Shaw Walters is the founder of Eliza Labs, an open-source developer of the Eliza agent framework. This framework facilitates the development of autonomous AI agents. Walters launched ElizaOS (formerly ai16z), a decentralized AI-driven investment DAO, to explore the potential of AI agents in finance and other sectors. [1] [2]
Walters first became involved in blockchain through his initial interest in game development, particularly in exploring how assets and characters could function across different platforms. This interest led to work on NFT projects, where he contributed as an independent developer and collaborated with brands unfamiliar with the space. Through these experiences, he learned Solidity and began exploring how AI could integrate with on-chain identity and wallets.
His interest in AI predates the release of ChatGPT. During the early development of the GPT model, Walters joined a small community testing GPT-3, which he saw as a major advancement. By late 2021, he was focused on developing smart agents, earning a living by creating AI-driven virtual characters, and working on NFT projects. At the time, the concept of intelligent agents was still niche, with few discussions of its application in blockchain.
Walters later joined Webaverse as a lead developer after its NFT land sale, working on AI-driven characters in a metaverse environment. The project aimed to create a 3D world where intelligent agents could interact freely. However, a hack resulted in the loss of its treasury, putting significant pressure on the team. Webaverse later transitioned into MoeMate, shifting its focus to AI characters.
Following this, Walters worked on various projects, including MagicML and Project 89, co-founding an open-source intelligent agent initiative called Magic. The goal was to develop a no-code tool for building AI agents, but community feedback indicated a preference for direct assistance in agent creation. This led him to develop Eliza as an example within the Magic framework. However, the limitations of no-code development prompted him to abandon that approach in favor of a more direct build. Lacking immediate traction, he continued sustaining himself through other projects while advancing Eliza as an open-source initiative. [3]
“I realized that intelligent agents are actually a web-based technology. So, I refactored the entire system in TypeScript and named the project VEGENT (because the name "Agent" was already taken on NPM). Afterward, I renamed it Eliza, which became the core of the project. When the Meme trend emerged, my technology was ready.”
In an interview with Unchained, Walters discussed the inspiration behind launching ElizaOS. Walters had been focused on developing AI projects since COVID, aiming to build something that would gain traction. Despite long-standing interest and work in AI agents, an early attempt with a Twitter-based agent struggled to attract attention. He believed the missing component was its connection to crypto, including DeFi cycles and speculative markets.
His engagement with the hedge fund manager behind the first DAOs.Fun investment fund led to the launch of VEGENT Spartan, an autonomous AI investor. Through this collaboration, he was introduced to Bowski, the creator of DAOs.Fun, whom he met in person in San Francisco. During their discussion, Walters shared his vision for a community-driven autonomous investor that would reduce risks associated with fraud. His broader goal extended beyond speculation, focusing on investments that generate value.
This vision led to the creation of ElizaOS. Inspired by traditional venture capital models, Walters wanted a decentralized version that was accessible to anyone. The project started informally, with rapid iteration and visual concept development, eventually leading to its launch. The ElizaOS experiment was initially aimed at raising $75,000 in funding, but its impact grew as developers adopted its open-source framework, Eliza, for various applications. The project evolved into a broader research effort exploring how AI can leverage market and collective intelligence for investment decisions with an expanding ecosystem of AI agent applications. [4]
On the Delphi Podcast, Walters discussed his journey in AI and crypto, detailing past setbacks, including losing control of an open-source project sold for $75 million without compensation. He continued building despite financial struggles and project failures, eventually co-founding Magic before shifting focus entirely to AI agents. His latest project, Eliza, gained traction, with the open-source framework reaching a $2.5 billion valuation and trending globally on GitHub. Shaw emphasized AI agents' potential in DeFi, particularly in automating liquidity management and improving user accessibility. While in Asia, he engaged with developer communities, exploring new applications for AI agents beyond social media bots. [5]
On the Mid Mic Crisis podcast, Walters discussed his unexpected rise to prominence in the crypto-AI space due to his open-source work on AI agents. Before shifting to AI, he recounted his background in game development, augmented reality, and social media-related projects. His early work included collaborations with Microsoft and Webaverse, where he explored AI-driven 3D environments. He emphasized that his code is open-source and aimed at enabling automated trading rather than direct financial speculation. He also addressed misconduct accusations, comparing his work to foundational software that others repurpose. His contributions reflect a broader culture of open-source development in AI and Web3, where ideas and code are freely shared and iterated upon. [6]
At CATSTANBUL 2025, Walters introduced ElizaOS v2, highlighting improvements over its predecessor, including a streamlined architecture, enhanced agent planning, and a unified wallet system to address issues with scalability and interoperability. The update introduced a package registry, CLI tools, and database-driven characters for evolving AI agents. The team also announced an agent marketplace with media generation features and "Block Tank," an agent-driven funding initiative. Additionally, they launched a $10 million Magic Fund in partnership with Jupiter to support open-source AI developers, emphasizing the importance of integrating AI and crypto. The research preview was scheduled for March, with a public release in April. [7]
편집자
편집 날짜
January 31, 2025
편집 이유:
New Wiki Created 🎉