Dr. Ben Goertzel on why artificial general intelligence should be decentralized

0:00I'm Ashton Addison from the Crypto Coin Show and today on blockchain interviews with we have with us Dr. Ben Gowartsel, CEO of Singularityet and the ASI Alliance here to talk about artificial intelligence, decentralized networks, decentralized AI, AGI, and so much more. Ben, welcome to the show and thanks for taking the time.

Yeah, thanks for having me. Yeah,

0:22excited to dive into your work on AI which is decades long now and now the world is catching up and blockchain is also beginning to merge more with AI and the principles of blockchain of decentralized collaborative in AI. There's a bit of a battle going on. I'd love to start out on a high level of sort of your thoughts on the vision of AI being built around collaboration,

0:48safety, and security rather than centralized control and sort of what we're seeing in the world with how AI is being built and being released to the mainstream right now. Yeah, I mean it's an amazing time for AI and for those of us who have been working in the AI field for decades, it's it's almost a bit disorienting that suddenly everyone loves AI, right?

1:10Whereas for most of my most of my career, AI was a sort of little corner of the research world and an even smaller corner of the business world. That said, the kinds of AI that are capturing all the resources today is a very small subset of what AI can do. And the way the way the business

1:39world tends to work is, you know, once one thing works and someone has a big success, everyone copies that exact same thing because they view that as the least risky thing to do. and then

the research world follows along because researchers want to do what will get them the highest paying jobs or the easiest venture money or something. So you have an interesting situation now

2:03where there's so much resource and so much enthusiasm going into AI and it is delivering real value and it is really interesting. On the other hand, it's a it's very limited and I think in the next few years, we're going to see a great flowering and expansion of AI, both in terms of the kinds of AI algorithms and approaches that are out there and in terms of how AI is

2:33deployed at the hardware and computer network level. Right? And I'm I'm working with my Singularet and ASI alliance team. I mean I'm working on both aspects like how does the actual machine thinking work and then how do you deploy that across machines and networks which is where the blockchain aspect comes in right and so we can we can talk about both of those

3:01and how how they connect together. Definitely, you know, and I've I've been a blockchain enthusiast since the early 2010s and now seeing AI come in so rapidly, I feel like there is a great synergy for the decentralized nature of it and you know there's already been some discussion around the AI that you know the everyday people are using which is right now just sort of

3:28glorified chat bots which as you alluded to what they are capable of doing in probably the near future is going to be astronomically more than what they're capable of right now. But even right now for the everyday person, it seems like a 10x in productivity. I can't imagine over, you know, the next few decades it'll be like 10,000x.

I think I think

3:51we're going to get to a human level thinking machine within five years. It could even be two or three years. And once you get there, then accelerating or magnifying human pre productivity is not even the right way to look at it, right? Because I mean, once you have machines that are so much smarter than than than people, then the machines are able to do the

4:20production and direct the production. And it's there's other big questions like are these our friends and are they enhancing and helping with human life or doing something else and are we even like choosing to

choosing to plug them into our heads and like fuse with the machine right but

but I think we're in a very interesting transition period now where AI is

4:45enhancing magnifying accelerating what we can do and in the transition period between here and full-on human level thinking machines. I mean, there's technologies to be built, there's businesses to be built, there's there's lives to live, right? And this will be a unique period in history when we're still in some ways smarter than the machines. But the ways to gain advantage

5:15and deliver good in the human economy mostly involve synergizing with these rapidly evolving sort of smart machines, right? And this is a it's a unique [snorts] period in history, right? which is quite fascinating to be to be living through it. In terms of blockchain, you know, in 2001 and 2000 on the Java 1.1 language, I was screwing

5:46around with like how do you bring distributed processing and strong encryption together,

right? And it was what you would now call early stage AI blockchain network. It was just very very slow. And it just seemed like we needed software to speed up before you could make that viable. It's funny. I even tossed around with my friends the idea

6:15of making a decentralized money, but we were just like, well, it's going to be really slow to clear any transaction, so who's going to want to use it? If I'd thought of money laundering and tax avoidance as use cases, right? Yeah. Then I'd be Satoshi, right? and we'd probably have AGI already because I would have had trillions of dollars to build it, right? But

6:43

the reason I mentioned that though is it's sort of a historical accident that blockchain began for decentralized money. I mean, decentralized money is becoming a good use case now that things get faster and faster on the blockchain side, but I mean, you could just as well have had blockchain emerge for decentralized AI or for decentralized medical records with data

