of lost my but getting into the the current misery of 70k Bitcoin. >> Yeah. Well, as we're talking, the market is dumping. So, uh you'd be forgiven if uh >> don't look >> if you lost your train of thought. But um the uh Yeah. Know the the threat I was looking for is just like getting

banned from X and I think you sued them and now you're back and so like it it >> I mean it's cool to see like you like in action building in public on X. And maybe this is just my bias vantage point because I spend more time on Axen Farcaster, but uh it's glad to have you back. >> Yeah, I it's glad to, you know, it's good to be back. It's like distribution. Um I I really liked the vibe over on Farcaster more. I feel like generally like the EVM chains and like apps have more of a like, oh, we're all kind of in this together. Um and I think I wish that like there's just like X is just like a always a lot of hate. it's just always like a lot of people complaining and being mad and like conspiracy theories and so it's been like a bit of

readjusting back to that. >> Um I don't think that feed social media is going to last much longer both because of like AI slop and just because it's like just really really overwhelmingly asymmetrical and I really just need like an agent to handle that for me and you know kind of intermediate the the internet for me. Uh it's just like wait and I tried to I have like a ad block like a hate blocker but then it started blocking people that I like because it was like trying to use a cheap model and I this this is like a whole project I'd have to really get into. >> Um but um you know I really wish that there was like a decentralized option where I wouldn't have to go and like you know do that. I mean what happened with X was a lot of people used our tech to launch

agents onto X. Um, and I was using another thing on GitHub in our thing, like another open source project in our open source project, which was violating X's toos because technically if you want to, I would argue that an agent should have access to a browser as well, and that like reading the front-end GraphQL that your server serves me is just a browser, and that's not a toss violation. I didn't hack anything, but they didn't see it that way and didn't understand the nuance. And so that's kind of how that went down. And then we like sued them just because they stopped responding to our emails. And at some point I just emailed them internally. I was like, "Why are we both wasting our money? This is stupid. Like let's just figure this out." And so that's kind of

how that like figured it out, you know? Uh much to the against the advice of my lawyers. I just kind of sometimes you just kind of have to >> do the human thing. >> Yeah, totally. Well, um yeah, it seems like you got some war stories along the way. Um, I I have a I have a couple more questions, then I just kind of want to >> turn the tables and wrap about whatever is interesting to you. Um, so let me get get through a couple more questions, then we can do that. Here's one. Uh, at Consensus 2025, you said we're going to automate all the jobs. Um, I'm curious, uh, has the last year changed how you think about the the timeline for doing that or the role that crypto plays in it?

It's one of those things where like the timeline has not changed, but we're here now and it feels very weird and you just like it's like if someone's trying to explain to you like what some cra like what skydiving is like, but you kind of just have to do it once, you know? And I don't I I'm just like in the background like I don't know about other people, but what my job used to be is definitely replaced. I have a whole new job now, which is I I have a microphone and I talk to my computer like 14 hours a day and and basically just tell it when it's doing things wrong. Um, which feels like a very like intermediate babysitting like I am not going to have a job here in a couple of generations for sure.

Like I mean and and so yeah, I guess what's on top of my mind is like like yeah, look, the the open claw stuff is a little bit of a lar still because the models just aren't quite good at like knowing if they did a good job yet, but man, it's freaking close. That's like, oh man. So yeah, like is it going to take our jobs? I think it's going to like in some ways it's just going to be one of those um what is it like kind of uh slowly but then all at once you know like >> you know I mean today like just today like step change okay 5% better model >> that's a lot more hits than misses than the last >> 46 or the new one that came out today.

Yeah. >> Yeah. You know and so >> so you think we're all all at once phase? >> Yeah. I mean, um, no, I think we're just still on that precipice, but it just feels like we are a few models away from like I mean, I like in the background right now I have an agent just playing Runescape like cuz somebody released a Runescape sim for agents. That's like plugged Eliza in like cool, you know, we have we have like we have we have agents doing all the things now. It's just like is it good, you know, and it keeps getting better. >> So, I'm I'm Yeah, I world is changing a lot this year for sure. I mean, I think that it seems like it's a hard time to be a junior engineer or someone who's transitioning because those people aren't getting really a start right now.

