Original | Odaily (@OdailyChina)
Author | Asher (@Asher_0210)
Yesterday (December 23), Shaw, the creator of the Eliza framework and the founder of the ai16z DAO, was interviewed by Odaily while attending a fan meetup in Shanghai, sharing the latest developments of ai16z, degenai, the deep cooperation project ELIZA, and the impact of the AI sector on Web3 by 2025.
The following is a transcript of the interview
Odaily: You have been very popular in the Web3 space recently, which must have brought a lot of pressure. Can you share your personal experience and why you chose to step into the spotlight from an anonymous developer in the second half of this year?
Shaw: I didn't intentionally want to step into the spotlight, and I never expected this to happen. It feels very strange because I was just a programmer before, and no one knew who I was. I was even anonymous at the time, but I gradually felt disconnected from the crowd, so I wanted to try to be myself and present myself authentically online. Then I released the Eliza framework, which unexpectedly became very popular and completely changed my life. I never anticipated these things happening. In fact, I wasn't even aware that the developer field also had a so-called "spotlight".
There is indeed pressure from the outside world. I used to really enjoy promoting some other projects, simply because I personally thought those projects were great, and I wanted to share what I believed in. However, when the DAO started to become very large, everyone became particularly focused on token-related matters, which put a lot of internal pressure on me, and I could no longer freely discuss the projects I liked.
Odaily: The Eliza framework is now one of the most popular projects on GitHub. What's special about it?
Shaw: I think there are three main reasons. First, it's ready-to-use. Currently, there are many tools related to AI agents and large language models (LLMs), and although they are all great toolkits, you still need to build the entire agent yourself. But with the Eliza framework we built, you can just start it up and it will run directly, containing everything you might need. At the same time, it also provides many plugins, such as the Solana plugin, the EVM plugin, which have all been integrated.
Second, the core of an AI agent is around the network. I noticed that many people limit AI to Python research and model development, but I think what's more important for an agent is to connect all these models to people through the internet. Therefore, we built it using TypeScript, and Web3 is particularly suitable for the JavaScript and TypeScript ecosystem, which allows many developers to start using it directly.
Third, focus on the social loop. We have built connectors to social platforms like X, Discord, and Telegram, so developers can immediately run our tools and have their bots deployed to X without needing to write any code. Then they can add their own features on top of it, such as integrating EVM chains, and you can add a wallet to your agent with almost no code changes. Or if you have your own software and APIs, you can also call them through the agent. Starting from social media is key, and agents on social media can self-promote, letting more people know about their existence, and thus spread quickly. I think this is the reason for its success: agents can easily deploy to X, and then everyone who sees it will want to have one, and then they can also quickly acquire and use it.
Odaily: Last week, ai16z not only launched on the Binance alpha section, but also on the OKX contract trading platform. Do you have plans for further in-depth cooperation with mainstream exchanges?
Shaw: Of course, we have a dedicated team handling these matters, and I think we may see more token listings and related progress after the New Year holiday, but I can't promise any specific timeline.I believe the listings on Binance and Coinbase are an inevitable trend, but I'm more focused on the content related to AI agents, so I can't make any guarantees or provide investment advice.
Odaily: What's the difference between ai16z and the previous DAO organizations in the Web3 space?
Shaw: I think the core problem with the previous DAO organizations was that while they did achieve decentralization, they didn't do a good job on the self-governance part. It's not truly "self-governing", and still requires a lot of human communication and coordination. This has brought many problems, such as: a small group of people holding all the power. Usually, some "whales" control all the voting rights; for newcomers, it's difficult to enter and participate, to contribute to the DAO.
This short-sighted behavior clearly exposes many coordination issues in DAOs. And we believe this is where our technology can solve, allowing more people to truly participate and contribute, and provide rewards for their contributions. We are researching a "continuous retrospective fund allocation" mechanism. The idea is: if you contribute code to the codebase, or assist in the Discord working groups, or provide help in any meaningful way, those contributors will receive rewards. And this is different from traditional airdrops, it's more like a continuous way to reward contributors. We plan to execute this reward mechanism every month. Currently, we are working on developing the relevant functionality.
Odaily: So, do you think the ai16z DAO will have its own business model in the future? What is it?
Shaw: Yes, our business model is based on the value accumulation of the token. At the same time, we are also making some investments, mainly investing in projects that we believe are very valuable, such as token launch platforms, DeFi platforms, and projects in the ecosystem that use AI agents.
We are not like a traditional startup with VC support, but we do have a relatively large fund, which was donated by a whale. We are using this fund to invest in the projects we think are important, and to get returns from them. We are also developing our own token launch platform, and are currently collaborating with the teams we have invested in, which is a bit like the Baidu Fund concept, and we are also deepening our cooperation with Baidu Fund and some other projects.
