Lumerin AMA Recap with Ryan Condron, Co-Founder of Lumerin

Dear Lumerians,

On June 27, we hosted a Twitter Space AMA with our Co-Founder Ryan Condron, our Engineer Scott Terry, and our Head of Marketing Alexa.

We’re extremely grateful to everyone who tuned in, participated, and shared their opinions and questions. However, if you couldn’t make, don’t worry!

Here’s a short recap so you don’t miss anything.

How do Lumerin and Morpheus coexist? What is the symbiotic relationship between the two protocols?

Lumerin’s first project was building a decentralized marketplace for buying and selling hashrate called the Lumerin Hashpower Marketplace.

We’re taking the same approach with a decentralized AI marketplace. We’re still exploring various options for combining them to maximize the value of both marketplaces.

Note: Morpheus is building a decentralized AI protocol where users can access AI models on-chain. Lumerin is contributing to the development and infrastructure of the Morpheus Network.

Can we dive into the energy redirected towards Morpheus and Venice instead of Lumerin? Wouldn’t it be best to focus on yourself and grow that marketplace?

The Lumerin protocol, the routing mechanism, is used on both the hashpower and AI marketplaces.

The difference is that the Morpheus token is used on the Morpheus marketplace, not the Lumerin token. We had to devise a strategy for bringing value to the Lumerin token through Morpheus. Part of that strategy is the Morpheus community earning more tokens through coding weights.

The Lumerin community is currently the third-largest holder of the Morpheus token. Our community owns a significant number of MOR.

We’re building a staking mechanism for using that token to airdrop our community, provide AI computing to our community, and actually provide liquidity to the Lumerin token itself.

That strategy will be implemented starting in July, but it’s been a very beneficial partnership.

Q: Can you elaborate more on providing AI computing to your community? What does that entail?

A: The really interesting thing about Morpheus’s architecture is that just by holding more tokens, you are given a certain amount of computing that you can use every day.

You don’t have to spend your Morpheus token to use the compute. Because of this, holding a large number of Morpheus tokens is highly valuable. As time goes on in the AI ecosystem, computing will become one of the most valuable resources on Earth.

Given that we have so much of the Morpheus token, we will actually control a very, very large amount of computing on the Morpheus Network daily without even having to spend the token. How we distribute that to the Lumerin community is still being worked out.

We’re looking at potentially staking the Lumerin token to access that compute tranche on a daily basis in ratio to your stake. But those mechanisms haven’t been developed yet, so we’re still working on them.

Right now, we’re kind of in the baby step phase.

Can you explain how this is going to provide deeper liquidity for the Lumerin token?

The plan is to sell a portion of our Morpheus tokens to build our own liquidity pool.

Obviously, we don’t want to sell all the Morpheus tokens because that’s not fair to the Morpheus community. We don’t want to dump the token. We are keen to be very good long-term partners with the Morpheus community.

However, we have a small portion of the Morpheus token earmarked to sell to build both sides of the liquidity pool. So, we can start building our own protocol-owned liquidity pools, which will help provide depth in the order book. I know there’s been a lot of concern about the current depth of the order book and the current liquidity pools.

This is our solution for building that over time. Because we get the Morpheus token dropped to us every day, we can buy up the Lumerin token and provide liquidity in the Uniswap pools on a daily basis. So it’s going to grow over time and become more robust.

Can you explain what’s stopping other large decentralized CPU and GPU marketplace competitors like Akash or Render from offering a similar hash power marketplace on their protocol?

From my perspective, there’s nothing wrong with that. It actually aligns with Lumerin’s goals and vision because the point isn’t to control the hashrate. The point is to decentralize it.

So, while we’re trying to build a decentralized hashrate marketplace to increase the decentralized nature of Bitcoin and mining, if there are other projects doing the same thing, so much the better.

But I do think we have an advantage… It took a lot of work, a lot of brainpower, and a lot of investment.

So, nothing is stopping them. But it really takes a lot of willpower to do it and make it work. The other power we’ve got is that it’s not only decentralized, but it’s open source.

The reality is they could build a hashpower marketplace, but AI computing and hashpower are very, very different.

They could start from scratch, trying to overcome the 10 years of experience that our team already has in the mining field. So, they could try to learn the Stratum Protocol and build granular hashpower routing systems and socket management systems. Or they could fork our open-source code and use the Lumerin Protocol.

That’s precisely what we would want them to do. There’s no point in them recreating the wheel. So, if Akash or one of these other compute providers wanted to offer a hash power marketplace, they could literally fork the Lumerin node and spin one up.

It would make our ecosystem all the more robust because rebuilding the wheel — I mean, it took us two and a half years to build it.

And that was with us knowing what we were doing. So it wasn’t easy.

Will Lumerin or the hashpower marketplace incorporate new proof-of-work protocols, such as KASPA, in the future?

It’s not on the roadmap at the moment.

In fact, the backend is currently built on the assumption that the providers and consumers are using Stratum. I don’t know what KASPA uses, but if it uses Stratum, in theory, an open-source developer or organisation could repurpose it for that or add that capability.

I say that because even if a different proof-of-work mechanism uses Stratum, there tend to be idiosyncrasies. There tend to be quirks in any slightly different system. Even within miners, we found we had to adjust for quirks.

Long story short, it’s certainly possible. If you use a Stratum, it would be relatively easy to set that up. It’s really just a question of will, who’s going to take that, and what the priorities are for the people involved.

It’s not currently on our roadmap, but if that’s where the community wants to put effort, that will be developed next.

How much focus is currently placed on the Morpheus node? Can you speak about when the node will be up and running and what the future holds for the Morpheus project? Also, elaborate more on how Lumerin will continue contributing to Morpheus and how these two ecosystems will live symbiotically.