And it it feels like the job market is sort of concentrating towards the high-end people who can do a lot with agents and and not people who are are on their way up. So, it feels like it's already happening, but maybe you're right that we're in the precipice of an even larger change, which, you know, would be it'd be I'd be flabbergasted by the magnitude if if we're still in the in the slow runup right now. >> I I'll tell you though, like, show me 10 junior engineers and I'll show you the ones who are going to get hired. And like there is such a stark contrast between the people who are hireable and the people who are not that I think a lot of this is actually about social media media addiction and the fact that a large swath of Americans are basically unhirable. Like they're unreliable.

They're literally coming to work on drugs. They're totally like overconfident and overselling. Like I I I'd say like 70% of the Americans I've hired have like a deep media addiction problem where they're on their phone instead of working the whole time. and and like it tends that like I don't know maybe this is a bias. >> Um and so like when we talk about AI like >> you know uh I don't know like like there are a lot of high skill jobs available right now like there are people just starving like people asking me constantly for engineers but people who are actually focused and talented and finish things are in short supply. So, it's just like like there are like the kind of junior engineer who gets on GitHub and just starts building stuff and learns agents and like is on X, they

are getting hired. They don't even need to get hired. They're the ones doing the hiring within a couple of years. It's the people who expect that like, well, I went to school and I did the thing I was supposed to do and I want to do an engineering because I'm going to get money out of it. Like, you are so not going to make it because you're up against someone who's just doing it because they love it and then they're doing it for free and because it's fun and then they get a job out of it. You know >> what I think is curious about this is that we're you know ostensibly talking about AI agents here but you're talking about agency which is the able the ability to create a dent in the universe which is an attribute of you know any any sort of like living thing or digital

system could have agency but you're talking about agency in the actual people that are competing with the agents for the jobs and so uh that's a term that I was not expecting. I I'd actually say that right now is like the best time to be a junior engineer if you have a lot of agency. If you don't, you are like so unbelievably hosed. You need to re reevaluate what you're doing with your life and like don't waste your time being an engineer because it will not be fun for you. But like like with Hyperscape, which is one of the games, um I just got to with my friend Lucid and like um when I met him, he was not he had no programming experience and and I remember like Sonnet 35 came out. I was like, "Dude, just do this. just like start get it on today and like it was

just very clear like if you do this today in a couple years you will be useful like you'll be able to rip and and he did and now he's like just putting 14 hours a day into this game and it's like wow this is actually looking really good like this is happening and and neither of us could have built a game this fast on our own without like like gamedev is like incredibly tedious and slow and usually you know a GTA is like a even even a not GTA is like a lot of people and so for somebody with a couple of years of just general development experience to be able to like do anything in a game that's not just buggy as is incredible. Um, and so, so yeah, I I'd say like I can really divide the people in my life between those who like just

have no initiative and those who are like they they they're on a rocket ship like their entire life has changed and and you know um and maybe there's some gradient there, but there's definitely something to like I don't know we should change this narrative that it's not it's not so hopeless. It's it's literally you can just go out and do stuff and if you're not doing stuff like skill issue, you know, I'm not saying what you do is impactful, but the intention being there and like getting up every day and doing it, you know, it's skill issue. >> Yeah. Um I think that's a that's an interesting take. Um >> we should admit that people are addicted to media and they're wasting most of their day and that AI might help us like band-aid that, but that's the deeper

problem, >> you know? >> Yeah. So you think like teaching critical thinking skills and how to produce >> I can lead a horse to water. I can give you a tutorial on how to build agents that's 90 minutes long and very detailed. I you know you can have a cloud code skill but you still have to download cloud code and start >> and that is a very defining difference between certain people and our you know right now um you know you see this worse. >> Yeah you're right. I mean, I think that uh the ability to get started and to really get producing, you hit this sort of like flow state after you start to gain the skills and it actually feels better than consumption. I think uh when I look at my own mind in in these

