Author: Weilin, PANews
To get closer to the community, Shaw, who had been working behind the scenes focusing on technology, stepped into the spotlight. In late December, during the important overseas holiday—Christmas, Shaw and his wife chose to come to China to meet with community developers. From Shanghai to Beijing and then to Hong Kong, Shaw's schedule was very tight, and he was queued up for questions at every event.
As the hottest AI agent project, the market capitalization of ai16z has surpassed $1 billion. The core of ai16z, ElizaOS, is a powerful multi-agent simulation framework designed for creating, deploying, and managing autonomous AI agents. It is built using TypeScript and can interact across multiple platforms. Many projects have already been developed based on this highly praised open-source framework.
On December 26, at an ai16z community exchange event held in Beijing, PANews engaged in an in-depth dialogue with Shaw, the founder of ai16z and the creator of ElizaOS, discussing the project's original intentions, progress, future development directions, and views on the future of AI Agents.
Shaw said he is not a money-loving person, but his development work aims to provide investment opportunities for ordinary people, allowing the crypto ecosystem to shift from speculation to generativity, moving away from gambling on tokens and instead investing in things that can truly expand the pie.
Regarding project progress, ai16z will soon announce the list of partners to improve transparency and will unveil the new proposal for ai16z's tokenomics around January 1. Shaw believes that in response to how to deal with an AI-hosted world, people should have their own data and models, and high decentralization is key.
The founding intention of ai16z: to give ordinary people the opportunity to become investors.
In the interview, Shaw admitted that the birth of ai16z stemmed from observations of the current investment landscape and a desire for fair investment opportunities. He mentioned, “If you want to invest in top startups like Uber or OpenAI, you have to be an institutional investor like a16z. Ordinary people have almost no access to these opportunities.”
Shaw believes that crypto ecosystems like Solana can transition from speculation to generativity, moving away from gambling on this token or that token, and instead investing in things that can genuinely expand the pie. Just like companies and software. For example, when a16z, Paradigm, or other large venture capital firms invest in software companies, they also incorporate them into their networks, driving value realization. They can make money because they create something of great value. Shaw hopes that similar opportunities can be widely used by ordinary people and communities.
He recalled the project's early stage: “I collaborated with baoskee from daos.fun. We had lunch together, and I said I wanted to create something like an AI version of a16z. He said, then just call it ai16z and put it together. So we started working on it right after that lunch.”
Latest updates: Partner list will be announced, and a new proposal for ai16z's tokenomics.
As the market capitalization of ai16z surpasses $1 billion, Shaw revealed a series of exciting project dynamics. In the short term, Shaw believes that many people in the crypto space do not understand DeFi and do not know how to make their funds in wallets earn more returns like banks or other financial institutions. ai16z hopes to build a bridge connecting DeFi with ordinary users.
Therefore, ai16z is developing some tools, such as an automated liquidity pool (LP) Agent. This agent can automatically manage liquidity pools for users, helping them earn income. The autonomous investor tool developed by ai16z allows the community to invest in projects they believe in. This autonomous investor is not solely determined by AI but is collectively decided by the community, with AI listening to everyone's opinions, organizing them into an investment thesis, and then making decisions on behalf of the community. ai16z is interested in how AI can solve many coordination and communication issues within DAOs and communities to help them make collective decisions.
Additionally, Shaw stated that the project team will officially announce the list of all partners to enhance project transparency and help the community better understand the relationships and specific scope of cooperation. Regarding tokenomics, in response to previous industry criticism about the lack of clarity in token economics, ai16z will unveil a new tokenomics proposal around January 1. This proposal is designed by professionals from multiple teams and will include LP pairing mechanisms, DeFi feature integration, and more. Moreover, the project team revealed that they have acquired a Launchpad company, which will be used for the official release of AI agents in the future.
Meanwhile, the recent AI Pool project that Shaw has supported multiple times on social media has attracted market attention. Shaw stated that AI Pool is created by partner Skely, so it is primarily his project, but it is built on our ElizaOS technology. It showcases an application process similar to pump.fun, but this process happens in a fairer way on social media, so users do not need to go elsewhere. This also demonstrates a truly autonomous agent, which is completely self-governing—ai16z has no control over it, nor can it access the private keys. Funds cannot be withdrawn, and it cannot be manipulated; it operates autonomously.
Facing competition from Zerebro and ARC, he admits they are good complements.
Shaw stated that projects like ARC (written in Rust) and Zerepy (written in Python) are good complements to ai16z's work on Eliza OS. “I know the reputation of Swarm; they are known for stealing the work of famous researchers and influencing them in practically ineffective ways. So I wouldn't list them here. But for any real open-source projects and true developers who enjoy collaboration, I am very willing to work with them. This week I contributed to Zerepy. We are also collaborating with Zerebro; although people consider it our biggest competitor, we are actually partners.”
Currently, DeFi, gaming, and social media are the three pillars of ai16z's work. Shaw believes that social media helps gain attention and rapid development; DeFi provides the greatest value to ordinary users, helping them earn returns more securely; gaming also attracts a lot of interest. Currently, Treasure DAO's Smolworld and Loot Realms' Eternum are both using Eliza technology for development. Although these three are the most obvious directions, more fields may emerge in the future.
In summary, Shaw's advice to the project team and users is that anything that helps users make money, save costs, or improve their lives is bound to stand out. “I think simply promoting a token, like flooding Twitter with messages, may not be as important as before, especially as the number of agents increases. The focus should be more on what this agent can do and what it can do for me. Therefore, it should focus on lowering the barriers to using things like financial products, bringing applications that originally required website access to social platforms, thereby simplifying the overall user experience.”
Future trends: Different stages of AI agents, how to respond to an AI takeover world.
Currently, AI agents have experienced different stages of development, from the initial meme to expansion into social, then to infrastructure and investment analysis. In Shaw's view, the biggest trend he sees now is that websites are migrating to social media.
All the purposes people used to access websites in the past can now potentially be achieved through agents on social media. Therefore, this trend indicates that people are looking for ways to integrate applications directly into social media. He mentioned that he had previously communicated with someone developing a travel booking agent, and they are also working with someone developing a real estate booking system. A clear DeFi example is Clanker, which is developed for Forecaster. I think this is a great example of this trend, where you directly place the pump.fun model onto social media, creating a social minting experience that is more engaging, self-promoting, and self-marketing, allowing users to use the application without leaving the social platform. I believe there will be thousands of applications following this trend in the future.
Looking toward the future, Shaw emphasized that diversity and high decentralization are key to preventing an AI takeover world. He stated that decentralization can ensure that each community has complete control over its agents and can effectively respond to the risks of an AI takeover.
“I think the most important thing is diversity, making AI agents very different from each other. They use different models, strategies, and methods, built by different teams with different incentive mechanisms, which can create a stronger network, making it very difficult for any single AI to take over. A very important point is that all of this should be highly decentralized, and we should have our own data and models, allowing each community to fully control their own agents. This way, in the hypothetical case of some AI takeover, that AI would need to contend with many other AIs. There should not be just one superintelligent AI, but many superintelligent AIs, making it much harder for any single AI to gain an advantage.”