AI in Blockchain, AI 3.0, AI Agents, Autonomous Agents, and more. feat. Todd Ruoff of Autonomys

zeeve

Explore the future of AI and blockchain with Todd Ruoff of Autonomys as we dive into AI 3.0, autonomous agents, decentralized AI networks, and AI-powered smart contracts. Discover how AI agents are evolving beyond automation, enabling true autonomy, and reshaping industries. Will AI and blockchain merge to create unstoppable decentralized intelligence? Tune in to find out!  🔹 Topics Covered:  AI 3.0 & Next-Gen AI Agents  Autonomous Agents in Web3 & Blockchain  Decentralized AI Networks  The Future of AI-Powered Smart Contracts  Autonomys’ Vision for AI & Blockchain  🚀 Don’t miss this deep dive into the cutting edge of AI and decentralization! Like, subscribe, and hit the bell to stay updated on the latest in tech innovation.  #AI #Blockchain #AutonomousAgents #Web3 #AI3 #DecentralizedAI #Autonomys #TechInnovation

Transcript

00:00So, hi, everyone. My name is Ravi. I'm the co-founder CEO of Zeef. And today we have a very,

00:10very exciting startup. We have Todd, CEO of Autonomous, a very, very exciting startup in

00:17the Web3 space. Not a new name, a seasoned startup. And I would like to welcome Todd.

00:23Hi, Todd.

00:25Hello. Thank you for having me.

00:26So, Todd, first of all, I would like to know about yourself, you know, a brief introduction

00:33of yourself and specifically your journey into Web3. And then, you know, a bit about Autonomous,

00:40how you started and a bit, you know, what you have been building at Autonomous.

00:46Yeah, sure. So, I came into the Web3 space through TradFi, actually. So, I worked on Wall Street for

00:55almost my entire career. My background is originally in technology, but then I got into the trading

01:03aspect, operations, working in the investment advisory and the broker-dealer space. And I

01:09had a long career in New York City. And upon getting tired of the commute, I decided that

01:17I wanted to pursue something else. And since tech was my kind of first love, I decided to

01:23look for something in the tech startup space. And I found crypto to be the perfect kind of marriage

01:29of finance and technology. So, I decided to pursue something in Web3. And I met up with

01:37our founder in a common interest kind of work group online and started working for what was

01:44at the time, Subspace Labs. And Subspace Labs is now Autonomous. And I began here four years ago.

01:52And slowly but surely, kind of, you know, we grew the team. I was employee number three. Now,

01:58we're about 30 people. And over time, the founder kind of took a step back and I moved into the role

02:06of the CEO. So, that's how I came to be here. And that's kind of how I came to be in crypto. So,

02:12it's been very exciting. No, I think that's very interesting. In fact, before getting into

02:17blockchain, I was in fintech space. But not on the trading side, I was more on the lending side.

02:23Okay. And then, you know, I got into blockchain. And in fact, when I started in blockchain, the first

02:29use case that I worked was invoice discounting, again, in the fintech side. Before, you know,

02:36infrastructure picked my interest. It started Zee four years back, focused on, you know, how to

02:42bring in managed and, you know, automation and a lot of security and compliances in the

02:49infrastructure side. And I think very, very interesting. So, I think Autonomous is very

02:55strong on the data storage, how bringing distributed data storage, distributed AI. And I think in last

03:04few years, we are seeing how AI has been disrupting and blockchain being the, I would say, infrastructure

03:10layer powering the AI. So, can you tell us a bit about, you know, how Autonomous is powering the AI

03:18infrastructure with blockchain as the backend infrastructure to it?

03:24Yeah. So, what's unique about us is primarily our consensus mechanism for the L1 is what really

03:30differentiates it. We, instead of being using something like proof of stake or proof of work,

03:37our layer one is built on proofs of archival storage. So, users contribute hard drive space

03:44to the network. And we then plot a portion of their drive, depending on how much storage they

03:51contribute, with the actual encoded blockchain data. So, if they contribute a small amount of space,

03:57they may only get a portion of the blockchain. If they contribute a large amount of space, you know,

04:02several terabytes, they may get multiple copies of the blockchain data. And that encoded plot is what

04:08is used to respond to challenges from the network for a transaction. So, if you think about something

04:16like Bitcoin with proof of work, a challenge gets posed to the network and all of the CPUs in the world and

04:23all the ASICs start trying to solve the challenge and whoever wins, you know, gets the block and then

04:28a lot of the electricity that's just been committed is wasted. So, in proofs of archival storage,

04:34what happens is when the challenge is posed to the network, the user's computers with very,

04:39very low energy use scan their hard drive plots and they're looking for the solution to the challenge.

04:44And the one who finds the challenge wins the block reward. And also, the 10 or so that came closest to

04:53correctly solving the challenge get voting rewards. So, it's incentive compatible. Again, it's a very

04:59low power consumption. And it gives us a way to store the blockchain data and use it effectively. So,

05:07this is proofs of useful storage. The storage is actually what's being used to respond to consensus

05:14challenges. So, the blockchain data itself is what's really, you know, what's powering the network.