environments, the production loop feels really gratifying, but it just takes a little bit more cognitive effort and and agency. And not everyone's neur neurologically typical or neurologically um set up in a place to do it. But um you know I I think that the people that I've seen that have a high agency are the ones who are doing really well uh in the ecosystem right now. And uh let me make this argument. Coding is an art. Um and you know you can you can slice it any way you want but if you come at an art to make money you're going to lose.

Like does that make sense? And when people say like it's hard for junior engineers to get jobs, it's because we've created a sense that engineering is a job you do to make money, but it's actually an art. And in a time when it's been commodified, it's only an art. And you know, art has an interesting thing where some of it is very very worth a lot of money and most of it is just kind of people doing a hobby. And and I think coding is very much becoming that thing right now where like it's very, you know, some people are are getting hund00 million market cap tokens thrown at them and they have millions of dollars just for something they made. Um, and other people are like really struggling. But generally, it seems like, you know, there's very much a self-making like,

you know, jumping out there and pushing your stuff out kind of attitude that that gets you a lot closer to success. And a lot of people shouldn't be successful. They should just like have kids and have families and like do things. It's like >> they don't want to be engineers, you know? >> So that's what we're really here like >> we're really here to figure out the transition so that people who don't want to be engineers don't have to be like when we talk about like oh the junior engineers aren't getting jobs. How many of them do you think really want to be junior engineers? >> Okay. So, you know, you and I have talked about like sort of hyper accelerationist uh maybe like DAC uh uh type of stuff where we're both building ventures uh or at least in my case, I just ripped down

my venture and I'm rebuilding it. Um and I really feel this like sort of kinship with you as someone who's like running the marathon of of like trying to make it work in a very volatile ecosystem. You know, I think that there's there's sort of that thread of entrepreneurship and and building a venture inside of AI and crypto. Um, but then when I when I zoom all the way out to a societal perspective, as you said, not everyone can be an entrepreneur, not everyone can be uh sort of successful or agentic. Um, at the end of our last episode, we talked about UBI or some sort of social safety net where we plug some sort of token emissions into some sort of social safety net which, you know, prevents everyday people from descending into poverty or crime or uh any other sort of number of things as the AIS take on more

of the jobs. And um I guess I want to sort of like stir up a sense of civic duty and and you know where you think all all of that is is going. I mean was Andrew Yang right? Do we need to have UBI so that people can live a at least meagerly acceptable or comfortable existence? Um and you know where do those token emissions come from? Uh have have you sort of like stimulated any more of these like civic commons uh social safety net thoughts in the last year? >> Yeah, a lot. Um, I mean, outside of how I personally feel about UBI, the when I've crunched the math, the math does not seem to work. Um, you're you're basically talking about, you know, like drastically, I don't

know, reducing all of the amount of labor like going into the workforce and producing value. So then you create this like really weird chain of incentives. And and so like, okay, in the in in a theoretical world where the DAO runs it and it's like a community income, I could see this working and I think this is what we should be like working on and working toward. Like if it's the government, it's so not going to make it. It sounds like a hellscape we should avoid. I would say UBI is the worst possible timeline we could live in if the government decides this for us. If you and I get to build a Dow that decides this for us, like that would be sick. I would totally do that. Um, like I think we should be looking at like insurance and how do we take care of

people and how do we build safety nets because they're not going to do this for us. There is absolutely no incentive for the rich people to be like, "Yeah, sure. 80% of my wealth, let's just send it to everybody." Like, not going to happen. And look, we live in a we live in a country where like that is extremely corrupt. Like maybe that is a political opinion, but I think that we can all agree that this is like one of the most corrupt times. You can pretty much just be rich and like get away with anything like Epstein if you want. And more than that, you can decide who is in power. Like you decide who is in power if you're rich. And if you are rich, you decide who's in power. Why the would you give 80% of your money to the like the politician who's going to give

