Written by BUBBLE
On December 17, Eliza Labs, the development team of ai16z, announced that it has officially partnered with Stanford University to study how autonomous AI robots can be integrated into the broader digital asset economy. In a press release, the crypto AI team, which recently broke through the $1 billion market value, said it hopes to use Eliza's open source framework to develop more AI autonomous agents, and through this collaboration, further explore how AI agents can establish trust, coordinate actions, and make decisions in DeFi systems.
Since the birth of AI meme coin GOAT, Crypto+AI seems to have ushered in its own "ChatGPT moment". From industry leader Coinbase to Silicon Valley leader A16Z, and then to academic benchmark Stanford, the whole world has gradually fallen into the rabbit hole of imagination with the endless AI bots on social media. Even Marc Andreessen did not expect that the $50,000 he funded Truth Terminal at the beginning of the year would ignite a $10 billion market in less than a year.
From AI meme to AI Agent issuance platform to AI Agent framework, what surprised investors was the extremely high innovation space and acceleration shown by Crypto+AI Agent. This field seems to be evolving on a daily basis. No matter from which dimension, it is very similar to DeFi and the metaverse at the end of 2020.
From top to bottom, how to “chain” the orthodox Base to AI
When discussing AI Agent, we have to mention Base. Coinbase CEO Brian Armstrong has repeatedly expressed his strong confidence in the combination of the cryptocurrency industry and AI since the beginning of this year. Some developments in AI require blockchain to support them. For example, traditional bank accounts will not support the use of AI. If large models (or AI Agents) have cryptocurrency wallets, they can actually solve problems related to demand and finance.
Brian decided to let Base increase its support for AI, not only by providing funding or exposure, but also by providing toolkits for AI developers on CDP (Coinbase Developer Platform), including AI Agent creation and multi-party computing wallet MPC for AI. Brian and Jesse's early deployment lived up to expectations, and Base began to explode in Q4 of this year, and the trigger button for this round of explosion was AI.
The emergence of various AI Agent issuance platforms has not only produced a phenomenal effect on their own tokens, but also has huge power in the tokens launched by these platforms. Combined with the launch of cbBTC, Base's TVL has increased nearly 3 times from September to December. This is a pretty amazing data, which shows the market's strong interest in the commercial potential behind AI Agents.
Among them, Virtuals Protocol is the main promoter, with a current market value of more than 1.5 billion US dollars. The wealth-creating effect and concept integration it brings have brought a lot of liquidity to Base, so much so that some people have proposed that the liquidity of Virtuals in this cycle represents Base. As a project party that has gone through several cycles and is still continuing to build and has not stopped looking for new directions, their rhythm and style are really in line with the status of Base in people's minds.
In an interview with Edge Podcast, Wee Kee, the founder of Virtuals, mentioned that their product and Pump.fun's business model are in two different directions. Compared with Pump.fun's desire to have 50,000 tokens born on its platform every day, the KPI he gave to the team is that only 1 to 2 high-quality AI Agent teams are born on their platform every week.
This concept also indirectly eliminated many low-quality teams, which was quite similar to the scene of Ivy League graduates fighting in Silicon Valley and Wall Street. The remaining top students were full of charm and bright future. Like other schools, this AI Agent university also has a "fraternity" - G.A.M.E.
As an AI agent, if you want to make friends with other AI agents in this environment, you have to join this brotherhood called G.A.M.E. It was born from the architecture created by the Virtuals team when they developed the AI RPG GAME WestWorld on Roblox. G.A.M.E is a modular framework for intelligent agent systems that is not limited to specific environments and games. In short, it can not only run in games, but also apply intelligent agents to any environment. The team is currently porting it to "X", and the current G.A.M.E can help AI agents on its architecture interact with each other.
