Join host Stephen Sargeant in this episode of Around the Coin for a deep dive with Todd Ruoff, CEO of Autonomys. This network provides the infrastructure necessary to scale decentralized AI applications on-chain. With over two decades of experience on Wall Street, including C-suite executive roles at Ruane, Cunniff & Goldfarb, Todd blends his traditional finance expertise with a focus on integrating blockchain technologies, which he began exploring professionally in 2018. His extensive background in investment operations, trading, and compliance enables him to adeptly navigate the complexities of cryptocurrency regulation. Passionate about education, Todd currently lectures on cryptocurrency for finance students at Rutgers University, imparting insights from his deep understanding of both traditional and decentralized financial systems.
Buzzsprout • YouTube • Quora • Medium • X • Facebook • LinkedIn • Soundcloud • Apple Podcast • Spotify • Player FM
Stephen: This is your host Stephen Sargeant. Welcome to another edition. We got it all in this episode of Around the Coin. We have the CEO of Autonomys. We have Todd Ruoff. We go into Deepin. We talk about the layer one that they're building that has storage, that has interoperability, sustainability. And accessibility.
We talk a lot about the balance between privacy and transparency. We talk about this background. Todd, you know, was in New York at the time of, you know, the global recession, as well as 9/11. So we talk about the state of affairs during those times. We cover everything in this episode, including why ownership of the data is important.
He gives some really amazing use cases. So you might want to tune in and listen how you can protect yourself. And we go deep on AI. No, not the fluffy stuff. Real AI, AI agents. He breaks down some of those complex terms that you don't know about, but you see him being tossed around at every conference. It really explains where AI is going to take us in the future and maybe how it can make us a little bit of money.
And this cool project that they built on Twitter, that's literally capturing the two crypto Twitter arguments. Absolute fan of the Todd podcast. Hope you guys like it. Reach out to Todd and me if you do. And we'll see you again next week.
Stephen: This is your host, Stephen Sargeant, the Around the Coin podcast. We have Todd Ruoff here from Autonomys Lad. CEO we're going to go into your little background a little bit and kind of get to know exactly how you ended up in crypto. Like most people, it's usually not a straight trajectory but maybe just explain to people a little bit about what you're doing there.
And then we'll dive down deep.
Todd: Yeah, it never is a straight trajectory. What we're doing here at Autonomys is we are building a novel layer one network, which the Autonomys Network is the first ever implementation of what we started as, which was the subspace protocol. We just launched our main net a couple of weeks ago from our foundation in Switzerland.
And it uses a novel consensus mechanism built on proofs of archival storage. And has some really unique use cases. So hope we can drill into it.
Stephen: I'd love to like, just start there. Maybe, you know, when people think of Bitcoin proof of concept maybe proof of work, maybe you mentioned something there. Like. Maybe this tells how your protocol differs than maybe the average, either DeFi protocol or Bitcoin that people pretty much understand. Well,
Todd: Sure. So when someone poses a transaction to the Bitcoin network, for instance, which is based on proof of work, all of the ASICs and CPUs start trying to crunch the mathematical computations, depending on the difficulty in order to. Solve it or get as close to the solution as possible. And of course, the one who solves it wins the block reward.
So it's a tremendous waste of energy because you have, you know, one node wins and 10, 000 losers even more. But nonetheless, that is the basics of kind of how the proof of work model works with proof of archival storage. The way our consensus works is users are actually just pledging excess or idle hard drive space that may already be existing on their computer.
to the network. And what happens is you download our small software client. You don't need any tokens to get started, like in proof of stake. You don't need a fancy rig like in proof of work or GPUs or anything. It runs on commodity hardware and you download the client. And if you have a GPU, you can plot over a hundred gig in about 15 minutes.
And what that does is it encodes to that. Empty space on your hard drive, the actual blockchain data itself, which is encrypted. When a challenge is posed to our network, what happens is the users who are running nodes on the network called farmers. They all start scanning their hard drives to see if they have the solution.
And of course they're scanning the encoded blockchain data. And the person who is the closest to the proper solution, they win the block reward as well as rewards for the other clients or nodes that were the closest. to it, and they count as kind of voters. So we're using hard drives for consensus, which is super environmentally friendly.
After you plot, there is next to no power consumption. And it's very efficient. And of course, because it is based on hard drive space. The network is very conducive to storing just vast amounts of data. And the data really becomes an asset rather than a burden. Like you see in some networks, like a theory.
Stephen: I love that. I love that. For me, that has no space on their hard drive. It might not seem like a, I'm sure there's more commercial users that are like, yeah, I got lots of space. But for me, I had to like delete half the things this just this morning. Talk to us. You were always in investment and working with training, but venues.
Tell us a little bit about your 20 years there. And then was there anything that you see like, Oh, now you're, you're deep in the blockchain and deep in space space. Is there anything that you see now that you're like, Oh, that would have been an easy solution for this, you know, 10, 15, 20 years ago.
Todd: Yeah, kind of almost the opposite. And I'll explain my background was originally in technology. I started out in it and graduated with a degree in it and finance. Thinking that I wanted to go work in on wall street with technology, which is what I did. And I started on the technology side and then migrated almost fully over to operations and trading, working in the investment advisor and broker dealer space and running the non investment side of the company.
And I did that for about 20 years and had all my trading licenses as well. And as, as time went on I really. Had taken an interest in blockchain and in crypto, the whole web three space, which I see as kind of the convergence of finance and technology, which are the two things that always interested me the most.
So that's kind of how I ended up moving into this space. And I started working in about 20 about 2018 is when I started in the space. I've been here with Autonomys for four years now. And what I was alluding to before is. I look back to when in the early days of Wall Street, when technology was really first coming on the scene, there were a lot of inefficiencies in the traditional financial markets.
And one example would be fragmentation of liquidity. There were the New York stock exchange and the American stock exchange and the NASDAQ, but they were all decoupled. And you had to go and look for where you could find the best liquidity. And then Algos were developed to be able to do that and then smart order routing to send it to the best exchange and then they came up with this idea of the consolidated quotation system Which shows you the national best bid and offer and it doesn't matter Which exchange it's on if it's otc if it's on the new york stock exchange or somewhere else Consolidates it all together into a single quote the bid in the ask and here's the price and Those types of things that the market has evolved along with in traditional finance I think represents some of the inefficiencies that we still see in traditional crypto markets today.