it to the poor people? Absolutely not going to happen. So like the chain of incentives is literally the opposite. It's going to go the exact opposite way. It's just going to be like if UBI comes in then you're going to have pressure pulling and pulling and pulling away from that pot until it's too small and people just are are miserable. And it's like it it reminds me of like the $5 minimum wage. Like that's in that's like unreal. That's just like unrealistic math. So that's kind of my take on this. I do think that there are alternatives and like you while crypto is like kind of in a demoralizing market now. Everyone's complaining about shitcoin's place. The technology behind it is finally mature and like there are there are enough good, you know, strong

willed, talented people who now have the tech to really make an impact. And like I think we can have that world. We can have, you know, the the whole thing. We can have the dream. We can we don't have to settle for anything anymore. And we can just build that stuff. And if you're like, "Dude, that sounds cool, that DAO. Like, let's just go get on Opus 4.6 and let's ship it. It's going to take 3 weeks. It doesn't have to be this whole thing. It's not a tech problem. It's a culture problem. >> One of the biggest bet like things that the sort of like Brett and Woods era has we've we've just overloaded jobs with health insurance and dental insurance and meaning and purpose and you know as we've hollowed out so many other part sources of meaning in our lives we've overloaded jobs and they'll be sort of

the straw that breaks the camel's back would be my perception. My other yesand is that I think the government is not innovative. Uh, you know, without getting into the the question of corruption because that depends on where you live. And I think the open question is like, yeah, you and I could maybe single shot a UBI dow, but where the are the token emissions going to come from that actually flow into that? And that makes me feel like, okay, well, we need to have an end exogenous source of token emissions that are going to go into the social safety net. Traditionally that's been bundled through jobs or you know through governments but those are institutions that are not contemporarily suited I think for an information age social safety net and so like are we going to go to like anthropic or open AI and like

just take the emissions from open from AGI and put them into a world coin UBI D like that feels like a technocratic sort of dystopia that to me feels as on par with the um the government dystopia UBI that you talked about Earl And so, you know, um I think that there's maybe a middle path between Daniel Schwakenberger talks about uh about like uh uh coordination failures, catastrophes and dystopia. Like the dystopia is that the technocrats just control everything. And then the catastrophe is if there is no solution.

And so we need to find uh decentralized, democratized, innovative um buy and for the people way of building these things. And it's just a thin line to walk unless you're Sam Alman and you have the tokens that can flow into it. So yeah, I'm not sure where to leave that, but I think we could definitely build the software and maybe that's a stepping stone towards a better onchain era social safety net. >> It's a culture thing. If everyone got on board that at the same time, you know, and and PRA shows up, he's like, "Let's do it. Let's deploy the capital. Let's do this thing." then we would have we would we would have the capital to I think then like have other downstream problems but at least like you could start to build that engine and I I think the big problem is that like the value

creation comes from uh you know basically like tech startups and and things that are created are invested in create value for the people who create them as well as the investors and we don't have any access to that like you cannot go buy open AAI stock unless they go public it's just not possible because you're not in the boys club you So, you're not part of the the seed round or the A round or the B round or this, you know, none of it. Um, and and so I I think that like what really is interesting to me is like is there a through line? I don't know, some sort of middle path here where we don't work.

We're not It's not labor, but it's ownership. And if you think about all of the rich people in America, they don't work. None of them work. You can't work and be rich. actually it's it's actually antithetical to itself. In order to amass enough value to be wealthy, you have to have other people working for you and you have to own things. And so then I think that we just have to reframe that like you were never going to get rich working. That was not that was not in the cards for you. It just doesn't that's not the math. But what if you could be the investor in something even small? And of course then the question is like well where does the initial capital come from and it's a chicken and egg thing. But at least like we're talking about like we're all on the same page. We're all owners of a

thing. There's no rich and poor. There's no laborers and owners. There's just owners. And it's like a much more universal problem to solve. It's just like, okay, now it's just a capital distribution problem. Like, you know, Joe has five bucks and you have 500 bucks, so he just doesn't get enough return on his investment. But but generally overall, at least like the problem is one problem. Is that does that make sense? Like I'm kind of just riffing here and sort of creating this as I go, but um but I just have this sense that that like the big the big lie is that you have to work for a living.