7:08sovereignty or something, right? So, I mean, it's a having started with decentralized money is a plus and a minus, right? It's a plus because I mean people made money and it also gives a way for blockchain to be used to fund other sorts of projects, right? ICOs and other more sophisticated mechanisms. On the other hand, it's also caused a sort of

7:33narrowness of thinking where the world just thinks about blockchain as a mechanism for digital money,

but it's not just that,

right? And this I think we're going to break through that next year with our launch of the ASI chain which is a new layer one chain and

this will be the first layer one chain. It lets you do largecale advanced AI processes on chain, right? Because

8:01

so I launched in 2017 Singularity, which is really the first serious blockchain infrastructure for AI and fetch, which we later merge with, launched just a little bit after that, right?

But what we do there, I mean, we have AI processes running in Docker containers on different machines. Then we have a sort of proxy by the side that has smart contracts. So you have like AI processes

8:27running in containers on different machines and the blockchain lets them communicate and coordinate together. But the AI itself is not in a micro level running on the blockchain. The blockchain is used to coordinate between a decentralized network of AI processes, right? And

that's what everyone is doing. It's what you can do.

With ASI chain, we can do things

8:49differently. we can very flexibly run AI on chain and the way this is enabled is as part of an open source AGI project that I've been leading for years which is called hyper which is a new version of the open cog open source AGI project in building hyperon which is not about blockchain it's about how do you make AGI by using not just neural nets but use neural nets use

9:16logic engine use evolutionary learning usear variety of AI paradigms, implement them together in this big shared enramm knowledge graph to make a system that can learn from experience and then learn from studying it its own self and its environment. But in building this hyperon agi project, we developed a programming language called meta meta. And two months after we launched it and

9:40named it that, Facebook changed its name to meta. But we not we have another t. We're not changing the name. And then meta with two T's also means loving kindness in Sanskrit. It's a kind of meditation, right? But so what we found is this AGI programming language we called meta that we developed for efficient large scale deployment of a variety of AI methods

10:08can also be used as a smart contract language. So we've repurposed our AGI programming language as a smart contract language. And what that means is we can take quite sophisticated scalable AI methods and we can run them as smart contracts on chain and we can sort of modulate at a micro level which bits of the AI run on chain or not. Like for programmers in the audience like each

10:33each function block in code you can encrypt with a num with a number of private keys, right? So you can decide like I'm encrypting this bit. I'm not bothering to encrypt encrypt this bit. And you can you can very flexibly put things onchain or offchain. So that's it's going to be quite interesting. And one thing that gets you is like right now all the big AI models are run

11:01on monolithic server farms like a bunch. It's distributed processing, but it's a bunch of machines like in one facility owned by one guy and there's a certain efficiency there, right? But with ASI chain, you can take advanced AI algorithms and run them on a heterogeneous network of machines sitting in all different countries and all different places, right? And of

11:25course,

there's certain efficiency that you get from having all your machines in one building,

but there's also an efficiency that you get by being able to run part of your server network like right next to a power facility in a place where it's cheaper to run a server farm or even run part of it on your phone or something, right? So I mean I think

11:50we're at a very interesting point where so we have now we've launched the DevNet version of ASI chain we'll go into test net early next year and then mainet after testing has been done with Hyperon the AGI platform like in the last couple months we finally got a really fast compiler for our AGI programming language so we're now just working on scaling up all of our research from the

12:17previous years. So, we're sort of at a transition point of moving from research into reality with Hyperon and with ASI chain. Now, big tech isn't slowing down either, but again, they're doing a certain thing. They're doing centralized server farms

and they're doing big deep neural nets trained using back propagation algorithm. So if that's the golden

12:43path to AGI, like big tech wins and whether it's Microsoft or Google or Tencent, we'll see. If

if using a greater variety of AI algorithms deployed on a decentralized network is a better path to AGI,

then then humanity wins instead, right? And then the decentralized ecosystem gets to play play a ma major major role, right? And so we're we're

13:10now where now when push comes to shove like when the rubber hits the road right it's a which is quite quite fun albeit a bit exhausting if you're right in the middle of it.

Mhm. Yeah. No, it's a it's it's crazy times right now and thanks for explaining the architecture behind the ASI chain and Hyperon as well. You know, in the last month we've seen AWS go

13:35down. It was actually affecting millions of businesses around the world. And then also this week, Cloudflare went down. X wasn't working. Meta wasn't working. At least Facebook and even OpenAI and the LLMs weren't working because of one server of one small issue. I feel like that is the thing that's waking people up to this decentralized AI that shouldn't run into

14:00any of those issues.