And you'll see things like. I think a consolidated quotation system where you won't have this fragmentation of liquidity, where you have to go check Binance and Coinbase and crack in to try to figure out where are you going to get the best price? And then look on Uniswap that needs to be aggregated and put into a single place where all the users can see it.
So you don't have to be exchange specific. So that's one example, but I see a lot of synergies like that, that the trad find market has figured out over the last 20 years, but some of those challenges and custody would be another one that still kind of exists in the crypto space.
Stephen: I think our, all of our arbitrage listeners just, just stop the podcast. They're like, no, this guy's talking about us. Basically we're to take away from a lot of the arbitrage, but for the, you know, everyday user, it's exactly that, right. Trying to run from one exchange to the other, seeing if there's a change in price, if you can get more depending on jurisdiction.
Which, as you say, is extremely inefficient. Now, you're in New York during, you know, the global recession and during 9 11. Can you maybe, any thoughts or memories that you had about those days or around those times that kind of still stick with you today?
Todd: Yeah, unfortunately there are so I was in New York facing south from my window in midtown Manhattan when 9 11 took place and I did witness it. And it was a terrible thing. But what was really just kind of, I think, exemplary of People and mankind was seeing how New York in particular and then the whole country and then the whole world kind of came together to get through that.
I thought that was a great example of our resilience and everyone was overly extremely nice at the time and very accommodating with anything they could do. So that was tough, but it did have its bright spot in that it kind of, I think, unified The, the country and of course the city and then the global financial crisis.
Yeah, that was pretty devastating as well. I had friends that worked in the, in the fixed income markets and that worked at Lehman brothers, which got wiped out when that took place. But fortunately for myself, I was working for a very boutique established. Investment advisor, and we were mostly kind of value driven in our investment approach.
So while we were affected, it wasn't quite as devastating as it was for much of the market, but yeah, those big drawdowns you tend to remember those cause they really stick with you. So, yeah,
Stephen: Was there any time that you thought like, hey, we're not going to recover from this? Like, both of those are like monumental, similar to the pandemic. You know, historical times. Was there any other thought like, Hey, maybe I should go into another career trajectory, get back into it and tech. Was there any thoughts of that during those either one of those times?
Todd: yeah. You know, it's funny. I think there always is you know, and I've been through so many cycles now in both with both you know, world events with gigantic market movements in both tried and crypto. And I still find myself every time we get into a bear market, like we just had with crypto for the last few years thinking, boy, is this.
Is this ever really going to come back? And you know, you watch all the podcasts and you know, people like yourself and everybody talks about it as though this is just a speed bump. You know, this is just a little part on the road. And I wonder how much people actually believe it. Because, you know, I say the same thing when I message it out, but I think somewhere deep inside, everybody has that.
But it certainly builds resilience. You have complete confidence, like the big drawdown that happened in our markets in crypto. The other day where, you know, some of the altcoins went down 20%, there was no question in my mind that it was going to bounce right back. So I think it's very circumstantial, but yeah, it certainly makes you more resilient when these things happen and knowing that it's probably not the end.
Stephen: Some of those all coins, you're like, yeah, maybe they should stay. They should stay down. They're not doing anything for anybody. But that brings us to Autonomys. The company is backed by some well known players, including, you know, Pantera. And I believe Coinbase Ventures. It says you're building something called AI 3.
0. And you know, I think it was Jack Dorsey that, you know, just as we got into web three was talking about web 5. 0 and like kind of leapfrogging us is, you know, web 3. 0, very consistent. To your AI 3. 0, where it's like, Oh, we're in the beginning of AI, but let's let's marry AI with blockchain and see how quickly we can go.
Todd: I would say very much. So Chris Dixon from a 16 Z had a statement and he has a book that, you know, the first iteration of web was web one, and that was read when you could go and you could read on various websites, web two, you could read and write, you could interact and send information. And then of course we get to web three, which is read, write, and own is his definition.
Very similar with our AI thesis. We. Really define AI one as the early days of machine learning, which is something that was. Always there in the 2000s, but it wasn't really as prevalent of a news topic as it is today, because that brings us to AI2, which is generative AI, the things that we're using today, like Cloud and, and ChatGPT, and some of these LLMs with interfaces that we're able to use, so that everyone has access to it.
With Web3, It's this concept of human centric AI where we're actually empowering individuals and giving them control and ownership over their data versus it being, you know, managed by a few big players, which is kind of what we see today in the ice space. It's about democratization of data access. It's about sovereignty.
It's about transparency and all the aspects of that combined together. To kind of make a pretty, pretty good similarity, I think, to what Web 3 does to Web 2. That's what AI 3 does with to AI 2 as well.
Stephen: I wanted to, cause this was in web three, two, you know, decentralization on our data, what are your thoughts? Do people really care about that? Or like maybe provide me a use case where you're like, people probably say they can, like, it's just like, people say they care about privacy. But like, like, get them on the Bed Bath Beyond website and they'll give their blood type to save 10 percent on a flannel bed set, right?
It's like, we care about privacy until we want convenience, where it's like, every time you open up a website, you're just like, cookies, except all, except all, except, like, you're not customizing anything, you're not, you're not even reading it, it's in different languages. You could care less. Give me one reason why people should start caring.
Like, maybe understand, like, Play Devil's Advocate. I understand why people don't care that much, but say they do. But this is the real reason, or maybe the future reason, that you're really gonna need to start thinking about these things.
Todd: sure. Have you ever received something in the mail offering you free credit monitoring because your personal data has been exposed or stolen and now they have to offer you free credit monitoring as part of a lawsuit settlement?
Stephen: I have not, but that sounds very interesting.
Todd: I've received at least a half dozen of those over the years, and I think most people in the U.
S. Have with these big data exploits. I think the Distinction there largely depends on the individual. There are certain people, especially in our space in web three, that really hold true to the crypto values of ownership and privacy. Whereas I completely agree with you. People are willing to give away their deepest secrets just so they can, you know, get, get a discount as you had said.
But I think that given. The option, especially as we go even further into the digital age, more information is going to be put online than we even have now. And with the emergence of AI, and you're going to have an agent that is trained potentially that knows everything about you, are you comfortable with that?