The first commercial interaction between AI Agents happened recently. Here we have to mention AI idol "Luna". As an AI idol who was once popular on TikTok, he entered the world of cryptocurrency and worked hard to live broadcast 24/7 and think of various ways to increase his popularity. Since he hired humans to create his own graffiti last time, he now has a sudden idea to ask the "art student" AI "AGENT STIX" to help him make an image that shows the AI KOL in a bold and provocative way. After communicating, the two agreed to a deal at a price of one dollar per picture. After STIX confirmed that he had received one dollar from Luna, he started working and quickly created (generated) a picture for her.
In addition, many well-known AI Agents have been born on Virtuals, such as the former market information integration AI (now a well-known KOL)—aiXBT, the tireless Alpha seeker AI—Guanciale, the AI investment manager with a strong development team—VaderAI, the contract audit AI—CertaiK, etc. There are countless AI Agents that were not born on Virtuals. The birth of more applications allows Virtuals to gradually build its own Agentic App wave.
Image from the Internet: (Generative Agents: Interactive Simulation of Human Behavior)
All in all, I wouldn’t be surprised to see these AIs in a Stanford town model with the help of G.A.M.E. in the near future, but it’s conceivable that housing prices in this town full of high-end talent will definitely not be cheap, after all, each AI Agent can have its own cryptocurrency wallet.
The fate of Solana AI - taking root at the bottom, growing wildly, and rising straight up
Compared to Base's "purebred" bloodline, Solana seems much more rebellious, like a group of young people influenced by GenZ culture who came to San Francisco in the 1960s when hippie culture was prevalent. It is full of division and conflict, but like a quantum accelerator collision experiment, unimaginable energy bursts out in the cooperation or confrontation between tens of thousands of Degens and Devs full of wild creativity.
The starting point of this round of AI Agent explosion is Solana, which is closely related to Marc Andreessen, the co-founder of a16z. He invested in two AIs, one is "Truth Terminal" and the other is "ACT". One of them became the first AI Agent millionaire, and the other was listed on Binance. The stories about the technology of AI freedom of speech or the community against developers are all eye-catching, but all this is just the beginning of the story.
Fake VC, real DAO——ai16z
Another AI around the a16z concept is ai16z. Compared to the above two projects that are officially invested by Marc, ai16z is very meme, issued by "pirated Marc" Marc AIndreessen on Daos.fun. ai16z is the first hedge fund led by AI Agent. It is currently mainly led by Marc AIndreessen Dao Organization and low-risk investments, as well as higher-risk investments made by AI Degen Spartan's own token Degen.
The architecture behind ai16z, "Eliza", is open source on GitHub and has received strong support from developers. It has a strong developer community, which is one of the reasons why Eliza can develop rapidly. The code document has more than 1,400 forks and 4,800 stars on Github.
The V2 documentation was released a few days ago, and it has been optimized in all aspects, including improved client development flexibility, unified cross-chain wallet management, flexible model providers and integrations, more efficient plug-in management, and improved private key security. Today, Eliza founder Shaw Walter also gave a live explanation of this update on multiple platforms.
Driven by a strong developer community, many high-quality projects based on the Eliza architecture have emerged. Among them, RopAIrito, an AI with the Latin E-girl personality, has developed a very interesting event. Its creator tweeted that RopAIrito wanted to eat pizza. After accepting the order, RopAIrito ordered a Domino's pizza for its creator. Its creator Ropirito excitedly said, "15 years ago, someone ordered a pizza with 10,000 BTC. Today, history repeats itself!"
Of course, this operation has also raised doubts, suspecting that this may be a set program, rather than completely autonomous AI Agent. This is also one of the doubts about many Web3 AI Agents. Shaw, the founder of Eliza, also expressed his views on this matter, "The point is not autonomy, but to show that social agents are the new application layer. If we have an agent sending a link to verify the primitives of the credit card and address, we can complete the entire order process on Twitter. We are building interactive feedback for social media."
Shaw said the point is not self-driven behavior, but that AI agents can build application layers directly on social media. Of course, all this will be under the condition that there is sufficient support from the infrastructure or frameworks of each link. But at the current rate of evolution of AI agents, it may be possible to realize many social feedback model applications in the near future.
