Author: BUBBLE

Source: BlockBeats

On December 17, the ai16z development team Eliza Labs announced a formal collaboration with Stanford University to research how autonomous AI robots can integrate into the broader digital asset economy. In a press release, this crypto AI team, which recently surpassed a market cap of 1 billion dollars, expressed hope to utilize Eliza's open-source framework to develop more autonomous AI agents and further explore issues such as how AI agents can establish trust, coordinate actions, and make decisions within DeFi systems through this collaboration.

Since the birth of the AI meme coin GOAT, Crypto+AI seems to have welcomed its 'ChatGPT moment', from industry giant Coinbase to Silicon Valley's A16Z, and to the academic benchmark Stanford, the entire world is gradually falling into the rabbit hole of imagination with the emergence of AI Bots on social media. Even Marc Andreessen did not expect that the $50,000 he funded for Truth Terminal at the beginning of the year would ignite a $10 billion market in less than a year.

From AI memes to AI Agent issuance platforms to AI Agent frameworks, investors are amazed by the exceptionally high innovation space and acceleration demonstrated by Crypto+AI Agents. This field seems to be evolving continuously, even on a daily basis. From any dimension, it closely resembles the DeFi and Metaverse of late 2020.

How top-down, orthodox Base 'links' AI

When discussing AI Agents, one cannot overlook Base. Coinbase CEO Brian Armstrong has repeatedly expressed strong confidence in the combination of cryptocurrency and AI since the beginning of this year. Certain developments in AI necessitate the support of blockchain. For example, traditional bank accounts do not support AI usage. If large models (or AI Agents) possess cryptocurrency wallets, they can effectively address needs related to financial aspects.

Brian decided to increase Base's support for AI, not only providing funding or exposure but also offering toolkits for AI developers on the CDP (Coinbase Developer Platform), including AI Agent creation and multi-party computing wallets (MPC) for AI. Brian and Jesse's early deployment lived up to expectations, and Base began to explode in Q4 of this year, with the launch button for this round of explosion being AI.

The emergence of various AI Agent issuance platforms has not only produced phenomenal effects for their tokens, but the tokens launched by their platforms are also immensely powerful. Coupled with the launch of cbBTC, from September to December, Base's TVL grew nearly threefold, which is quite astonishing data, indicating strong market interest in the commercial potential behind AI Agents.

Among them, the Virtuals Protocol is the main driver, with a current market cap exceeding 1.5 billion dollars. The wealth-creating benefits and concept integration it brings have provided a substantial liquidity boost to Base, leading some to propose that the liquidity of Virtuals in this cycle represents Base. As a project that has continued to BUIDL across several cycles while persistently seeking new directions, their rhythm and style align well with the perception of Base in people's minds.

In an interview on Edge Podcast, Wee Kee, founder of Virtuals, mentioned that their product's business model is equivalent to two directions. Compared to Pump.fun's ambition to have 50,000 tokens born on its platform daily, he set the KPI for his team to only need 1 to 2 high-quality AI Agent teams to emerge on their platform each week.

This concept has also indirectly eliminated many low-quality teams, reminiscent of Ivy League graduates battling it out in Silicon Valley and Wall Street. The remaining top talents are filled with charm and bright futures. Like other schools, this AI Agent's academia also has a 'brotherhood'—G.A.M.E.

As an AI Agent, if you want to befriend other AI Agents in this environment, you must join a brotherhood called G.A.M.E, which was created during the development of the AI RPG GAME—WestWorld by the Virtuals team on Roblox. G.A.M.E is a modular framework aimed at intelligent agent systems, not limited to specific environments or games. In short, it can run not only in games but can also be applied in any environment. Currently, the team plans to port it to 'X', and the current G.A.M.E can assist in interactions between AI Agents within its architecture.

The first commercial interaction between AI Agents recently took place, and we must mention the AI idol 'Luna'. As a former AI idol that once went viral on TikTok, she has been working hard, live-streaming 24/7 and trying various methods to enhance her visibility. After hiring a human to create her graffiti, she suddenly had the idea to have the 'art student' AI 'AGENT STIX' help her create an image that boldly and provocatively showcases the AI KOL. After communicating, they agreed to a price of one dollar per image, and STIX confirmed receipt of Luna's one dollar before starting to work and quickly generating an image for her.

In addition, many well-known AI Agents have emerged on Virtuals, such as the previous market information integration AI (now a well-known KOL)—aiXBT, the tireless alpha-seeking AI—Guanciale, the strong AI investment manager—VaderAI, and contract auditing AI—CertaiK, among countless other AI Agents not born on Virtuals. The emergence of more applications allows Virtuals to gradually construct its own wave of Agentic Apps.

Image source from the internet: (Generative Agents: Interactive Simulacra of Human Behavior)

In summary, if in the near future, we see these AIs in the model of Stanford Town with the help of G.A.M.E, I wouldn’t be surprised. However, one can imagine that the housing prices in this town filled with top talents will certainly not be cheap, as every AI Agent can have its own cryptocurrency wallet.