Yeah. I mean it's not a panacea right but I mean it is it is more robust right I mean no no no no no no question I mean that's that's the nature of it the challenge

has been to get that robustness without an egregious

loss of efficiency

and I think that's what the ASI chain gives you and it's been subtle

14:30like we've been working closely with Greg Meredith who led the Rchain blockchain team

and

I mean that kind of crashed and burned for tokconomic reasons but we took the open-source code from our chain integrate it with our meta language and some of some of our other code and we've now come up with something that I think it basically solves the blockchain

14:58trillemma but in a sort of artfully dodging way like the blockchain trilmma is can you know can you get something that's you know scalable and secure and decentralized alto together right and the way we do that with ASI chain is partly that different shards of the chain can have different consensus mechanisms so the whole thing is secure

15:29and the whole thing is decentralized but to get scale There's not one consensus algorithm that will give scale for every application that you want. So what you need is a system that lets you put different consensus algorithms on different shards and then lets them all communicate together for consensus on the overall network. So for example, if you want to

15:50run a DEX that does high frequency trading on crypto tokens, then you need a consensus mechanism that's really really fast in certain senses. Then you may require the nodes in that network have a lot of multiGPU servers, right? And there's a consensus mechanism called Casanova we can we can use there that was developed by one of our lead developers at his previous company Pyro

16:16effects. Right now, if you want say something can run on a mesh network so that everyone's phone can run some of the AI and you have like a globally distributed AI commons where anyone can contribute some processing power then they can democratically vote and what that commons does, right? And the validators may pop on and offline randomly because they're on people's

16:39phones.

For that, we can run a consensus algorithm called Cordial Miners. It's not as fast as you would want for high frequency trading, but it's very robust. The validators popping on and offline all the time. So you can have one shard doing the mesh network thing, one shard doing the high frequency trading thing. They have different consensus mechanisms. They're all running on

17:00they're all running on the same secure decentralized infrastructure. So you I think to solve the blockchain trillemma you have to cheat a little bit like that because there's not there's not one consensus to rule them all

but you don't you don't need there to be right and so that then the other interesting thing is once you've decided to do multiple shards that may or may

17:23not have different consensus mechanisms you can do tokconomics that way right so you can you can say okay take your ASI token which we're going to rename the fetch token from from our network is ASI token with the launch of ASI chain. You can take your ASI token, you can stake them on a shard and then as reward for that staking, you can get a liquid

17:44staking token associate associated with that shard. And then

then we have an ecosystem a little bit like dynamic dynamic tao from from bit tensor

but we're working out the tokconomics differently to so that the emissions are tied to reputation and revenue rather than to market gaps. We really want to we want to disincentivize pump and dump and memecoin and incentivize

18:11projects that are actually using AI to deliver deliver value on the chain, right? So I think if this all comes out right, I mean then we're building the next generation AI economy, right? like what we're and inside each node we have the instrumentation to run scalable neural nets and also the hyperon system for neural symbolic evolutionary AI right so what we want to

18:38do we want to seed the next phase of AI development toward AGI going beyond deep neural nets to neural symbolic and integrating with different AI methods but we want to seed this next generation of AI on the next generation decentralized econ economy. So like when you have you have a couple kids out of Bangladesh or wherever who have a new cool idea for an AI company instead of

19:04struggling to raise VC money which is going to be very difficult in their country, right? then they can launch a deepend thing on the shard of ASI chain and if they have a product that actually delivers value to people that will get them a lot of LST emissions they can then trade that LST on the DEX shard of the ASI chain and you have a whole decentralized

19:29AI AI economy there right and we're we're in the middle of building this now like the devet is launched developers who want to help chip in or test it you can you can you can find it online. And this is going to be quite epic during 2026 I believe. And so you interviewed Hamayun from our partner project Fetch a couple months ago. So

19:53

we've had a bunch of conversations about the Fetch agent verse

and how do you run parts of that on the Shard and ASI chain, right? Because that's now run on Cosmos blockchain.

Mh. which is great for some things, but the fetch agents can be crosschain. You have part of it on Cosmos. You have part of it in a shard of ASI chain and they can all communicate and

20:16synergize together and in that way you sort of jumpstarting what we can do on ASI chain with all the things Fetch averse can do already.

