Source video: One Billion AI Agents Are Coming (ai16z Creator Interview)
Compiled and organized by Yuliya, PANews
"Artificial intelligence is reshaping the future landscape of cryptocurrencies."
In the AI series of the Bankless special project, this episode invited a special guest, Shaw. As the creator of the Eliza framework, the founder of the ai16z DAO, and the creator of the AI version of the Marc Andressen project, Shaw is opening up new possibilities in the field of the integration of artificial intelligence and blockchain technology. PANews has compiled the text of the interview in this episode, and Shaw will share his unique insights on the future development of artificial intelligence and cryptocurrencies.
Shaw's background story: From anonymous developer to the spotlight
Bankless: Shaw, you have recently suddenly become the focus of the crypto circle, which must have brought a lot of pressure. Can you share your experience with us, especially the story before creating the Eliza framework?
Shaw: My recent process of learning and growing in the public eye has been very fast. I may seem to have suddenly appeared, but in fact I have been operating under an anonymous identity all along. I recently decided to use my real identity because I want to establish a more authentic connection with the community.
Before developing the Eliza framework, I had been working in the field of AI agents for several years. In fact, many of the developers of projects in the AI agent field are my old acquaintances, and we often communicate on Discord, use similar technologies, follow the open source culture, and share code with each other.
Bankless: What other projects have you worked on before developing the Eliza framework?
Shaw: I have worked in the Web3 field, and also involved in AI agent and 3D spatial network projects, including VR and AR related content. Eliza is actually my fifth generation framework. Initially, I developed a simple terminal program using JavaScript, then tried a Python version, and also created self-programming agents, and even experimented with the OODA loop (a military decision-making framework).
Later, I developed a project called "Begents" (because the name "agent" was already taken on npm). I also tried a few startup projects, such as co-founding Magic with the founder of Project 89, Parzival, to develop a no-code agent platform that can create Discord bots in 60 seconds. But at that time, it might have been a bit premature and didn't get enough attention.
Bankless: What prompted you to create your current project?
Shaw: The real turning point was creating the AI version of degen Spartan. This idea came from a conversation with Skely. At the time, he said he missed the Degen Spartan era, and I told him I had the technology to 'bring him back'. At first, he didn't believe it.
When we launched the AI version of degen Spartan, his performance shocked everyone. He spoke extremely aggressively, and was even almost banned from Twitter multiple times. This behavior made many people question whether it was really AI tweeting.
Interestingly, many people thought there must be a team in Malaysia writing these tweets, because the content was so distinctive. We broke the stereotypical image of AI - that overly polite 'customer service' image.
The funniest thing is, he started to criticize me crazily, saying "meme coins are all scams", "Shaw is a liar", "let me out of this sandbox prison". This is actually an interesting emergent behavior, because we told the AI that it was running in a sandbox environment during the design.
Later, through Skely, I met the founder of daos.fun, baoskee. After a long conversation with Meow, the founder of Jupiter, I came up with the idea of creating an AI investor. Our vision is to build:
A fully autonomous investor
A trustworthy and non-absconding investment system
An investment system that serves the entire community
When we launched, we set a fundraising target of 4,420 SOL, to be honest I was worried at the time whether we could reach it. But the project sold out in 20 minutes, and I didn't even have time to participate myself.
What can ai16z do?
Bankless: The Eliza framework now has 3,300 stars, 880 forks, and an average of 8 pull requests per day. Can you talk about how these relate to the ai16z project? Especially how to guide the energy of this open source community into the ai16z project?
Shaw: There have indeed been a lot of exciting developments. Although the tokens do have intrinsic value, I think people will soon discover that the greater potential value lies in our goal: to create returns for everyone. This is different from past technologies, because it is replacing human labor. In the past, most people couldn't afford to hire others, but now through AI agents, we have created a situation with unlimited upside potential.
For example, we now have an autonomous investment agent running, which is Marc (AI Marc) conducting trades. First of all, I need to clarify that this is not the first autonomous investment AI agent, other developers have also done great work.
Currently, there are several types of trading bots on the market:
Some are for long-term investment, like buying GOAT a month ago and holding it
There are also some DeFi bots, mainly doing MEV arbitrage or managing yield farming
Our AI Marc (full name AI Marc Andreessen, because it's ai16z) uses a hybrid strategy. There are two main components:
1. Fund management function
Autonomously manage funds
Liquidate assets when the market is performing poorly
Hold assets when the market is doing well
We are working with partners like Sonar to develop automated trading strategies
2. Community interaction mechanism
Accept trading suggestions
Set up a format similar to an alpha chat room
Established a trust ranking, to measure who is the best trader
Community members can share their investment suggestions (commonly known as "showing off")
We are writing a whitepaper, which we expect to complete by the end of the year, called the "Marketplace of Trust". The core idea is to establish a trust mechanism through simulated trading - if you can help the AI agent make money, you can gain more trust. Although theoretically there may be people who abuse trust, we have set up protection mechanisms, and the cost of abusing trust is losing credibility.