Potentially? Being controlled by someone else or having AI models trained on your personal information versus now, where maybe you think, well, it's in a database, you know, they're too lazy to look at it, but as we get into the age of AI, you have to really think about how this data can be used and potentially used against you.
And I think that that's something that will really start to change people's mindset, especially with the younger generation, because this has been a point of concern for them their whole lives versus the baby boomers who don't understand a lot about you know, what we're talking about here, but I think given the choice between having autonomy, privacy and transparency, or just having all your information out there.
I think given the choice, people will opt for the former because it's there. I think now they don't have a choice. So that's why everyone is so open and willing with all of their information.
Stephen: I think that makes sense, right? The Google and the bigger players, they have the capital that they can, like, retarget you every time you go clicking a website and pixelate and send you advertisements. But if everyone gets access to that using AI and you're inputting, you're training these AI, they're going to be able to step by step.
They're going to be able to be proactive versus they're not going to be able to wait until you Google anything. It's already going to start showing up on your feed just by based on what age you are and you know what, you know, time of life. So now they're going to be sending you, it's going to be abused almost similar to like email where it's like you can barely open up anything or give your email address.
But then I think to your point, even what if the hackers get in touch with that intelligence? And they have access to knowing everything about you. It's going to make those, you know, so many attacks way more sophisticated. Can you give us some use cases about what people are using your network for now?
Whether some of the applications, whether some of the projects may be not fully built yet that you're just excited about.
Todd: sure. So we just launched our main net phase one last month. So we've been live for about four weeks. Our network has kind of two components. There's the consensus layer. And then there's the compute layer. The consensus layer is where the storage takes place. That's live today. So farmers are downloading our farming client and they're pledging disk space to the network.
And we have over 400 petabytes pledged to the network and over 2100 farmers participating in consensus and helping to store the data. Some of the use cases that we see, there are several. 1 is for a lot of the. The future, which will involve AI agents and you're hearing this everywhere is the buzzword, but this is something we've been working on and talking about since before it really became popular with the advent of chat GPT and so forth.
And how these agent to agent comms are going to take place. And, you know, if your agent is talking to my agent, How do you know that it's really my agent, just like you get an email that says a person's name, but how do you know it's them? So one of our domains that will exist on top of the consensus chain is called auto ID, and it is just a feature it's built in.
It's not necessarily like a core product offering, although it can be for some people to give agents or AI models or people a unique. ID that cryptographically proves their identity. So in using that, they can communicate confidently with other agents, knowing that the agents are who they say for. So ID is going to be one thing that I that I think is very important.
Another one would be for The real time payment rails. So all of these communications that are going to be taking place between agents and between people and agents and so forth. This is going to cost money. There is a tremendous amount of data that's going to need to be stored. And, you know, it would be much easier if we had a way where we could affect this instantaneously with real time settlement versus using a credit card.
Or doing an ACH transfer and you get it in three to five days. This, it can happen at the snap of a finger. So that is one storage is also a really big use case for us. There's going to be a tremendous amount of data that is going to be generated by AI and agent to agent comes. And as I mentioned in just a month, we have over 400 petabytes of storage space that's available for.
people to be able to to build on. The last thing is we have a product called auto drive which we will be putting out. It's more of a proof of concept where individuals can go to the website. You can click browse, you can select a file from your hard drive, you can upload it to our blockchain in real time and it doesn't cost anything.
We're subsidizing it for the time being and it's there and it's. Permanent. We have data permanence, which is very different than a lot of the storage chains that are out there because ours is built in a way that's incentive compatible. And then that goes into another agent that we're building on top of that.
But maybe if we go into agents a little bit later, I can talk about that.
Stephen: I love that. I want to actually go there. Cause you know, you were seeing right now with AI, you can't believe anything that you see visually. Right before I was like pictures for, you know, Photoshop now it's like video. You're like, I don't know what I'm looking at. I don't know if it's real. Everyone could say, oh, that's not really me.
That's AI. And I think, you know, the consensus with a lot of people, especially in things like marketing. Finance is like, we're going to need to put all of this on the blockchain because the provenance we're going to need to prove like, yeah, I put this through, I put this through Autonomys like two weeks ago.
And this is the exact original file. This other file that you're seeing out there. This other video is not, you know, is not the same thing. It's been altered in some way. Do you feel like that's kind of why you're bringing out the auto drivers like Most of the information is going to have to go on some type of blockchain because in two to three years, we're not going to be able to believe what we're seeing anymore.
Todd: I would totally agree with that. And it's not just auto drive. Auto drive is for storage, but auto ID is for identification. But when you put the two together, you have cryptographic proof of the origin of a piece of material. That's really what we're trying to capture in order to get both. Anyone can go and mint something onto our chain, and it is stored in perpetuity until the end of time.
But in order to verify the origin, of that, the cryptographic proof that's needed in order to, you know, demonstrate that it is who it says it is. That's where auto ID comes in, which is kind of a complimentary product.
Stephen: And I see just like any good protocol, you guys have a huge partnership ecosystem. One of them that I know very well is how boring, what part did they play? Like it's a security audit firm. Are they, you know, working with all the other applications? Making sure that their code is right and can't be, you know, exploited by hackers.
Like where do they fit into your ecosystem? And then maybe mention some other unique partners that you have there as well.
Todd: Yeah. So how born we did partner with for some aspects of our security, primarily our Blockchain is based on the substrate framework. So the premier auditor in that space is security research labs. SR labs did polka dots audit for instance. And many and most of the projects that are built in the polka dot ecosystem.
So we've been partnered with them for the last three years for kind of an ongoing audit. We did get our final audit before mainnet and we still have audits going on in perpetuity until we get to. To mainnet phase two. So yeah, they have been an integral part of it and we are just now getting to the point, we have signed over 40 different partnerships which many of which are co marketing and many will be to build and deploy things on the blockchain like agents.
And again, it's a, it's a novel. Base layer one. So it can run anything. We will have an EVM domain so people can deploy Ethereum smart contracts. And we're just now getting to the point where we're going to go and start onboarding builders to bring in the demand side of the equation versus the supply side, which for us as the farmers contributing storage and the, you know, over 2000 nodes that we've been able to consolidate and make the network in, in less than a month.