Yeah, it's very cool and great to hear Greg's name. Actually, we were large proponents of our chain back in the in the 2017 days. Greg lives like three miles away from me here in the Seattle area. So yeah, and we're both musicians,

20:39so I've gotten to know him super well, actually.

That's great. Yeah, tell him I said hi. We've had him on the show half a dozen times when our chain was building. We'd love to hear from him again. Well, so that yeah, it's interesting when when you run open cog hyperon when you spin up the inramm knowledge graph, what you do is you run what until last month was called

21:03an R node and it's it's an R node from our chain running the knowledge graph. It's it's now been renamed Fire Node because his new company is Firefly, right? But for a while I think his company was R Unchained which was was pretty cool. But so yeah Greg Greg has been a great collaborator and we've I think we've taken all the computer science and data structure

21:30stuff for the A&R chain and we've worked together to take it to the to the next level on the back end of ASI chain.

Yeah.

Yeah. It's very cool. And when we had Humayan on with Fetch, you know, we were talking about AI agents and how we can get them into the hands of more people. And you know, Open AI has made it really easy for people that, you

21:52know, grandparents that don't really know how to use tech that well to query answers and do basic chat. Now the functionality is coming in where the AI agent can start doing things on the website for them or book the hotel, etc. And when I was looking into Fetch's AI agent verse, you know, it was still quite a bit of technical expertise for even myself to run a AI agent.

22:20

It's not at that level yet in terms of tooling and that

that is an advantage that big tech has, right? like they have

we've been very backend oriented and fetch even has been backend oriented not quite as much as SNET but

I mean if you're open AI you know they've got 4500 employees right so I mean I mean they

and they're much smaller than Google

22:46they're much smaller than Microsoft or Tencent but I mean

so they can put a lot of work on easy to use tooling I mean that said I think we're using their vibe coding tools for interface stuff along with along with everybody else. It doesn't build the AGI but it can it can build the it can build the interfaces right. So I think I think by the time we launch

23:14ASI chain mainet which I'm hoping is sometime middle of middle of next year though we're not giving a hard date yet right

I think we'll have some pretty nice tooling there also to make it easy for people to do to do things and I mean we you know there's some things that are an uphill battle because I mean OpenAI has been operating at a

23:38loss right so I they will give you access to all their servers and they're spending money for it and they're getting huge amounts of money from V VCs for it and we I mean we don't have that much money as they do to operate at a loss but we see that as a challenge to make more and more efficient AI AI algorithms right like we may not be able to operate at a loss but we can let you

24:03run part of your AI on your on your own phone and on your own laptop which they're not able to do with it with their with their centralized infrastructure. Right? So as we move toward lot toward launch of ASI chain I think we will have easier to use tooling than you've seen on singularity net or fetch with the ASI main chain launch and we hope to overcome the inability to

24:29operate at a huge loss like openai and anthropic by the lower cost base that comes from a decentralized hardware hardware infrastructure right and the these are challenges, but I think, you know, we've got an amazing tech team and I think we can we can overcome these these challenges. And open source is a huge help, too, right? Like we've got a

24:55growing open source development community helping us build helping us build all this stuff. And I mean, big tech is embracing open source, but only in a very limited way. like the core of what they're doing is closed and they'll they'll open some models but not the code used to train the models and so forth, right? So I mean we're Hyperon is all fully open source. ASI chain is all

25:22fully open source. So I mean as these things roll out and show more and more capabilities, we can get a positive feedback loop by pulling in more and more developers around the world to help make it better and better.

Definitely. It's it's a good model to have to have a base foundation on and I was reading as a part of the ASI chain, you know, part of the goal is to

25:44have like billions of autonomous agents and have everybody hopefully easily be able to build something for their specific use case. Do you

I mean the way you get billions of agents is agents that build agents, right? which I mean we've been we've been talking about since the launch of Singularity Net, but we now have the infrastructure that can actually do it.

26:05Yeah.

Mhm. Yeah. I wanted to touch on the humanoid robots. You mentioned it briefly near the beginning and I know that, you know, you were building the Sophia robot much before, you know, Tesla and these other humanoid robots sort of came into the picture into the mainstream.

Where do you see that at right now? It seems like there's a inflection point

26:27where you know at least as a part of Elon Musk's new proposal that it's part of the terms is to make millions of bots.