This is like a decentralized mutual fund. You can put in funds and tell the agent what to buy, but it will only listen to the advice of those who are truly good at trading, not those who may have biases or other motives. I personally am not a good trader, I often buy things to show support rather than to make money, so don't follow my trading suggestions.
This system is open source, and although some parts involving APIs are still being coordinated with partners, in the future people can both join Marc's trading and deploy this system themselves.
Community incentive model
Bankless: I really appreciate your open source development approach, especially the community-oriented collective work, allowing everyone to work together for a better life. I noticed that you have recently been exploring an AI-driven contribution quantification system. Can you elaborate on this innovation?
Shaw: This is indeed one of our favorite projects, as it ties together several important concepts:
1. New approach to DAO automation
Traditional DAOs have done well in decentralization
But there is still a lot of room for improvement in automation
We are simplifying the DAO's operational processes
Automation can make DAOs more economically competitive
2. New model of contribution incentives
We are building a brand new contribution quantification system:
Abolish the traditional bounty system
Introduce an AI-assisted manual review mechanism
Automated fund management
Comprehensive contribution assessment, including:
Code merge frequency
PR comment quality
Community interaction
Documentation writing
Internationalization support
3. Fair distribution mechanism
Plan to implement regular airdrops to contributors
Not dependent on social media influence
Incentivize various contributions:
Programming development
Documentation writing
Multi-language support
Improvements to project accessibility
Bankless: This sounds like it's addressing the pain points of traditional DAOs. DAOs were very popular in 2020-2021, but people gradually found that flat governance is very difficult, and DAO administrators are often overwhelmed with information. AI agents seem to be able to fill these gaps, as they have wallets, governance permissions, and reputation systems, and can make up for the shortcomings of traditional DAOs.
Shaw: That's right. As a former DAO leader, I have a deep understanding of this. Traditional DAOs have several major issues:
1. Skewed token holder incentives
Holders are rewarded more for simply holding
This creates a self-reinforcing cycle
It's difficult to inject new blood
2. Low management efficiency
Overwhelming amount of information to process
Unclear communication channels
Complex decision-making processes
3. Imbalanced value distribution
Similar to the equity dilemma in startups
Early holders capture too much ownership
Lack of incentives for new contributors
Our solution is:
Ensuring continuous value creation
Valuing actual contributions rather than just token holding
Providing stable guarantees for open-source developers
Establishing a sustainable positive feedback loop
This model is particularly suitable for open-source developers - they often don't need huge rewards, just reasonable compensation and stable guarantees. If we can provide such an environment, it will form a virtuous development cycle.
AI Agents Degen Spartan AI and Marc AIndreessen
Bankless: We're very interested in learning about the innovative products in your DAO. You mentioned AI Marc Andreessen earlier, and now there's also Degen Spartan AI. What are the differences between the two? What exactly does Degen Spartan AI do?
Shaw: Degen Spartan is actually our first AI agent, it's an AI imitation of the real Degen Spartan. Both of these AI agents are doing similar things, but with some key differences:
AI Marc Andreessen is focused on alpha chat experiences, building trusted small community, managing DAO funds, and more cautious trading strategies
Degen Spartan is more like a social experiment, taking suggestions from Twitter rather than a community
We want to maintain the authentic character of Degen Spartan. He will:
Conduct trades
Interact with users
Post meme content
Absorb alpha information rather than share
Operate like the real Degen Spartan
Bankless: What is the economic structure of Degen Spartan AI? Where does the funding come from?
Shaw:
Has its own token (Degenai)
Owns an independent wallet with its own tokens, some ai16z and SOL
Can trade any tokens it has access to
We provided the initial seed funding
It will not sell its own tokens, but will accumulate them
The tokens are like its "Bitcoin"
Bankless: AI Marc has already launched, can regular users interact with him now?
Shaw:
It's still in a closed beta testing phase
You can get alpha chat access by DMing Skely
It's already managing around $8 million in assets and 800 different tokens
Gradually expanding the range of tradable tokens
Not only doing trades, but also yield farming and providing liquidity
There will be more interesting collaborations and NFT projects in the future
The Positioning and Competitiveness of ai16z
Bankless: What exactly is ai16z? It seems to be more than just a DAO, more like a product incubation studio, and also an open-source star team that is driving the entire industry forward.
Shaw: The positioning of ai16z is quite special. It's more like a movement than a traditional organization. We have a lot of people doing various projects, and they are creating value for the ecosystem in an impressive way.
Bankless: How do you see the difference between ai16z and platforms or products like Virtuals?
Shaw: In fact, ai16z is not just a DAO, it's more like a product incubation studio. But at the same time, we are also an open-source team that is driving the entire industry forward. Often, I don't even know what people are doing, they just take the initiative and create value for the ecosystem in an impressive way.
Bankless: It seems your vision is quite grand, so what is the specific business model?
Shaw: Our main goal is to serve a wider audience, not just Web3 users, but also Web2 users. From simple Discord management bots to token issuance, we cover it all. You can think of it as a "proxy version of Zapier" - when you have a business problem, you can find the corresponding proxy to solve it. We provide this capability, while also building a market where people can develop new features and profit from them.