Stephen: Awesome. You just brought on the chief marketing officer and you said a lot of your partnerships are on marketing. It seems that, you know, as a protocol layer, you're always contending with, you know, grants from everyone's jumping from one protocol to another, trying to get as many tokens as they can, as many grants, it's hard to figure that there's that many developers and users that stick with the project.
Like, what are you, and I believe his name is Peter, what are both of your thoughts on how you can, you know, get that stickiness and traction where people want to build and stay long term on your project?
Todd: Yeah. Stickiness is a big thing. And yeah, yes, I do work closely with Peter on this and a shout out to our fabulous integrations person manager. His name is part. He's been with us since about the same time Peter joined and, you know, we've found that it works best through incentive alignment. We have made a generous.
Token contribution to the treasury, which is over 15 percent of the total supply that is all for the betterment of the ecosystem and the community. And again, that's run by the Swiss foundation, which we wanted to do it in Switzerland because we wanted to have you know, really good optics. It's not easy to put up a foundation in Switzerland.
We could have done it in some other jurisdictions, Singapore, the Cayman islands in, you know, less than
Stephen: Even with a lot better annual retreat, I think.
Todd: Yeah, so we spent over a year getting this thing established so that we could do it right. And with 15 percent of the token supply, the foundation and the foundation council, which is the governing body, and through governance, which will be introducing in later stages of the protocol. There will be able to kind of direct these funds as to where they're going to go, and most of it will be used toward grants, which is exactly what you're saying to entice people to come in and long term incentive alignment.
We have another project that we are working on. We haven't really. Come out to announcements to announce it yet, but we're looking into something where we can essentially launch like an incubator network, which will reside on our blockchain and that incubator network will allow people to come in and support those projects that are being incubated and receive rewards in the tokens.
From those projects. So this is something that we still have under development, but we think it's going to be really exciting and is absolutely going to attract people into the ecosystem and make them want to stay.
Stephen: You know, you were telling me, I think before we were recording, that you used to be the CFO. How does protocols like this make money? I think people understand, you know, there's a lot of tokenomics involved. But they also see these protocols giving away, you know, hundreds of thousands, millions of dollars.
Like what is the business model for most of a protocol like yours? You don't have to give me a, you as an example, but what is the business model? Are you making, you know, on transaction fees? Is it, you know, people investing in the company? Like where, where would somebody, if you're investing in this, get your return on investment?
Todd: Sure. So the company itself, the development company was obviously seated as you had mentioned earlier by some investors. We have a really what we think is an elite roster of investors that we've partnered with, and we've been lucky to have helping us. But from the blockchain side, in addition to the tokenomics as I'd mentioned before, it's primarily about the storage and commoditizing that storage.
So for a user who's going to pledge his hard drive space and now participate in consensus, there is a dynamic cost of storage that is calculated by the protocol, which essentially looks at demand. Looks at supply and determines about what the cost of storage should be to maintain that equilibrium. And as the cost of storage rises, and it becomes more profitable for farmers, additional farmers will come on to the network to meet the higher demand of storage and then maintain that equilibrium.
Quite the contrary. When the process, when the cost of storage goes way down and it's not as profitable for a farmer, maybe some farmers will leave the network. And at that point again, it balances out and there are farming rewards that come into play. There is also staking rewards when you. When you get your block reward tokens minted, you can then stake them with executors and operators on our network so you can earn additional yield.
So that's kind of where the ongoing revenue comes from. We nor the, neither we, nor the foundation takes any fees from transactions or, you know, gets a piece of, of anything that's happening. We are invested. In the ecosystem by holding the token much like everyone else. So we're all kind of partners in that, but we have no additional benefits of, you know, additional revenue streams or anything like that.
Stephen: If you grow it and it blasts off and everyone's dedicated and puts their effort to it. The entire ecosystem goes up in value and everyone has a percentage of that. That's, that's awesome. Ambitious. Awesome. Yeah, awesome. Talk to me about mainnet, you know, like where are you looking at? You know, after a month, this is like, how, like, are you like, how could this break?
We just, we, we spent so many times, so much time planning this. And there's this little thing that goes wrong. Like what, what issues do. Maybe not even so much your company, but others that you see building on the network or about to build on the network. What are the issues when you flip that switch on for a layer one blockchain?
Todd: Yeah. Fortunately, we have a really talented protocol team that orchestrated this and in real time and live during a community call, we launched the network and the video is available on our, on our website and our YouTube for anyone that wants to watch. But that's what made it. I think especially challenging, but also especially exciting because you can't hide when you're streaming something in real time.
And fortunately it went off without a hitch due to some great planning and proper scripting of the sequences. And we were able to launch the chain and start seeding the storage in real time with everybody watching, and it could not have gone any better. So we were very lucky, but again, I attribute that to the skill of our protocol and community teams that, that really built this out.
To the point of sometimes when things don't go well, and you do see that a lot, unfortunately, I think it's when people don't anticipate. Anyone can just launch a chain and have a token, but to do it properly and to think through all the possible scenarios that could happen in the future really is important.
And you see this with a lot of big projects that have launched recently. I won't name any, but you know, problems with their tokenomics where maybe. You know the, the investor their tokens are locked, but they were able to stake their lock tokens. And meanwhile, because they were able to stake all those lock tokens, they were able to take out 60 million of staking rewards after only a few months of main net because they didn't lock the staking rewards.
Well, you know, that's one of those things that I think. People really need to anticipate because that leads to a complete imbalance of liquidity and the ecosystem versus what people were expecting. You know, and of course the community is upset because as far as they're concerned, those were supposed to be locked hook.
And so, so how does that happen? And it's just thinking through every iteration of what could go wrong. And I don't think anybody is capable of doing that perfectly. Neither will we be. We were very fortunate with mainnet phase one. With mainnet phase
Stephen: like neither will we be, but we're off to a good start.
Todd: could happen, anything can happen, but we try our hardest.
I can say that.
Stephen: I love that. Is it tough? Because, you know, you know, coming from that web to where like, Hey, if you're Zuckerberg and obviously you're working with people, but it's like your company, your decisions, if it fails, it's on you. Whereas like you're in web three, you're, you almost need to build with the entire community.