Odaily: Last weekend, the ai16z Official Twitter mentioned that they will integrate their ecological token economic model, is it through staking, burning, etc. to increase the token value?
Shaw: I think burning tokens is not an effective strategy to some extent. At least based on my experience, destroying tokens sometimes gives a sense of "selling", and this approach doesn't seem to really help the long-term value of the token. In fact, if I burn the tokens myself, it may cause some dissatisfaction, because people may feel that this will weaken their incentives. However, I think staking is very important. I believe any mechanism that can lock up tokens will be very popular. What we are really focused on right now is that our tokens face huge liquidity issues after listing. For example, we have a token with a market cap of $1 billion, but only $4 million in liquidity, which is clearly not enough.
Odaily: Regarding degenai, you previously said on the X platform that you will soon launch its AI agent. Can you reveal what kind of AI agent it will be and what its core competitiveness is?
Here is the English translation:Shaw: Yes, degenai is about to debut, it is the first character we have launched, and it comes with a built-in wallet. degenai can obtain trading signals through social media, we will track whether certain KOLs (Key Opinion Leaders) truly share quality recommendations. If we follow the advice of these KOLs, will we be able to profit, or are they actually "harvesting the leeks"? The core of degenai is to establish a trust scoring system, and make trading decisions based on this data. The ultimate goal of developing degenai is to make these tools not only serve us, but also be used by users in their own communities. For example, users can trade through this proxy without having to manually perform complex trading operations on Phantom or DEX Screener.
At the same time, we are currently developing another character, the two characters use the same technology, but with slightly different functions. Although some information has been publicly disclosed, we are in the testing phase, and have only opened a very small range of Alpha chat groups, where invited users can share trading information in this Alpha chat.
Odaily: Regarding the ELIZA project that has a deep cooperation with a16z, you could have issued it yourself, but you handed it over to the AVA team. Do you think they are more professional and will make ELIZA better? How will you help the project in the future?
Shaw: The AVA team once showed me an early version of the project. I thought it was cool at the time and was willing to work together to ensure the project truly reflects high-quality standards. In addition, their team has some excellent developers. Their vision is to build a platform that not only provides decentralized functions, but also far exceeds the existing Eliza project in terms of user experience and functionality. They hope to redefine the positioning and application scenarios of the Eliza token, not just a token or AI project, but a comprehensive platform that encompasses a variety of tools, services and ecosystems. The support for ELIZA includes:
Technical support: We will continue to provide support at the technical level, such as code review, optimization and assistance in realizing the core functions of the project;
Resource integration: We will utilize our own ecosystem and resources to engage in deeper collaboration with them, including sharing market channels and industry networks;
Community promotion: We plan to use our own community resources to help them expand their user base, and ensure that existing Eliza token users can smoothly migrate to the new platform or accept new functions;
Financial support: If necessary, we will evaluate whether to support the long-term development of the project through direct investment or participation in the token economy.
We are very optimistic about their potential, and also believe that this project meets market demand and user expectations. They have the ability to promote the development of Eliza in a more free and creative way, and our support is to ensure that the project starts steadily, is consistent with our overall mission, while respecting its independence.
Odaily: Whether in the Web2 field or the Web3 field, everyone is optimistic about the AI sector in 2025. Do you think the performance of the AI sector in the Web3 field in 2025 will surpass the GameFi Summer and DeFi Summer of that year?
Shaw: I think it's possible. Ultimately, it still depends on the performance of the global economy. If mainstream assets like Bitcoin continue to perform well, the market may continue to rise. In addition, I firmly believe that AI will fundamentally change the way the internet operates, and AI agents are just the beginning. For example, when it comes to the X platform, I'm now seeing most of the information seems to be AI-related content, and this change really feels crazy.
Odaily: Finally, what advice do you have for new users entering the Web3 AI field?
Shaw: Do your own research (DYOR). I see a lot of people losing money because they didn't do any reading or understanding at all. They just blindly followed the trend, and only started complaining when they lost money. For behaviors like PVP, the trend changes so quickly that a few minutes can determine whether you make 1x, 10x, or lose all your money. This rapidly changing environment leads to a phenomenon of blind following. But those who really do a lot of research, understand the differences between all the projects, if you can spend a lot of time asking these questions and finding the answers, your chances of choosing the right investment will be much higher.
In addition, I think earning money is not just about buying a project, you can also do liquidity mining. Buy top-performing tokens and then put them into liquidity pools. Although you may not get a 10x return by missing the early opportunities, you will get long-term stable returns, which is the key.