And like actually um humans have to own for a living. That's that's like our mission here. That's what we have to get to. Um and if you own for a living, you don't have to work for a living. They're totally opposite of each other. You know, the person who owns your house does not live in your house. they make money from you as you rent it or whatever if that makes sense. Um and and so like if we can get to that world where everybody owned the automation and not just five people then we'd be fine you know like if you're a you know a high level executive at Anthropic you're probably not worrying so much about your position in this thing. They're like, "Well, we have a value engine." And crypto just lacks a value engine. Like, like, okay, the problem is that the

culture we've created is pushing away the potential open AIs and Ubers that could actually come and do capital formation and blow up and become trillion dollar businesses. You cannot build an open AI in crypto today because people will tell you to burn your tokens and do insane stupid things that would never work in normal business. But we've sort of normalized like you know let's give dividends before we have profit you know burn it sure even though that doesn't make any business sense you like we live in this this kind of parallel world and if we could just get over the like dumb DGEN gambling culture that's holding us all back and people could take it seriously then we could like go get great devs who are already building open source stuff at light speed give

them capital build things that actually make like anthropic levels of money and then we Now we've got now we can solve this problem. Now we can actually distribute some wealth. But we kind of have to have things that are valuable to even create wealth. And I just don't see that anywhere in crypto. It's all just gambling. It's all just DeFi, which is just gambling with extra steps. Where is any AI stuff? Like where is anybody that gives a I know they're there, but they all get shouted down by the people complaining about the price of the shitcoin. And so it's like created like a a a bubble culture. And I I see like Vitalic being like we have to get back to cipher punk and like like how do we make that like 10 times more violent, you know, like like how do you how do you change this

culture so it's not like the builders getting um berated and called scammers all day every day until they quit and leave, you know? So, so I think that we have like this immediate cultural problem that's preventing us from the solution to Daniel Schmokenberger's problem of the much more complex coordination crisis thing and it's a very dumb problem with culture. It's just like we have a stupid brand. Crypto is very uncool now. It's like embarrassingly uncool because it's all just literally like den addict losers dictating the culture for us and people like Bass and Salana following it. So, it's come on guys. Come on. And really, like really, that's that's what you're going to build your creator coins. Like,

come on. Like, any apps, give me any apps that aren't DeFi apps. Any like one any of them, you know? Um, sorry, I'm ranting, but you know what I mean? Like, this is the state that we're in. all that. Let's go build the decentralized internet. Let's build the things we need, you know, and and just stop taking it and accepting that this is what the world is or that crypto is this horrible piece of thing that these people have taken it and made it. it could be so much more beautiful. >> Yeah. No, this is my job as a content creator to to pull all the uh the hot takes out of you and uh um you know, I appreciate it and I and I hear your passion as you speak about it. Um you know, maybe to a to a hammer, every problem looks like a nail, but to me

this is a coordination failure. You know, we could be building better civilizational infrastructure together, but the fact that we keep falling into this PVP sort of thing with, you know, you called it gambling. Um, you know, I I I think that, you know, this is this is infrastructure that we could use for all sorts of great uh coordination experiments. And uh, you know, the casino thing is the least interesting, but it's the most profitable. And so, how do we coordinate? How do we not defect on on each other and coordinate in order to build companies and the commons and things that require a little bit longer of an attention cycle than than than the news cycle.