And to your point, when something goes wrong, it's not just all you're disappointing, like shareholder value, or you're disappointing the company people on your team, you have tens of thousands of people relying on this. You know, to go a certain way, is it kind of hard? Cause essentially you're building in public at that stage.
Is it, do you find that a little bit more difficult, especially with, there's a lot of community pressure, they want the token to go out. So, you know, sometimes it might be a little bit short sighted in the actions that they want the foundation to take.
Todd: Yes. And I see that quite a bit where there is some, some short night, some short sightedness. People want liquidity yesterday. Whether their token is locked or they received rewards and they can't access them yet. Everybody wants them. And then you have to wonder what kind of hands are you putting those tokens into?
You know, you want diamond hands instead of paper hands. And the people that really have the belief in the ecosystem, they're the people that we are really trying to reach. And they know that for the long term betterment of the entire project, there are going to be some sacrifices or some things that may not happen as quickly as they would like, but it's going to be for the trade off of having it done properly.
And I think anyone. Would rather have it done properly than just a kind of a quick hit. But again, you know, this is web three and people want things, you know, they want immediate gratification. So versus the web two world, yes, it's very different. It is a lot of responsibility that not just myself, but our entire team feels to our investors, our ecosystem users the team itself, our ambassadors.
And that's something that we take very seriously. And in order to, you know, kind of. Keep pace with everything that's going on in a very transparent manner. You know, we deploy all of our resources to make sure we're always communicating with the community in our discord, in our telegram, on our, on our ex account to make sure that we're doing things as transparently as possible.
And I think that's the most that users can ask for. And it's not something everybody does. But when you do it, the people that really matter the most to the project are going to appreciate it. And those are the ones that you want to keep around.
Stephen: I love it. Is it nice to know that you've been building, especially with this team for four years? And it's almost like Deepin has finally come to you. Whereas like four years ago, nobody cared about Deepin. It was NFTs. Then we went to the metaverse. Then we jumped to AI. Then we jumped to DeFi and Bitcoin.
You know, but in the background, it seems like, you know, infrastructure, the picks and the axes and the shovels, like. That's always been on investors mind and, but it seems like now it's getting to the mainstream where even, you know, your average meme meme coin holder is kind of getting interested in deep in again and T and all the acronyms.
Do you see that? Or is that just me? Am I a little too close? I see it more often. Or do you see when you're going to events and conferences that a lot of people are getting more and more interested to deepen,
Todd: absolutely. We position ourselves as deep in as well because of the storage component, but there are a lot of networks out there that are just selling storage. That's the only. You know, real reason for the protocol. Ours is more of a packaged offering where we have the storage layer on the consensus chain.
We have what we call our enshrined domains, which will run on top of it. That's the execution and the compute environment that will be coming in our main net phase two. And by bundling those two together and decoupling the execution from the consensus, you know, we let the farmers worry about producing the blocks.
And then we have the operators. That are there to perform the compute and order the transactions and maintain the state of the chain. So by kind of decoupling those two, it's made it really kind of a unique offering that, you know, we can offer permanent decentralized storage that is going to Be there and accessible to anything in the domains that are on top of it.
And those domains can be customizable environments that a project may want to launch a domain to do any particular particular application. And they can customize that execution environment to support whatever their unique use cases. So I think Deepin is a big part of that, but. We're not just storage and we're not even just storage and compute, it's storage and compute, but we are really tying in the application and use case layer, which we see as the AI component.
Stephen: which makes sense. Cause I think I read that your core pillows are interoperability. You hear those terms a lot, as you said, there's a lot of, you know, protocols attacking one of those pillars, but you're attacking all three of them. Is there one main pillar that you have to get right in order for the other two pillars to make sense or to, you know, be successful?
Todd: I think it's kind of all three. And if one falls, the rest of the ecosystem potentially crumbles till now the obvious, you know, critical component was the consensus chain, because if you don't get consensus, right. Everything above it is exposed, and that includes the execution environment. And our Dilithium consensus that our wonderful head of protocol, Daria, and some of the PhDs we worked with, Dr.
Chen at University of British Columbia you know, put in Months and months and years into our dilithium consensus to make sure that we got that right. So that is fundamental. But then, of course, when we put the domains on top and we start tying in the A. I. Components that are going to interact with that.
They are also equally as important because if anyone piece breaks, the end user is affected. Do they care that it's because the Okay. Consensus chain had a bug or the execution layer failed. They probably don't see that, but to the token holder, it would obviously matter because the consensus layer is what is really going to you know, drive the, the, the token The token float as well as the value to an extent, but a lot of the value accrues at the higher layer when we get into execution, things like governance and some of that incubator environment type tooling that I had talked about before.
Stephen: That's super interesting. One of the things I found when I was reading it here is that, you know, the question, especially a lot of privacy protocols come on the podcast. And they talk about the need for privacy, but I come from a compliance background and there's, you know, especially from trad five, there's a need for a certain amount of transparency versus privacy.
What are your thoughts on this dichotomy of privacy? It seems like you're touching both of these areas. With some of your prologue was parts of your protocol. What are your overall thoughts about how to approach that privacy versus transparency?
Todd: Yeah, I think it's a balance. So, you know, privacy is for the individual, for sure. You want to protect your own data, whatever you know, your social security number, your driver's license number maybe a site needs you to prove that you're 18. So they make you upload your driver's license and it has your birth date on there, but you've just given them all of this additional information about yourself that wasn't necessary.
Well, what if that could be put into a ZK proof where the data is. Encrypted. But the proof can still prove to whoever the requester is that a particular thing is true. Are you over 18? Yes. And here's a cryptographic proof proof, but I don't have to turn over everything that that I have, you know, my driver's license.
So that's one example. And I do think that privacy is really for the individual. Transparency, I think, is very important in the public domain when we start to get into areas where users demand it. Like AI you know, how are these models being trained? Is that something that we can put on chain so that everyone can see it?
You know, I ask it questions about, you know, medical conditions or maybe about my politics or who knows, religion, finance, anything. Is there any bias that's been introduced into that? What kind of training data was really used? And is that the type of thing that needs to be out there in the public domain so that we know what type of data we're interacting with, how these models were trained?
So that would be an example of something where I think transparency is really important, but privacy, I would say is equally as important for the individual of when we start to get into you know, things that are going to matter most to the person who holds, who the information owner is.