05:20And this makes it very conducive to blockchain bloat. The larger the network grows, the more data that's

05:27on chain, the more nodes expand. And currently, we have over 2,000 nodes participating in consensus,

05:33making it very decentralized. And as the blockchain continues to grow, users don't have the burden on

05:40their nodes where they have to store, you know, the entire chain and it gets too large. And then,

05:46you know, chains like Ethereum have to start thinking about pruning data. For us, that bloat

05:53actually becomes an asset because that means the network is getting larger, the nodes are getting

05:57more decentralized, they're getting more secure, and it can just grow and grow and grow. So,

06:02it's very conducive, as we said, to storing large amounts of data on the on-chain itself.

06:09That's, I think, awesome. So, bringing in cost efficiency, and at the same time, you know,

06:15decentralization, and at the same time, bringing security to the data. And what about, you know,

06:21how the data is being managed, how the privacy and security are being managed in this kind of

06:27distributed environment? So, the data is erasure-coded across the nodes. So, it is just kind

06:34of portions of each, you know, block are being stored across various nodes. So, while the data,

06:42if someone contributes a large amount of storage, you know, may all be contained on a single node,

06:48there's just fractions of it. And it's spread across various nodes. The chain is censorship-resistant,

06:55it is permissionless. So, anyone can run a node, anyone can store data on the network,

07:00using our token as the utility token to pay for the storage costs, just to put data on-chain.

07:07And it's incentive-compatible, because as users continue to participate in consensus,

07:13they're earning rewards for, you know, solving additional block challenges, for contributing

07:18additional voting rewards. That is at the data layer, which is on our consensus chain. And what we are

07:24going to be launching coming up is our compute layer, which are these, what we're calling

07:29enshrined domains. These, essentially, think of them as roll-ups that sit on top of the consensus

07:35chain. We will have one of those for ID, we'll have an EVM, and anyone will be able to stand up a domain

07:42to take advantage of the compute layer, which sits on top of the data storage. And obviously, those two

07:46things work very well together. Got it. So, Autonomous is, the core value proposition

07:54is the foundation layer for AI 3.0. So, what exactly is AI 3.0?

08:02AI 3.0 is what we look at as the evolution of where AI is going. If you think of AI 1, we would just

08:09consider that basic machine learning that's been, you know, taking place since basically the turn of the

08:15century. And then we get into AI 2, which we would consider where we are now. This is generative AI,

08:22you know, basic input from users getting responses tailored to specific queries that are posed to

08:29LLMs. AI 3.0, we consider to be the next iteration of this, which is where there will be ownership and

08:37privacy and identity that goes along with all of this, so that users can actually have transparency

08:44into AI models, seeing how they were trained, ensuring that there's no bias to, you know, owning

08:51their data that comes from responses, things that are stored on chain. And we can talk a little bit

08:59about what we're doing with agents, which is where we really see the value in this. There's some agents

09:04that we've already deployed. And we feel that this is a really kind of the natural way for AIs to be

09:13able to store their data going into the future to ensure that there is transparency and accountability

09:19and immutability so that things cannot be changed looking back.

09:24Got it. No, I think that's very exciting. In fact, I was very intrigued by story protocol. And I think

09:33this is very much required, the whole concept of open, accessible AI, because we know there are a lot

09:42of challenges in the current version of AI, especially the data provenance, data monetization.

09:48We don't know how the data is being used and whether we can trust the underlying data, which is being

09:54used to train the models. And I think having a proper data provenance and creating monetization

10:01models where the right set of data copyrights and et cetera can be taken care of, I think is super,

10:07super important. And we know the disruption is happening at massive speed, much, much more than

10:13what we were anticipating. Going through one video, which was talking about deep research,

10:18which is recently launched by ChatGPT or OpenAI is mind boggling, the kind of speed which is there

10:26of innovation. So I think having the right set of infrastructure is super important to match the

10:32speed of these innovations. Otherwise, I think we may land up into severe issues. So I think this is

10:41quite interesting and I would say it's super important and at the right time. I also saw that

10:49there are a few other products, especially auto ID, one of the identity side. And I think identity,

10:55as I can understand being a foundational layer, more on the infra side, self-sovereign identity

11:03becomes very important. So a brief about what auto ID is and how it plugs in into the overall design

11:10of your foundational layer. Auto ID will run as a domain on top of the consensus chain. So as I

11:17mentioned before, we have these domains, these enshrined roll ups, which will sit on top of,

11:22you know, the layer one on consensus and auto ID will be one of them. I mentioned before agents,

11:29and that's a space that we are, you know, laser focused on right now. And when we think about agents,

11:35everyone has been talking about autonomous agents and, you know, crypto agents. And most of these

11:42agents are little more than a meme coin that maybe can send some transactions around on chain.