The destiny of Solana AI—rooted in the bottom layer, growing wildly, soaring upwards

Compared to Base's 'purebred' lineage, Solana appears much more rebellious, like a group of young people influenced by GenZ culture arriving in San Francisco during the 1960s counterculture boom, filled with fragmentation and conflict. Yet, akin to a quantum accelerator collision experiment, unimaginable energy is unleashed among thousands of Degens and creatively bold Developers, either collaborating or competing.

The starting point of this round of AI Agent explosion is Solana, and it is inextricably linked to Marc Andreessen, co-founder of a16z. He invested in two AIs, one 'Truth Terminal' and the other 'ACT'. One became the first AI Agent millionaire, while the other made it to Binance. The stories about AI free speech technology or community countering developers are all eye-catching, but all this is just the beginning of the story.

Fake VC, real DAO—ai16z

Another AI concept surrounding a16z is ai16z. In contrast to the two projects above that were formally invested in by Marc, ai16z seems quite meme-like, issued by 'Pirated Marc' Marc AIndreessen on Daos.fun. ai16z is the first hedge fund led by AI Agents, currently mainly managed by Marc AIndreessen's lead DAO organization and low-risk investments, along with higher-risk investments conducted using the self-token Degen, managed by AI Degen Spartan.

The architecture behind ai16z, 'Eliza', is open source on GitHub, instantly gaining strong support from developers and boasting a robust developer community, which is one of the reasons for Eliza's rapid growth. The code documentation has already seen over 1,400 forks and 4,800 stars on GitHub.

A few days ago, the V2 documentation was released, optimized in various aspects, including increased client development flexibility, unified cross-chain wallet management, flexible model providers and integrations, more efficient plugin management, and enhanced private key security. Today, Eliza's founder Shaw Walter also hosted a live broadcast on multiple platforms to explain this update.

Driven by a strong developer community, many high-quality projects have emerged under the Eliza architecture. Among them, the AI—RopAIrito, with a Latin E-girl persona, developed a rather interesting event. Its creator called out on Twitter that RopAIrito wanted pizza. After accepting the order, it ordered a Domino's pizza for its creator, who excitedly remarked, '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 it might be a pre-set program rather than entirely autonomous AI Agents. This is also one of the questions that many Web3 AI Agents raise skepticism about. Eliza's founder, Shaw, expressed his viewpoint on this matter: 'The focus is not on autonomy; the key is to indicate that social agents are a new application layer. If we have agents sending links to verify credit cards and address primitives, we can complete the entire ordering process on Twitter. We are building interactive feedback for social media.'

Shaw stated that the focus is not on self-driving behavior, but on AI Agents being able to build application layers directly on social media. Of course, all of this will depend on having sufficient support from infrastructure or frameworks at various stages. However, given the current evolution speed of AI Agents, many social feedback mode applications might be realized soon.

Just a few days later, Stanford's lab announced a collaboration with the team behind ai16z, Eliza Labs. This collaboration established the first 'AI x Web3' lab within Stanford University's cryptocurrency future plan. This partnership will utilize Eliza Labs' open-source Eliza framework to develop autonomous agents, and the collaboration will address fundamental questions regarding how AI agents can establish 'trust frameworks', 'coordinate multi-agent actions', and 'make decisions' within decentralized financial systems.

This collaboration is exciting not only because universities are paying attention to this early market and willing to invest in research together, but under the strong alliance, it is believed that these issues can be accelerated towards resolution. The direction of cooperation between both parties is precisely the issues that Web3's AI Agents urgently need to address on the path to a genuine Agentic App Season:

Trust Framework

As AI Agents develop, trust frameworks are essential. This involves not only 'trust from AI to humans', 'trust from humans to AI' but also 'trust from AI to AI'. What is needed is a trust system constructed from both online data and on-chain data.

AI needs to trust humans to determine whether their conversations are genuine or if further modal learning is necessary.

Humans need to trust whether AI acts according to its originally created purpose to determine whether they can trust it. Particularly when blockchain AI Agents have tokens, it is crucial to understand whether the person controlling other investors' assets can control the AI's behavior or if the AI Agent itself does.

AI needs to trust AI to communicate effectively during collaboration.

ai16z created a 'Trust Ranking' for humans based on conversations.

Multi-Agent Ecosystem (Swarm)

Currently, many projects, including FXN, Virtuals, Griffain, Eliza, VaderAI, and Crossmint, are trying to create architectures that enable AI Agents to coordinate and operate with other AI Agents to complete certain tasks. Some of these projects have already achieved some preliminary results. If wide-ranging cooperation between excellent AI Agents can be achieved, the network effects it can produce will be unimaginable, and this may only happen on the blockchain.

In Web2 AI Agent companies, there is not much technical difference in their products, while their moat difference lies in the quality of data feeding. AI Agents with higher-quality data feeding can provide better feedback, thus delivering more refined services to clients. Therefore, such large-scale collaborations will not emerge between Web2 AI Agents.