Robotics robotics is major growth, right? And I'm I'm taking a walk to visit a humanoid robot right here. I don't think I'm I'm gonna turn her on, but you can see this is a

this is Dez Deona, who is Sophia's Sophia's little sister.

26:54

Wow. She sings in our rock band Desdon's Dream where I play the keyboards and occasionally sing along. She [laughter] she composes music and whatnot. And we've yeah, we've got

we've got under repair Grace who's a who's an elder care humanoid robot here. Right. So I mean this is a is a major growth area for sure. I

27:24mean I think that robotics is now it's reduced to a software problem

plus a lowcost scalable manufacturer problem, right? Like we have the robot hardware that we need to carry out pretty much every practical task that we care about. And the reason we don't have robots doing our plumbing and electrical work and picking up our kids' toys is mostly the AI software hasn't

27:53been smart enough,

right? And

once AI software is smart enough,

then the work will be done to tool factories to make the robots cheaper and cheaper and cheaper. And I think over the next few years,

the AI will become smart enough, right? Which is which is

is awesome. Now, I sort of think most of the robots out there are not really going to be humanoid.

28:21

I mean, for most tasks, you don't need a humanoid body. Even if you think about like working in my kitchen at home,

I mean, there's low down cabinets where you'd like to have an arm on your feet. There's high up cabinets where you want to like send a tentacle up there or something, right? for so I mean I think a humanoid form factor will be a minority of the robots that

28:47are actually out there in use. Now the work I've been doing in robotics with Hansen Robotics who made the Sophia

robot the first robot citizen which I led the development team for and then Dezdona and Grace Sophia's little sisters then also we have a company called Mind Children which is making three and a half foot tall robots which are aimed at being teachers assistants

29:15initially to be rolled out in Korea. So, we're focused more on the social and emotional side of robotics. So, I mean, of course, they need to move around and they need to be able to pick up things and use use their hands, but the re the real thing you need from an education robot is just the ability to connect with kids and understand what they're

29:39doing and help them. And with elder care, again, if you need a robot to pick someone out of bed, you don't really need a humanoid farm. You just need like a an automated bed and wheelchair that connect to each other. What you need a humanoid form for is cognitive and emotional interaction be sort of nurses assistant assistant with the nurses and nursing assistants

30:03are busy. So I think that's also an interesting problem. It's interesting to me that Tesla and Figure Unitary, these big guys in the robotic space, they're not focusing on that. Like these are kind of faceless scary looking things, whereas Hansen Robotics and us with mind children are focusing more on the kind of soft soft

30:32side of it. But I mean, the beauty of being a robot is you can mix and match body parts, right? So like you could take a handsome robotics head, put it on a unitry body or take a take a mind children teacher head and add on arms that someone else made that let it help kids with their with their fine hand work better. Right? So, I mean, I think it's a

30:57

it's going to be huge and

it highlights the need for data sovereignty, data protection, all the stuff that the blockchain gives you because once you have robots in every classroom, and every house, and like

robots in your bedroom, and your kids playroom, and so on. I mean,

where's all that data going, right? M

I mean on the one hand yes all that data

31:21should be used to help train and make the robots smarter. On the other hand none of us want big tech and big government to be mining everything we say and everything we do. So the capability to store all that data in a encrypted and secure way and let different views into that data to be used to train the AI models without leaking any of our private info. I

31:48mean this is a this is something that blockchain delivers. It's something that hyper running on the ASI chain delivers. It's something that big tech tech stacks do not deliver

because they don't care about that. And

at what point people start to care about that? Will be interesting to see like

everyone claims to care about data sovereignty, but they still give all

32:12their day to Google and Facebook, but and maybe that robots cruising around in their bedroom is where they draw the line and actually do start to care about about data sovereignty, right? In which case, blockchain based infrastructure has has a very large advantage.

Yeah, we'll we'll see. I feel like right now there's definitely more awareness

32:33going on with privacy at least we've seen you know Zcash which may or may not be fully private but that sort of bringing to people's minds about privacy and then Vitalic's latest talk at the DevCon talking about how Ethereum chain is more moving into the privacy era. So hopefully that translates to sort of overall awareness maybe outside of the blockchain

32:57bubble that you know people once they see a robot staring at them while they're sleeping or in their private moments.

It's it's interesting how little people care in practice about this sort of thing, right? And it's a I mean outside of the sort of blockchain neo cipher punk bubble it's it's interesting how little care people actually have

33:26about that. But no in Europe people do care right a lot a lot more. In China people care a lot less.