We brought on a community manager specifically because we would like the community to take the hashpower side of the marketplace and run with it as much as possible.

We want to provide incentives for that because, ultimately, Lumerin is not about a particular marketplace.

It’s about decentralizing marketplaces across the board and creating a secure decentralized protocol for doing so because there’s a lot of need for this in various industries. So that has always been our focus in a lot of ways. I alluded to this a bit before.

The hashrate marketplace was a proof of concept. At least any front ends or integrations we build for Morpheus will be like a proof of concept. We’re really trying to build pipes to make it possible.

So, at the moment, I would say we’re really close to having something ready for the test net. I’m going to say really close vaguely for now, but we want to make sure our initial users and partners, integration partners, are able to succeed.

So we are really focused on that, on making sure that the people who are going to be building and maintaining the front end and the integrations more long-term, we want to make sure we’re serving them really well. So we’re really focused on the quality we’re providing for those integrations and some proof of concept integrations ourselves.

How are Lumerin and Morpheus ensuring that user data security and user private information are protected in a decentralized network?

I know that decentralized networks have some positive aspects, such as protecting data and stopping it from being exploited at a centralized point.

It’s such an interesting problem because decentralized networks and blockchains are about sharing data across nodes, but you don’t want all the data in the system shared with everybody.

So, it’s always a delicate balancing act. The short answer is that we use the transparency of the blockchain on Ethereum by putting the governance code in smart contracts, but we also secure the essential private data for the individual user. So, we use encryption mechanisms, obviously, to avoid having those exposed.

We avoid exposing any of that individually identifiable data across peers. If someone is using their own individual private data in the system for a use case, like purchasing, that’s never going to be exposed to any other peer in the system.

What practical problems are Lumerin and Morpheus aiming to solve using AI?

It’s funny because the team has different attitudes towards AI tools. I’m an AI maxi, so I love to use AI tools as much as possible to optimize my productivity.

I mean, we’ve experimented with AI chatbots for automated code review. We’ve at least looked into tools that automatically provide security update recommendations and even code pull requests. There’s only so much you can do when you’re trying to push a product forward.

We can’t spend all of our time playing with shiny new toys, And it’s really amazing how quickly new tools leveraging AI are getting produced, and new models are being produced at a ridiculous rate in the open-source community. But to give you a flavor, some of us use AI tools to help speed up the process of writing coded tests.

The other half of the question almost sounds like you’re asking: What is Morpheus’s point? What are we trying to accomplish in the AI world?

I have an excellent example of that. So, an article came out talking about an AI agent that was built, I believe, using the TAL network. They had built an agent and gave it about $10,000 to invest in it.

Within a week, it returned about a 24% return on investment. It made about 84 trades, like an auto trading bot. One of the visions we’re looking at is that we can actually build AI agents that can consume a very large amount of data, analyze data, look at the patterns inside that data, and autonomously make financial decisions.

So, having Web3-enabled AI models that run locally on your machine and are trusted to work with your wallet is actually the core vision of the Morpheus smart agent system.

In order to get there, we have to be able to have interconnectivity between a bunch of different chains, a bunch of different models, and a bunch of different datasets, both public and private. As long as the private dataset is obviously exposed to a Morpheus network, it may even be behind a paywall.

So, having all that different data interconnected is what Lumerin specializes in. We are working on a proxy system that connects all the different nodes together, where users and providers can interconnect, and then the AI agents can basically crawl across this spider web of Morpheus nodes. So, the aspect of an autonomous AI agent working on your behalf that you financially enable is a very, very powerful thing.

It will get to the point where any human interaction with the system is going to be considered a failure of user interface and UX because if we do our job correctly with these AI agents, we shouldn’t have to interact with them, and we shouldn’t have to tell them what to do. They should be able to consume the information appropriately and make decisions based on the information, and most times, probably better than we would be able to do it ourselves.

So it’s a scary vision and an incredibly powerful vision all at the same time. Let’s go with incredibly powerful.

Lumerin and Morpheus are combining some of the latest and most innovative technologies to create something unique.

How are you guys planning on attracting new users to your platform and expanding the community? What’s the sales pitch on why folks should join and use Lumerin?

We had an incredible call with some of the core guys at Filecoin.

They see that one of the leading restrictions on AI’s growth and progression is the need for real-time storage and hot storage. This is going to be increasingly a bigger issue as context windows for AI models need to be larger and larger. We need to have better and more robust storage options.

If we’re dealing with a decentralized system, what better to pair that with than decentralized storage?

So, the next natural fit for partnership is with Filecoin to tie into the Morpheus ecosystem. That brings in a massive number of people into this community. We’ve talked to people from The Graph who are interconnecting different nodes, almost like decentralized block explorers.

AI agents will need that information to consume on-chain data. We’ve been talking to many other protocols because, at the end of the day, all this gets tied together into a solid mesh, right? We start with computing, and some of the world’s largest miners are now knocking on our door, saying, hey, we want to be involved. We’re retrofitting a large amount of rack space from Bitcoin mining to AI computing.

And we want to actually back both marketplaces, the hashpower marketplace and the AI compute marketplace.

It’s this vision of these decentralized systems and autonomous agents that all come together almost like puzzle pieces. When you take a step back and look at the ecosystem from a macro level, you start realizing that we’re building something like this incredible tapestry, where all these different systems will be able to talk to each other, depend on each other, and interconnect in a very decentralized permissionless way.

It will all be outside the banking system, outside of Google, Apple, and Open AI. And because it’s decentralized and permissionless, they can’t stop it, control it, or censor it.

So, all the things we’ve been building, from Bitcoin to Litecoin to Dogecoin to Filecoin, are building on top of each other to create this opus of Morpheus.