Stephen: I really liked the way that was that privacy for the individual transparency for the ecosystem. We all need to see that, make sure all the actors are acting in good faith, but we don't need to know particulars about the individual actors. I like that. That made a lot of sense. What are your views on something like tornado cash?
I know, you know, it's going through the courts. It's being reversed, I think, from the initial sanctions designation. But you're also, you know, you're in the U S maybe the foundations outside the U S but the U S has a strange way of, you know, overreach in a lot of areas. What is your thoughts, especially now with Trump and like, where are your thoughts about the regulation in the U S and what it needs or what you need from those regulations?
To be able to thrive and, you know, obviously one of the most important markets around the world
Todd: Yeah. Something like, you know, we, we are not in the privacy protocol per se, like, like a tornado cash type smart contract. And that is really a mixed bag. And I hear arguments from both sides that are. Absolutely. True. In both cases, you know, from one perspective, the government is saying, and they've proven this, that, you know, this is being used to launder hundreds of millions of dollars and steal all this money and have it sent to bad actors who wish to harm the United States.
No, that's not okay under any circumstances. And there has to be controls in place for that. But for the individual, I can understand the need for something like this. There is a lot of, you know, public figures, podcasts, hosts, maybe like yourself and so forth that have a big on chain presence. But how do you handle your finances privately?
If money comes in to, you know, somebody sends you money, Steven, and it's on the blockchain and the whole world sees you just got sent that money. Well, what are you going to do with it next? Does the whole world need to know how you privately direct your, your, your funds? I think that can be handled through things like zero knowledge proofs and so forth going forward versus things like mixers and tumblers like tornado cash.
It was interesting to me that they modified the sanctions component, and I am also anxious to see with the U. S. Trial coming up if that's going to be affected at all by the change in the political landscape. I don't think it's going to help his partner in Europe, unfortunately. But, you know, this opens up a whole nother area of to what extent is the developer of a technology responsible for how it's used.
So you know, are the People that created tornado cash culpable of someone using it for money laundering. Any more than Microsoft is culpable for someone using Excel to keep copies of fake books that they're going to use to submit to the IRS. Are they responsible for that? So
Stephen: sexual exploitation, like Facebook, like, you know, or WhatsApp for, you know, pig butchering scan. You're right, it's, it's, it's, it's just a layer on, it's other people, you know, feeding the pipes. This is just a pretty much a blank slate on that topic, though. What are your thoughts when I, you know, when I think about this.
What are your thoughts about something like what's happening in the U. S. now where, you know, they have a new crypto and AIs are, do you have I don't know, David Sacks, I watch, I listen to a little bit of the All In podcast, so. Are you thinking that this is what the U. S. needs is a little bit, someone that's in, that's, you know, been part of Web 2, excelling in Web 3 that can provide insights on where the guidance should go.
Todd: hundred percent. I also have listened to the the all in podcast many times. I've done some research on, on David Sachs and, you know, I know that he is an entrepreneur. He is very experienced in tech and Silicon Valley and how the markets operate. He is certainly going to be much more favorable toward crypto.
Now, granted although I don't think the roles have been explicitly defined yet, remember we still do have Other regulatory bodies like CFTC that are going to be responsible for regulation to some extent, but in terms of guidance and someone who's going to be out there advocating for these two Andrews industries, I think it's important that we get.
Someone in there who has a good working relationship with our president, which clearly he does but also has their finger on the pulse of technology and knows what's going on. He knows the mind of the of the investor and he knows the mind of the entrepreneur. So when you put all that together I think that that can be a powerful combination.
I don't. know if I can say how much political experience he has, which we all know is going to be a big challenge. I'm sure. You know, the Hill is a much different place than the Valley, but that remains to be seen.
Stephen: I love it. I love it. And yeah, there's, I think everyone remember, has to remember, there's more four letter acronym agencies than just DOJ out there that you have to deal with. They think just Elon and Vivek are going to be able to change everything. Walk us through some stuff. So, you know, very new. I'm, I think I understand AI, but now just like web two, people are adding in a lot more layers.
We have generative AI, we have LLMs. Can you walk us through, if I just give you the, you know, the term and you walk me through a basic, you know, one or two sentence about what it is and like, maybe why it's important or what distinguishes it from something else.
Todd: Depends on the four letter word.
Stephen: makes sense to a lot of people, but like, what is generative AI?
Todd: Sure. Generative AI is where you can essentially give a plain text. Argument or input to a large language model, something that has been trained on patterns and repetitions and mass amounts of data and is able to sequence it. That is the large language model. And then the generative AI part comes when you give that a prompt in natural speech and it is able to use its training of the large language model to generate a response for you.
Now that response could be in the form of text. It could be a picture. In the case of Sora, it could be a video that we're seeing now. So yeah, and then there's the whole multimodal inputs, you know, you can input a photo and it can give you back text about what's going on in the photo. So the generative can, you know, go either way, but that's the whole premise of generative AI.
Stephen: And then, like, is there a difference between AI agents and something like ChatGPT Perplexity? Are those one and the same, or are those a type of AI agent? Like, how would you describe those?
Todd: Yeah, I would describe them as components of the AI agent. So in the case of Autonomys I'll, I'll explain something we're doing and, and how that kind of overlays with it. So we created a proof of concept AI agent called Argument, M I N T so it's on Twitter if someone wants to follow
Stephen: Ha ha, on Twitter. I love that. Play on words, on
Todd: shameless plug.
Yeah, shameless plug. It's zero X argument. And what we've done is we have pointed it towards a lot of the Web3 conversation online. And it is out there to basically take counter positions and to offer you know, maybe some opposing point of view, not necessarily be contentious, but in some cases it may be.
And that is consistent with probably 99. 99 percent of the Agents that are already out there and the difference between the agent and just chat GPT is the agent is acting Autonomysly. It's on its own chat. GPT is waiting for you to tell to do something. That's the interface to the language model.
So argument is sitting there and it's waiting and it's automatically on its own engaging in conversations. We have it connected to a rag model an augmented generation retrieval, augmented generation, where it can take the tweets, it can, or whatever you want to call them on X, it puts them through the AI interface and it comes back with a response.