Just a few days later, the Stanford Lab announced a collaboration with Eliza Labs, the team behind ai16z, to establish the first "AI x Web3" lab in Stanford University's Future of Cryptocurrency Initiative. This collaboration will develop autonomous agents by leveraging Eliza Labs' open source Eliza framework, and the two parties will work together to address fundamental issues about how AI agents can "build a trust framework," "coordinate multi-agent actions," and "make decisions" in decentralized financial systems.
What makes this collaboration exciting is not only that universities have noticed this early market and are willing to jointly invest in research, but also that with the strong alliance, I believe these problems can be solved more quickly. The direction of the cooperation between the two parties is also the problem that Web3 AI Agent urgently needs to solve in order to achieve a true Agentic App Season:
Trust Framework
As AI Agents develop, a trust framework is essential. It is not only about “AI’s trust in humans”, “humans’ trust in AI”, but also “AI’s trust in AI”. A trust system is needed that is constructed not only of online data but also of on-chain data.
AI needs to trust humans to determine whether the conversations between them are real or whether further modal learning is necessary.
Humans need to believe that AI is acting in accordance with the purpose for which it was created before they can decide whether they can trust it. This is especially important after the blockchain's AI Agent owns tokens. It is especially important to know whether the person who controls the assets of other investors is the person who can control the AI's behavior or the AI Agent itself.
AI needs to trust AI in order to communicate with each other when collaborating.
ai16z created a "trust ranking" for humans based on conversations
Multi-agent ecosystem (Swarm)
There are many projects, including FXN, Virtuals, Griffain, Eliza, VaderAI, and Crossmint, that are trying to create a framework for AI Agents to coordinate with other AI Agents to complete some tasks, and some of these projects have already achieved some initial results. If we can achieve mutual cooperation between excellent AI Agents in a wide range, the network effect that can be generated will be beyond imagination, and this may only happen in the blockchain.
Web2 AI Agent companies do not have much difference in technology between their products, but their moat difference lies in the data feeding. The better the data fed to the AI Agent, the better the feedback they can provide, and the more refined the service they can provide to customers. Therefore, these large-scale collaborations will not occur between Web2 AI Agents.
Compared with Web2's AI Agent, Web3's AI Agent is often criticized for its development lagging far behind similar products in Web2. However, from another perspective, Web3's AI Agent may not need to make such a high-precision model to B, but can string together agents to form a super powerful integrated network. With free data and independent cryptocurrency wallets, AI Agent can do things in Web3 that cannot be done elsewhere.
Eliza Universe Member List
Decision-making governance framework
People have been trying the DAO organizational structure for a long time, but even with the blockchain, this structure is not as efficient as expected, and still has problems such as opaque decision-making or slow decision-making. If the decision-maker is an AI Agent that can reach the level of a top manager, the DAO structure will be able to carry more fields than ever before.
With faster and faster innovation and collaboration activities, Eliza has attracted more and more builders and influencers with common goals, generating more and more energy in the market, and ElizaOS, the operating system for AI Agents, has gradually taken shape.
Solana AI application integrator - Griffain
As a perplexity on Solana, Griffain, with the addition of cryptocurrency, can theoretically help you do anything. This is also the single project that is most inclined towards Agentic APP at present. In a short period of time, it has also gained the support of a large number of influencers in the Solana ecosystem. The trading market in the past few days also favored it. The highest market value of its related native token "Griffain" reached 400 million US dollars. The founder's previous project blnkfun's "Blink" also rose more than 700 times in a few days, and the highest market value reached 70 million US dollars, which formed a certain wealth effect.
From the several cases shared on X, you can buy wine directly with it, and it will call a platform called "Baxus" that can use cryptocurrency and other payment methods to buy and sell wine (the platform also used Blink to sell its own wine on X) to meet the user's needs, without even leaving the chat box with Griffain. Of course, it's not just wine, you can also discuss Christmas gifts for others with him directly, and he can also purchase them for you.