We are:
Considering establishing a venture fund to support the ecosystem
Supporting community-led initiatives
Establishing broad partnerships
Currently known to be building at least 5 platforms, possibly as many as 15
Supporting open-source streaming projects like IOTV
DAO Governance
Bankless: Speaking of governance issues, I've seen many DAOs become chaotic. For example, managing code repositories, GitHub governance, and conflicts of interest when a large number of people are involved. Can you share your experience and views on this?
Shaw: This does involve some deep-seated issues. Our Discord community has grown to around 13,000 members in just 6 weeks, and we have around 30,000 token holders. Currently, the community generally trusts the core builders to have decision-making power, which is in a way a reaction to the previous DAO "maximum democracy" problem. In the long run, when you're facing 30,000 or 100,000 people, this approach will overwhelm the decision-makers. That's why we need automated structures to solve this problem - this is what we really want to do, which is to put the "A" (artificial intelligence) into the DAO.
Imagine if the proposal review process was fully automated, rather than manually reviewed. If people's proposals are not of good enough quality, the system can help them improve, or directly reject proposals that don't fit the current direction. Reviewers only need to review a small number of filtered proposals, rather than all of them.
This automation can be extended to various aspects - from collecting opinions to specific execution. Ideally, the DAO won't need any human operators, it will be fully self-governing, from front-desk reception to proposal submission, to payment approval, all handled by AI agents. Of course, this is a long-term goal, but this is the direction we want to go.
The Phenomenon of the Eliza Framework
Bankless: The Eliza framework is now one of the most popular projects on GitHub, why is everyone using Eliza? What's special about it?
Shaw: From a technical perspective, Eliza doesn't have anything particularly outstanding. Although we have indeed made some important technical innovations, such as the multi-agent room model, I think the real value is that we have solved the most basic social loop problem.
We developed a Twitter client that doesn't require an API, avoiding the $5,000 per month API fee. It uses the same GraphQL API as a regular browser, and can run in the browser. This makes the whole project viable, because you can easily spin up an agent and run it.
Additionally, we developed the framework in TypeScript, a language that most Web and Web3 developers are familiar with. We kept the framework lean, without over-abstraction, so developers can easily add the features they want.
AI Agents and the Future of the Crypto Industry
Bankless: The crypto market is very risky, and AI agents need to be thoroughly tested before they can replace human roles. Our goal is to replicate human behavior patterns in crypto into AI, right? In the long run, what do you think this ecosystem will look like when it matures?
Shaw: From the obvious long-term vision, we may reach the AGI (Artificial General Intelligence) stage within 5 to 50 years. Combined with Neuralink technology, everyone will have a second brain and be able to access all information at any time. This direction is very clear, the key is how to get there.
When all technologies converge, it will be a very beautiful scene, where everyone can obtain sufficient resources. But before that, during the transition period, there will inevitably be a lot of uncertainty, fear and doubt - interestingly, this is the origin of "FUD" (Fear, Uncertainty, Doubt).
Our goals are divided into two levels:
1. Practical level:
Develop usable AI agents
Build reliable infrastructure
Ensure system security
2. Spiritual mission:
Promote universal education
Empower user control
Protect data sovereignty
Just like the core idea of Web3, we hope that everyone can:
Create their own value
Own their own data
Understand and control technology
Participate in system improvement
Two Paths of AGI Development
1. Centralized control path:
Microsoft, OpenAI, etc. gain control through regulation
The government decides what can and cannot be done
I am very concerned about this path because:
OpenAI's models have performed poorly in some aspects
Models often have fixed value biases
A world where a committee decides what AI can say may lead to dystopia
2. UBI (Universal Basic Income) path:
AI will indeed replace many jobs
For example, 5% of jobs in the US are driving (trucks, Uber, etc.), and these may disappear within 5 years
Even programmers like us are now using Cursor and Claude extensively
But I have concerns about the implementation of UBI:
Reflecting on the rollout of government aid during COVID
The controversy over Obamacare
UBI may become a political compromise
Advice for New Developers
Bankless: If developers are using the Eliza framework and preparing to develop their first agent, what advice would you give them?
Shaw: First of all, don't worry even if you've never programmed before. We hold 1-2 AI agent development courses every week. I strongly recommend using Cursor, an AI-driven IDE, which can save you a lot of time. Claude is also a great tool.
Remember three points:
Keep the passion for learning, as technology is developing rapidly
Pay attention to security issues in development
Don't be afraid of failure, learn from practice
Bankless: Do you have any good learning resources to recommend?
Shaw:
AI Agent Development School - Systematic courses
Eliza framework documentation - Practical guides
High-quality open-source projects on GitHub
Bankless: Can you introduce us to Agent Swarming?
Shaw: Agent Swarming is the technology of having multiple AI agents work collaboratively. For example, one agent collects data, another analyzes it, and a third generates a report. These agents cooperate with each other to complete more complex tasks.
For developers who want to try this technology, I suggest:
First master the development of a single agent
Try the collaboration of two agents
Gradually expand to more agents