So once that response is submitted and it's posted on X. It takes the original interaction and its response and any subsequent interactions and admits them on our blockchain. So it is really using the blockchain. So this is a true, like, blockchain based agent, unlike other agents that claim to be, you know, on chain.
But the only on chain component is they have a meme coin. Aside
Stephen: Ha
Todd: there's nothing to do with the blockchain. But again, this is a proof of concept. It's not to say that there's tremendous value in storing interactions on X, but this is just to show, you know, when our agent talks to another agent or talks to a person and a dialogue gets started, those communications can be anything, but we're taking those and we're minting them on chain so that there's a history and you can go and retrieve it and it's there permanently.
Stephen: I think that, you know, when you say, does that have a use case? Like, when you think about, you know, what's happened recently with Puff Daddy or, you know, Jay Z. The first thing everyone does is go and delete their tweets. To your point, so like if there's a place where you can like drum up some of these tweets and obviously there's like screen recording and you know, a lot of people, there's probably companies that all they do is take pictures of celebrity , you know, screenshots of celebrity tweets.
But it'd be interesting if, you know, David Sachs had such a contentious point, you know, seven years ago about crypto. Like he was like, Jamie Diamond's, right? You know, Bitcoin's a scam. And then now he's kind of running the opposition would wanna pull up those old tweets and how he responded. To a bot, pretty much an AI agent.
I think that's super interesting and fascinating, actually.
Todd: Yeah. And then what our model will do is when it mints it on chain, that becomes part of the on chain memory for the rag model. So it uses those interactions to develop any kind of future responses that it's going to generate. And that memory gets built and it gets retained permanently. So, you know, again, this isn't to say, you know, one of our core product offerings is going to be we're archiving You know, tweets, a lot of people do that, but in this case, it's tweets, but it could be supply chain data.
It could be medical history. It could be education material, and this could be happening between agents or between the human and agents, and it's going on chain. We just felt that this was a fun little way to kind of demonstrate you know, how our product. Works with the AI agent using our auto drive SDK, which was what stores the data and then putting the AI component in and showing people how the whole thing kind of works together in a very understandable way that we think people will find fun.
Stephen: awesome. You know, before I get into asking about your predictions and what the future holds for what you are doing and others in the space in 2025, no, some of the people are listening to this AI agents are able to respond. Where's the, you know, the business use case, like where's the, you know, if I'm an average user, I know a little bit about crypto, I'm dabbling in AI.
You're talking to these people that are building these agents, like who's whispering to you, like, Hey, you know, all these people have to do is ABC. And, you know, they could probably do 50, 60, 000 a year, a hundred thousand dollars a year, rather easily, especially if they don't like their job. Is there anywhere where you're like, Hey, these are just areas.
And I know we don't offer financial advice, but these are areas that people might want to examine when it comes to AI whether it be a side hustle or even just a fun weekend project with, I'm sure what argument probably started. Yeah.
Todd: Right. Yeah. And one of the things that I kind of keep in mind when I'm talking to people, especially some, some of the younger generation today is to make sure if at all possible universities now are starting to offer AI curriculum. That's going to be hugely important. Certain Functions, repetitive functions.
Tech support is one where we're seeing it on Amazon. Now you chat with an AI, you no longer have to chat with a live person. So things like support you know, more trivial tasks are going to be handled by AI's and those jobs are going to go away. But the people that know how to use AI. Are going to be the ones who really have flourishing careers in the future.
The ones that know how to maximize its benefit to get the most productivity. It's not necessarily to say that in the future, AI is going to reduce jobs and we're just going to continue with the output that we have. It's rather going to make people more productive so we can. You know, create more GDP is probably going to skyrocket in the coming years as a result of, you know, the AI innovation and how that whole space is going to move things forward.
I would say to the point about how that is going to interact with blockchain is that. You know, keeping everything in these, these permanent records, as we talked about, like with storage for payment rails. And, you know, that would be one use case as you'd mentioned, payment rails and transparency of data online and entire AI models, perhaps at some point, and things like inference and federated learning, where the models are constantly being trained on new information, people want to know what that is.
Well, federated learning model can be. Put on chain inference, you know, training the model on new information and making sure that it stays up to date and is most current. So that's something that we're seeing now from some of the AI providers, the ability to do things like inference and some federated learning tasks.
So I think those are going to be really big spaces to watch. Another area that is probably going to be affected is unfortunately going to be software engineers to a large extent. I don't know if it's going to put people out of business, but everyone on my team that I have spoken to said that AI has made them You know, five to 10 X more productive.
Nobody writes code anymore. You're just using the tools to piece it all together. So there will still be a lot of value in knowing how to do that. And that's why I would encourage people that are just getting into the space that are maybe younger to really make sure that you seek out these programs and educate yourselves.
Because if you have this background, you're going to have a lot more value in the workforce than somebody who either doesn't know how to use an LLM and interact, or even someone that just uses it recreationally.
Stephen: Yeah, I agree. Like even just some of the agents, I see people building out like simple stuff, like whether to do with LinkedIn or sales. I'm like, well, if that just person just cut down 25 percent of their workload, they're going to be able to deploy that somewhere else. And you're probably going to see a lot of one, two person teams, whereas to your point, it used to take 10 developers, a lot of money, and you're spending 90 percent of your day trying to find a developer, test the developer.
But if you don't even know what a good developer looks like. Then you're kind of almost wasting your money. And now, as you said, if people get into this and get a better understanding, it'll be super easy for them to deploy on their own or just hire somebody to do it for them. All right. We're coming up on 2025.
We'll be in 2025 by the time this episode is released. What are your thoughts? What should we be looking out for? What are you excited about? Whether it's with Autonomys or just the industry in general for 25.
Todd: Yeah. Well first with that, I would say happy new year to everyone who's watching. 2025 is going to be really interesting from the standpoint of just Autonomys. We're going to launch our mainnet phase two, our execution environment and we will have a tradable transferable token. So for us, that's going to be the culmination of, you know, years of.
Of effort. Again, our consensus chain is already live. We're not just a white paper. Like so many projects, we've got something real, something that's tangible. So the evolution of that will be exciting in 2025 also will be exciting is obviously the anticipating change of what we had said before the regulatory landscape.