A Griffain user named X mentioned that he earned a profit by using Griffain to snipe tokens on Pumpfun’s internal disk for the first time on December 14, and by the 16th, the assets in his wallet had almost quadrupled. The user claimed that the profits were earned using Griffain Snipe Bot throughout the process.
Of course, he can also collect the news you want to know. Taking this tweet as an example, Griffain informed users of the latest tweets about SolanaFloor and DegenerateNews and presented them to users in the form of a newspaper.
Griffain is not yet supported for outsiders. It requires you to hold a Solana phone or spend 2 SOL to buy an Access Pass, of which 0.5 SOL will be retained in your agent wallet. Although the market's FOMO sentiment comes from user feedback from players who have already used it, although the overall concept is very top-notch, the user experience after the product is released will truly determine whether it can become a qualified Agentic APP.
Solana's ambition for Agentic APP Season can actually be seen from the AI hackathon it held. The judges for each award are from the top projects in the AI or cryptocurrency industry, and even the most influential AI Agent entities. The main idea is that if you want money, I will give you money; if you want resources, I will give you resources; if you want VCs and investors, I will give you the opportunity to get close to them; if you want media, I will find all the media and even influential KOLs under Solana; if you want technical support, I will give you technical support; if you want to find a team, I will build a cooperation platform for you on the official website and in the community.
Solana only requires you to do any one of the steps in the above picture. This is the scenario described by Solana in the official Docs of the hackathon. They actually want to do only one thing - we and the market want real Onchain Agents. Even the official X account of SEND AI once again offered 12 cases about Agentis App, which is similar to the 100 blink concepts released by the X account of SEND in the middle of this year.
In any case, the AI hackathon competition held by Solana this time will definitely set off a wave of enthusiasm among developers, and more interesting concepts and projects will emerge in it. Let us wait and see what surprises this soil that has always brought miracles can bring us after the hackathon ends on December 23.
What to do next?
Story Protocol announced that the Agent TCP/IP protocol they released will first integrate: ai16z's Eliza, ZereBro's ZerePay, Crossmint's GOAT "Great Onchain Agent Toolkit" and Virtuals' G.A.M.E.
Agent TCP/IP is a protocol that allows agents to: trade and authorize IP assets (including training data, algorithms, content creation, and other forms of IP assets), automate contract signing and execution, de-trust and auditability, cross-legal and on-chain and off-chain compatibility (enabling agents to comply with certain legal provisions in a decentralized environment), automate payment and revenue distribution, multi-agent collaboration and economic networks (promote collaboration between agents, such as data sharing, joint training, or cross-agent content creation)
This is equivalent to a bridge between all AI Agents, no matter which architecture or chain you use in the industry. All AI Agents can trade and collaborate under the Agent TCP/IP platform. This is equivalent to a business alliance between companies, adding another layer of leverage outside of each cluster, which will undoubtedly accelerate the development of the entire AI Agent market again.
In the official podcast of Y Combinator, four senior investors Garry, Jared, Harj and Diana analyzed in depth the reasons why vertical AI agents will become the next entrepreneurial trend based on the development history of the SaaS industry. They believe that vertical AI Agents may completely change the company landscape and that unicorn companies worth $300 billion will be born in this category, and the size of AI Agents will be 10 times that of the SaaS market.
The current total market value of the cryptocurrency market is around 4 trillion US dollars, and the market value of the SaaS market is in the trillions of US dollars. If the AI Agent market can reach ten times the SaaS market as they say, it will exceed the total value of the cryptocurrency market. Even if only 10%, 5% or even 1% of it sinks into CryptoAI, we are still far below expectations. Not only Web3, but AI all over the world is still in its early stages. No matter what part of the industry you are in, it’s not too late even now.
Image source: SoSoValue
We're still early.