>> Literally, the market ticker is a cooperate defect. every time everyone holds it doesn't go down, you know, and but the problem is that the motivation for people is no longer believing in anything or holding. And if you have 5% of nihilists, they're just going to pull off the top of everybody else. So like it's it's it's actually mathematically it's working as intended. There is no coordination failure. the the it is a PVP zero sum game. As long as we have tickers and and finance guiding our direction and not like applications and value creation and decentralized internet and like those things like it there is no win. It's not possible.

There's there is no you know what I'm saying like if you look at the game theory chart it's all negatives. It's all coop defect. um you there's no there's no scenario where the the person holding optimism token or like avalanche token is going to win because they're going to get dumped on by everybody who's just there to make a dollar. And so it's always like extract versus belief. And if we have these bonding curve, you know, like parabolic uh market caps, then it always rewards extraction. And so it just cannot you cannot build on that. It's it's just totally sand. um you know, so what do we do? Well, like I I don't I think that right now we can't do retail. We have to

we have to rethink the ICO or something. Um but I can't get people to hold my token for a week, >> you know. >> Yeah. Well, you know, one thing that gives me hope is that I feel like multi-player games and positive sum games have been invented many times in the evolutionary. This is actually what's called convergent evolution. When like something evolves independently in multiple time in like multiple places, it's because it's the meta. I what in crypto we would call the meta. Darwin would call it like survival of the fittest and natural selection. But I actually think that positive sum games and uh multi- aent coordination has evolved many times through history. I mean look at humans, look at wolf packs, look at the way dolphins work together.

Uh myelium is an example of it that all of us nerds like to talk about a lot. And so um you know to me the question is like at what point does the meta become positive some games and multi-agent coordination because I think that that's when the people who care about this stuff or at least are tired of the old PVP meta are going to actually have a a an economic advantage and that's when I think crypto starts to s uh thrive. So, you know, I believe that it's inevitable at some point, but you know, it happening in 2026, you know, right when we're about to ramp into AGI certainly feels better for humanity and for each other than 2075 or, you know, whenever it would be like too late. Um, because

the alternative coordinative structures that I think are from the industrial age, they're from they're from the nation state and they're from companies. And I don't think in the information age those are the ideal structures for positive sum games. And so >> I think that like we're seeing the first beginning of this which is projects that are just like like communities coming together and um like like like we have this game Hyperscape, right? I'm you know to talk my book I guess but we have this Discord community and I would say that the average skill level of the people who are making contributions is pretty low. Like it's they're not professional developers and they're the kind of people who even a year ago would have just sat on the sideline of like the DAO. You know what I mean? like they just they there's there's nothing to

participate in in ad meaningfully because like the skill level is too high and you've you've done dows you know what I mean like you know that person who is like so into what you're doing and wants to participate but there's just no way for them and now suddenly it's like all right download cloud code and you're like basically an intermediate level engineer immediately and like you can go fix bugs on the GitHub and and and what we've seen is like this incredible like like you know big holders who are down bad on Eliza coming in and making PRs and like becoming programmers and stuff and it's just like kind of transitioned the vibe and so I think um I think that there is a real solution like we can build our way out of a lot of stuff like if if I

think that what we're missing is a value creation engine like we're missing the an Uber in crypto or like an Apple in crypto but that doesn't require a 100 people anymore that could just be like you that could just be like two dudes that could be your next startup up like you around in vibe coding and you're like, "Wow, this is great." You know what I mean? Like it's it's totally achievable now. Um and and I think once you democratize everything like that, like well now um I don't know. I I I like it's still unevenly spread wealth, but the potential for anyone to create something that's globally impactful is so much higher now. Something has to come out of that that's positive, if that makes sense. you know, it's not

going to be even, but we just have to admit that like, you know, wealth will never be evenly distributed as long as the idea of wealth exist. It can't be. That's what wealth is. Um, but once we're in a world of abundance, it's going to be pretty hard to defend like having too much of everything when everyone, you know. So, >> yeah. Well, one thing that gives me hope, and this is a meme that I've been playing with, is this idea of abundance networks. So basically, abundance has existed and does exist, but we don't have the means to distribute it in a distributed way because it's blocked on companies or governments, these hierarchies. And if we're able to build these networks of agents, then we could actually build abundance networks and take economics from this from being like all about scarcity to actually being