In America people pretend they care but they but they don't act like it. But it's a interesting to see how it evolves. But I mean you may be just one intrusive step by the government away from people carrying a lot, right? Like if Trump's IC starts bumping

33:54illegal immigrants based on their social media posts and then suddenly may people may care a huge amount, right? So it'll be very interesting to see how it evolves.

Yeah. I feel like some world events or some specific triggers are definitely going to wake up and hopefully that will help bring more awareness to the decentralized.

We can we can bring the we can build the

34:18technology.

Yeah.

So that it's it's it's there when people realize they need it. Yeah,

definitely. Well, I'm looking forward to seeing how the growth of ASI chain and decentralized AI is now, you know, at least that the greater blockchain community and web 3 developers are realizing that this is important to be working on.

Yeah. And I think we need to we need to

34:43also focus on interoperability and crosschain stuff because while I mean the blockchain world can feel big when you're in the middle of it. It's growing rapidly. It's like a tiny speck compared to the centralized tech world. Huh. I mean, one thing we've gotten from Greg's technology with ASI chain is I mean, we're going to launch hybrid nodes between ASI chain and Cardano, Ethereum,

35:11Cosmos, and then other chain. So, you should be able to build crosschain liquidity pools like stake tokens in one chain and get rewards in another chain. You should be able to run an AI process where part of the network is on Cosmos, Cardano, Ethereum, part of the network is within within ASI chain. And I think how all that will evolve is quite

35:36flexible. But we need all the different chains to be cooperating together as much as as possible to sort of overcome the advantages that big tech has in terms of money and momentum.

Yeah.

Yeah. Definitely. We need the capital to flow as easy as possible to where it needs to be. So I think that interoperability is important. And now we don't have a lot of time left,

36:03Ben, but with the [clears throat] ASI chain in DevNet and then moving to testnet, what can develop developer-minded people do right now or if people aren't developers, when and how can they get involved in this? So with DevNet, you can jump in and see what's there and understand the architecture and you can start prototyping AI software or other DeFi tooling,

36:32whatever is your thing on top of the DevNet. for AI developers. I mean, you can dig into the hyperaround codebase in GitHub and start experimenting with how to use different sorts of AI technology within your AI workflows, which could be AGI or it could be as simple as like using using a more flexible knowledge metagraph inside Graphrag for an LLM or something. I

36:59mean, there's all sorts of things you can do. Now when we move from DevNet to test net it gets more interesting in that will be more like the devet has the structures there but it's not yet scalable so you can use it for prototyping when we have more scalable consensus which will come with the devet like like it's just some we're moving [clears throat] some components from

37:22scala to rust basically then then we'll have a lot more scalable then then I mean ironically scala is not actually the scalable language in spite in spite of the name, right? But

yeah, I mean I mean then it still won't be with full tokconomics, but you but you'll be able to put AI applications on the test net and run them on a big a

37:50big network and sort of sort of see see what they can do. And there's a so there's a lot for developers to play with. Mhm.

Abs. Absolutely. I mean, from an investor standpoint, I mean, I'm not I'm not a money guy. I'm an AI guy. I'm not going to hawk our token, but you can you can you can look and see like a the fetch token may be a very good deal at

38:15this particular moment due to recent historical events that have have pushed the price probably below what it's it's logically worth, right? So there's

lot of interesting things there. Now tokconomically next year when we launch ASI chain

there's going to be very interesting opportunities to sort of buy a license to operate a shard and then kind of

38:46monetize your AI or DeFi ideas on a shard and then profit from the LST emissions on that shard. And we've tried to learn a lot from [clears throat] dynamic towel. It's very interesting what they did with the subn network economy. On the other hand, we're working hard to make sure the tokconomics and ASI chain reward real products that are generating revenue

39:12from delivering valuable [snorts] services.

Yeah. No, I love that you're switching and a lot of ventures are switching to emissions from revenues and you know sort of the traditional business rather than just market cap to start you know stockpiling tokens and I can leave more information on the ASI chain launch the DevNet just

39:33singularity net and the ASI alliance and the fetch token everything else in the show notes below because we covered a lot of topics. I really appreciate your insights into all of it, decentralized AI, Sophia and her family there that you're showing in the humanoid robots and wishing you and the team all the best and everybody in the ecosystem growing together and I

39:57would love to follow up again in the near future.

Cool. Thanks a lot. Yeah.