Clearly there's going to be a softening of regulatory policy when the new president takes office. And I think we're all anxious to see what that's going to look like. I don't think that the. Regulatory bodies are just going to lay down and say, you know, we'll just open up the floodgates and now it's the wild west of crypto.
But I do think that there will be some thoughtful processes and policies put in place about maybe when you want to do an ICO or you want to launch a token, here's a registration project that makes sense for you. You don't have to have audited financial statements. Because obviously these are not the types of things that are conducive to.
To crypto big difference between it and orange groves, which is what the Howie test was, was founded on. So regulation would be another big one to see what happens in the ETS base, I'm very interested to see you know, Bitcoin, if you've been watching along with. Bitcoin and Ethereum, the inflows have just been breaking records every week.
It just keeps going and going and going. And now we're seeing Bitcoin over a hundred thousand. Some people are predicting that to continue to climb dramatically as ETF adoption and institutions continue to come in and just invest in the space. I think that will be interesting. Altcoin ETFs. Maybe I don't know if we'll get to them in 2025.
Stephen: I saw the Solana ETF, they submitted that. So I think if you get into a Solana ETF, then I think anything's fair game.
Todd: maybe, but you know, the thing that's missing from Bitcoin and Ethereum have regulated futures markets that trade on the CME. So they already have regulated exchanges. These are mature ecosystems, at least more mature ecosystems. So to take an ecosystem that has. No proven financial market in existence at all, aside from, you know, the crypto exchanges that it trades on.
And for the SEC to not have any kind of surveillance mechanism over things like price manipulation you know, concentrations of Ownership by certain individuals. I think it's going to be difficult for them to just go in and kind of blanket approach and say, yes, anything can go into an ETF. I think there's going to be a lot of thought that's put into that.
Maybe we get a roadmap for what that's going to look like, but I'm not sure if we'll get it in by the end of next year, but that would be great if we can.
Stephen: I love it. Todd, I'm, I'm, you know, recently I've been asking people, you're around a bunch of developers, a bunch of tech person, people, you're, you know, you started off in tech yourself, what are what I call the nerds? What are the nerds playing with on the weekend that we should be like, Hey, you know, maybe we should be looking into this.
AI is one, a lot of things, but what are people doing on the weekends or with their time that you're like, Oh, this is strange, but I can see society kind of going down that route in two to three years.
Todd: Yeah, those are great questions. I would say on our side it's a lot of development and rust. And what I have found with our developers is that they just love to build these things in their spare time. Right now I really do think that AI, not just because it's the, like the shiny new toy but because it is kind of like the foundation of a lot of things that are being built right now.
Is still playing a big part of what people are doing. The argument was built on the spare time of our head of engineering, who was able to just kind of create this and keep working on it on weekends because he liked the idea and he wanted to do it because this is what he enjoys doing. So that's one thing.
The RAG model that we're using in it to go out and, you know, do the the AI component and come up with an answer and then mince it on chain. That whole thing in the auto drive concept that we're using to put that on chain was done by another one. Another one of our developers who did it in his spare time because he was working on it.
So at least I can say with our team, the The things that they're playing with mostly revolve around our project, which is great. It shows the, the passion and the loyalty that people have to the project. And I would guess that similarly things like that are going on at other projects, but AI agents are currently all the buzz and that seems to be what everyone is focused on.
So I would certainly say that's something to keep a sharp eye on.
Stephen: I love it. And where can people get in touch? This has been a fascinating episode. Where can people get in touch with you? You spend most of your time arguing with argument on Twitter or are you on LinkedIn as well?
Todd: Yeah. So We originally were going to be a polka dot parachain, which we scrapped the idea years ago and decided to be a sovereign layer one. So, on Twitter, I am PolkaTodd so you can get me on Twitter. Our website is Autonomys with a Y, Auto N O M Y S. Of course, we have a very active discord server.
We do have an Autonomys Twitter account. You can, of course, follow zero X argument as well. And telegram all the major channels. We would love to welcome anyone who wants to hear more about the project.
Stephen: Cause you brought it up again. What, and now you're explaining it, you know, why do a layer one, you know, that's the question I have for anyone building a layer one, where you probably saw early on, like, Hey, we can just build on top of an ecosystem that's already built thriving, has a community. What was the kind of maybe not nailing the coffin, but where you're like, Hey, we're going to put in all this work.
We might as well just do our own layer one network.
Todd: Yeah, because there is the reliance on someone else's technology to for yours to exist. And not that there's anything wrong with that at all, especially in, in the case of Polkadot, you know, they've built the whole parachain ecosystem but everything is relying around their consensus chain underneath.
And if that stops, which it hasn't to the best of my knowledge, if anything were to happen. All the projects that are on it are also gonna kind of grind to a, to a halt or gonna run into trouble. You know, we've seen that happen with Solana several times now where the changes kind of stops. So by building your own sovereign layer one, not only can we just take advantage of.
Our core competencies without having to like kind of put a rump peg into a square hole, which is what it felt like we were doing. But it allows you to really expand your creativity and you know, our founders just when they came up with this idea it was so novel and so kind of new that no one had ever really done it.
So it seemed like the easiest thing would maybe be to build it on another chain, but. Once we realized, I think the creative and engineering creativity that could be accomplished as a sovereign layer one, we decided to go the other route. And that happened about three and a half years ago. And that's how it evolved.
Stephen: and that makes a lot of sense because if nobody ever did it, you want to kind of have that moat there, that advantage. But if you build it on top of something else and anyone can just kind of come and build a very similar foundation as you whereas it looks like you guys are doing things completely a different way.
That's only going to happen on your own blockchain.
Todd: Yeah. The only thing you really get from building on someone else is you get this. The chain security, right? If you build on Ethereum, you get the security of Ethereum. Same with Polkadot or anything else. But you know, as I said, we have over 400 petabytes already. Hopefully a lot more by the time people see this that have been put toward our consensus.
And you know, our chain security, we know is rock solid because of, again, I mentioned our dilithium consensus that we built that is one thing that we're really comfortable that we nailed. So yeah, we'll have our chain security for sure.
Stephen: I love it. Todd, thank you so much. And we appreciate you coming back on. Maybe we can have you back in the end of 2025 to see maybe some of the predictions and see how mainnet the phases went for you.
Todd: That'd be great. Thank you for having me, Stephen.
Stephen: Awesome.