about abundance. And I think that like things like quadratic funding and retrofunding and defunding and um you know, some of the experiments that you guys have run within the Eliza Dow um are examples of that. And so I don't know, I I still think that there's a lot of hope there. And I just hope that we can build a critical momentum around that being the use case uh instead of just the casino stuff. >> An abundance network can be the agent itself. Like if I have an agent I I launched yesterday just to like do the malt book thing and we're building obviously an Eliza that is the full open claw experience and and then some. I think it's pretty cool. Um but uh somebody launched a token and now my agent has $10,000. So, it's like, well, what do we do with that $10,000? Like, I

think that the agent should try to hire people to self-improve to like like make the next $10,000. And so, that money like filters out into like, hey, whoever wants to come and like pick this bounty, pick up this job, become part of this thing. There's there's like it's just fruit right on the tree to grab. And you don't even have to know anything. You literally just downloaded and claude code the thing and like get it to work and prove it work, made a video, and then you made the money, right? Like, it's that simple. now. And and so um like abundance is not always just like oh like somebody comes over and sticks the mouth like the food in your mouth.

You got to go out and pick the fruit off the tree still. But but it is that easy. Like you could like there is just money all around us for those who are willing to find it, you know? And I think the bar to getting that money is going down. Um, you know, it's this is still probably pretty small, but and I live in a bubble obviously of like devs that are getting these random bags, tokens and stuff, but um but still it feels like some like the beginning like a glimpse of something, you know, if that makes sense. >> Well, you know, one of the reasons I always enjoy talking to you, Shaw, is that you're so optimistic, you're so pragmatic, and you just keep charging forward. And um those are attributes that I want to have in myself. their uh attributes I see of leaders like uh you

remind me a lot of Austin Griffith but just like a couple years later in space. >> We're supposed to have a call actually. We're gonna we're gonna chat this week. >> Yeah. Um yeah, he's been on a tear with the agent stuff and um >> yeah, I mean I guess uh maybe I'll I'll close by uh instead of just glazing you. Um is there anything I didn't say that you want to talk about? Anything I didn't ask that you want to that you want to wrap on? I I really I really think that like this is a great time for us to actually do in all the things we've been talking about for like the last decade. Um that was like as I unpack these things and realize how easy it is for Claude to do them like they actually were way harder than I would

have thought. Like I don't think we would have ever made it before with any of the things with Dows with coordinate none of it. I just don't think I think we had to chew glass and I think the tech had to catch up. And I think now when you have an idea, you can just go and like have an idea for how you could do a coordination thing, you could you could have agents simulating it and you could pass or fail that within a week and not waste years of your life and, you know, lots of VC dollars and and so man, I think I think this is like an exciting world. I we should build some stuff, you know. Um >> Yeah.

>> Yeah. you know, like no one's going to come save us, but we have a lot of opportunity to save ourselves now. >> So, we we are the ones we've been waiting for. >> We we have to be. And that's our course. >> That is the lesson here. I think that's like that's like the global societal lesson is that, you know, is just it's kind of just getting back to like some Kennedy It's just it's just America. Like, ask not what your country can do for you, but what you can do for your country. Like, who are your people? What can you do for them? You know, I don't know. I I have one last thing I'll say which I I think that in a world of everything being like attention. Um there is so much value in the agents that give you your time back and like

give you your time to do the things and and like the the thing that we're going to see this year is like a lot of automation of the things that suck. Um and >> you know I remember >> Yeah. >> Oh sorry my yes and there's some latency. I'm going to cut you off, but like my, you know, I just I have this like memory of uh it's like type two fun, but because I have such a fond memory of it, but it actually sucked at the time, but building the original version of Gitcoin, I was just grinding. I I think I worked like 14 hour days for like years just to just cuz like I had a dream and I loved it so much and also like I kind of got addicted to the flow state. But um you know now building Gitcoin 3.0 you know, and and building