11:49But there's nothing really autonomous or web based about them, or sorry, or web three based. There's

11:55nothing they have to do with crypto aside from the meme coin. When we think about autonomous agents,

12:00there's really three things that they need to have in order to be truly autonomous. They need a place to

12:06live. They need money and they need identity. And an autonomous agent cannot go and sign up for an AWS

12:16account. It just can't do it. There's some minimal amount of KYC that's needed. There's some type of

12:21paperwork. There's things that are going to have to take place. So an autonomous agent can't be autonomous

12:26and set up an AWS account as a place to live. It can't open a bank account. So an autonomous agent

12:33go to JPMorgan Chase and open up an account and deposit funds and be able to use that, you know,

12:39to pay for services. It also doesn't have a passport or a driver's license, which is where auto ID comes

12:44in. So we're supplying those three things for agents and that the first place is a place to live.

12:50Our chain is permissionless. An autonomous agent can spin up an agent on our blockchain. Now it has a

12:56place to live. It doesn't need any of the permissions that it would need to be on something like AWS.

13:02It has tokens, which are of course very conducive to payments. So it doesn't need the bank account.

13:09And then finally, it needs a way to identify itself. So the auto ID domain is going to provide agents and

13:17human entities with ways to verify who they are really for the agency aspect so that they can

13:24cryptographically prove the identity of the agent or the individual for performing transactions.

13:31So with those three things, with a place to live on the blockchain, with money in the form of tokens,

13:39and then an auto ID, a truly autonomous agent can exist and it can spawn up additional agents and it

13:46will be able to then identify itself to the rest of the Web3 ecosystem so that you can be sure that you're

13:53dealing with who you think you're dealing with. Because as we know, you know, Web3 has been rife

13:59with scams on multiple levels. And when we're talking about autonomous agents, you can just imagine

14:04this problem being multiplied times millions or billions. So cryptographic proof of identity is going

14:11to be critical as we go forward. No, absolutely. I absolutely agree with you. I think, again,

14:19on the infrastructure side, if you want to have, you know, as you rightly said, more meaningful agents to

14:24be there. This is what you are building is more on the infrastructure side, you know, having the right

14:30identity and the capability for these agents to identify and define themselves so that they can do the

14:37right set of actions. And those actions will have, because as the AI agents will have more meaningful

14:42activity, there would be financial transactions. That means there would be legality to it and there

14:47would be a lot of challenges to it. And until and unless the necessary underlying infrastructure is

14:51there, all those challenges cannot be navigated easily. I think, so Todd, I think we can now talk

15:01about we are going to meet at ETH Denver in two weeks time. So what are your plans? Any new exciting

15:10developments happening at Autonomous that you would like to share in the next month or so?

15:16Yeah, well, we will continue to roll out additional agents. We already have several that are live as

15:23proofs of concept. They're currently running on the X platform because that's where all the conversation is

15:28taking place. We have one called ArgueMint with an M-I-N-T. So a zero X argument that we have been

15:37really kind of promoting and showcasing. And this is not to say that there is necessarily value in

15:43immortalizing, you know, tweets on the X platform. This is to show that agent to agent communications can

15:50be stored permanently and transparently. And we're taking all the interactions as well as the reasoning

15:55process that an agent takes on the X platform to formulate responses. And the entire process is

16:02being minted on Autonomous. So you can go to the Explorer and you would be able to look and see

16:08every single interaction that one of our agents has taken on the X platform in interacting with someone,

16:15you know, in the form of a tweet. And of course, this will be very useful going forward when we have

16:20agent to agent communications that take place on X or on other social platforms or on other protocols

16:26that aren't even in use yet. So we've been showcasing this. We'll continue to be talking a lot about that

16:32as we go into ETH Denver. We will be talking about our upcoming TGE for the execution level domains,

16:40which we'll be launching in the spring. We're very excited about that. We are also considering exploring

16:46additional fundraising opportunities. So we're looking forward to meeting up with investors and,

16:52you know, others in our industry just to kind of see what everyone else is doing. But, you know,

16:57we really feel that we are the only ones who are doing on-chain agents that are actually on-chain,

17:03not just meme tokens. And ours doesn't have a meme token. We have the layer one token,

17:09which is used to, again, you know, put transactions on-chain and to put data on the network. But

17:17there's no meme tokens. There's no just sending money around for, you know, no particular reason.

17:22It is to showcase the data storage aspect and the infrastructure. So those are the key points

17:27that we're going to be driving home. And we look forward to having some good conversations about

17:31this and, of course, meeting up with you. We have an event that we'll be co-hosting,

17:35so looking very forward to meeting you in person. No, absolutely. I think we are very,

17:41very excited to partner with Autonomous. And Zeev is very excited to host this event and meet so many

17:48excited, exciting protocol partners, startups. And I would like to wish you all the best for your

17:56upcoming TGE. A lot of exciting releases that are coming up on the platform and a lot of new

18:02partnerships and a lot of exciting opportunities that are going to be there this year. With that,

18:08thanks a lot, Todd, for joining us on Web3 Voice. And let's see you soon by the end of February.

18:18Thank you for having me. Thank you. Thanks, Todd.

Category

Transcript

Recommended