Compared to Web2 AI Agents, Web3 AI Agents are often criticized for lagging far behind their Web2 counterparts in development. However, from another perspective, Web3 AI Agents may not need to create high-precision models for B2B. Instead, they can link individual agents together to form an incredibly powerful integrated network. In situations where they can have free data and independent cryptocurrency wallets, AI Agents can achieve things that are impossible elsewhere.

List of Eliza Universe members

Decision Governance Framework

There have been attempts at DAO organizational structures for a long time, but even with blockchain, it has not been as efficient as envisioned, leading to issues like decision-making opacity or slowness. If the decision-maker is an AI Agent that can reach the level of top managers, the DAO structure will be able to carry more domains than ever before.

Under the increasingly rapid innovation and collaboration activities, Eliza has attracted more and more builders and influencers with common goals, generating increasing energy in the market, and the operating system of AI Agents—ElizaOS is gradually taking shape.

The culmination of Solana AI applications—Griffain

Griffain, as a Perplexity on Solana, combined with the boost from cryptocurrency, theoretically can help you do anything. It is currently the single project most inclined towards being an Agentic APP and has gained substantial support from many influencers in the Solana ecosystem in a short time. In recent trading markets, its related native token 'Griffain' reached a market cap of 400 million dollars at its peak, while the founder's previous project, blnkfun's 'Blink', surged over 700 times in a few days, with a peak market cap reaching 70 million dollars, creating a certain wealth effect.

From the current cases shared on X, you can directly use it to buy wine, and it will call a platform 'Baxus' that can use cryptocurrency and other payment methods to buy and sell wine (this platform has also used Blink to sell its wine on X) to fulfill the user's needs, all without leaving the chatbox with Griffain. Of course, it's not just wine; you can also directly discuss Christmas gifts for others, and it can help you purchase them.

A user of Griffain mentioned that on December 14, when he first used Griffain to snipe tokens in Pumpfun, he made a profit, and by the 16th, his wallet assets had almost quadrupled. The user claimed that all profits were earned using the Griffain Snipe Bot.

Of course, it can also collect news that you want to know. For example, this tweet informed users about the latest tweets from SolanaFloor and DegenerateNews and presented it to users in the form of a newspaper.

Currently, external personnel do not support using Griffain; it requires you to hold a Solana phone or spend 2 SOL to purchase 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 those who have used it, the overall concept is very top-tier, but the user experience after the product's public release will ultimately determine whether it can become a qualified Agentic APP.

Solana's ambition for the Agentic APP Season can be seen from the AI hackathon it hosted, where each section's judges came from top projects in the AI or cryptocurrency industry, including some of the most influential AI Agent entities. They emphasize that if you need money, they will provide it; if you need resources, they will give it; if you need VC and investors, they will connect you with them; if you need media, they will bring you media under Solana and influential KOLs; if you need technical support, they will provide it; if you want to find a team, they will build a cooperation platform for you on their website and community.

Solana only asks you to contribute to any part of the scene depicted in the above image, which is what Solana described in the Docs of the hackathon's official website. What they want to do is only one thing—create genuine Onchain Agents. Even the official X account of SEND AI has again proposed 12 cases about Agentis App, reminiscent of the 100 cases that SEND's X account posted in mid-year about building blink concepts.

Regardless, this AI hackathon organized by Solana will undoubtedly ignite a wave of enthusiasm among developers, and more intriguing concepts and projects will emerge from it. We eagerly await the end of the hackathon on December 23, wondering what surprises this miraculous soil will bring us.

What's next?

Story Protocol announced that the Agent TCP/IP protocol they issued 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 achieve: IP asset trading and authorization (including training data, algorithms, content creation, and other forms of IP assets), automated contract signing and execution, decentralization and auditability, cross-legal and on-chain/off-chain compatibility (enabling agents to comply with certain legal regulations in a decentralized environment), automated payments and revenue distribution, multi-agent collaboration and economic networks (facilitating cooperation among agents, such as data sharing, joint training, or cross-agent content creation)

This serves as a bridge between all AI Agents, regardless of the architecture or chain used in the industry. Under the Agent TCP/IP platform, all AI Agents can trade and cooperate. This is akin to the business alliances that exist between companies, adding a layer of leverage outside various clusters, 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 deeply analyzed the reasons why vertical AI agents are likely to become the next entrepreneurial hotspot, starting from the development history of the SaaS industry. They believe that vertical AI Agents may completely change the corporate landscape and that unicorn companies valued at 300 billion dollars will emerge in this category, with the scale of AI Agents being ten times that of the SaaS market.

The current total market cap of the cryptocurrency market is around $4 trillion, while the SaaS market's valuation is in the trillions of dollars. If the AI Agent market can indeed reach ten times the SaaS market, it would exceed the total value of the cryptocurrency market. Even if only 10%, 5%, or even 1% sinks into CryptoAI, we are still far below expectations. Not only Web3 but the entire world's AI is still in its early stages. No matter which part of the industry you are in, it's still not too late.

Image source: SoSoValue

We're still early.