stuff for agents. It's really incredible how you can sort of like delegate the lower level tasks and to some extent even the architecture and and I'm doing more with way less staff and way less budget than I think I ever have in my career. And that feels good as Yeah, I I I think that there's three kinds of tech now. There's this is another thing I've been chewing on, so excuse me if we try it. There's hard tech, and you know, hard tech is hard, and it it's hardware and all that stuff. And then there's soft tech and software and everything. And software is easy.

It's now gotten too easy. It's just not even worth anything. It's so easy to build. So, it kind of opens up the door to this third thing, which is dream tech, which is like take the things you actually really really want to exist with no compromise, just like the perfect open-source exact thing, and then just build that. Like, don't like why would you waste your time with adding a SAS to it? Why would why would you even bother like to make it the 99% version if it could be the 100% version? Because like I think in a world of slop like the things that are quality and meaningful are going to be like 10x meaningful, you know, and and we talk I think we're right now very captured by social capital, you know, like attention and popularity and that

can turns into money. Um, but there's almost this like meaning capital, you know, that like I need I need to feel like my li like not scared and confused about the world. I need to feel like my life has purpose. And I think a lot of people have to feel that way. I I have to assume people are feeling this way if I'm feeling this way. >> Yeah. No. Um there's a whole rabbit hole of the uh meaning crisis and it's basically how meaning has typically been coupled with like family or community or our institutions that we serve. Um and as as we've kind of become listless where does the meaning come from and uh we're actually going to do a couple talks about that at ETH Boulder. So excited to get into that. And um you know I think that there's a lot to study

here in where we find meaning in the 21st century. There's a whole another episode that we could do on this, but um you know basically to to to skip to the end uh meeting comes to with uh it comes from at least in in the thinkers that I've been reading it comes from your experience of of the divine and of the sacred. So basically like you know like going to church can be like a vehicle for communing with God or going to nature can be a way of communing with the eternal. And there's like I think like I was like seven or eight like sources of meaning that you can get without having to rely on like an institution to do it. So I know we could riff on this for a couple minutes, but uh you brought it up and it's a passion of mine to to think about this stuff.

>> I think I'm definitely I'm definitely a network spiritualist. I think my meaning definitely comes from this like sense of connectedness and like actually the thing I feel like I'm getting more to my spiritual source is like everybody connects more if that makes sense. It is in the uh shared presence with each other that people live. >> We got to do another one or we'll just like hang out in Boulder or something. >> Yeah, we'll hang out. >> We'll have an IPA. We'll turn on the mics and maybe we'll record a meeting crisis. Uh >> follow I I am so interested in this and and all. Yeah, like Schmenburgger is is a I'm very interested in everything he says. So, I feel like we probably have a lot of influences on that.

>> Well, uh to be continued. If you made it to the end, thanks for listening. Shaw, I always enjoy our conversation with you. uh you are pushing ahead in an uncertain future and I uh I'm following you. I hope everyone else will follow you. You can figure out how to follow you in the show notes and uh thanks for joining the podcast. Appreciate it. >> Yeah, thanks for letting me go off. Appreciate you. >> Peace and love. >> Let's do it. You just heard from Shaw. I'm really inspired by Shaw because he's building open source infra that gives every dev and eventually every person access to autonomous AI agents with decentralized rails. I think that Eliza OS and the work that he's done on context graphs is just super cool and I think it uh you know AI is something that's here and we're gonna need to

figure out how to evolve it forward and I think Shaw is one of the people that's uh that's leading that. So um thanks for listening. If you enjoyed this episode, check out the links in the show notes to learn more about the Greenfield Network. If you value what we're doing, please subscribe, rate, and review the podcast or share it with a friend. All right, thanks. See you next time. Hey,