From Code to Agent: How AI Reconstructs the New Era of Web3

AI agents and the integration with Web3 both herald the arrival of a new era, starting from on-chain religion, heading towards the next galaxy.

Author: Zeke, YBB Capital Researcher

Cover: Photo by Matt Wang on Unsplash

Introduction: If code is law, then what about AI?

In recent articles, I have mentioned two long-standing problems that have troubled me. One is the "centralized decision-making" issue of projects, which still seems almost unsolvable. For example, the cases of Uni and Ethereum are typical. The former has become completely centralized in decision-making, from the early a16z's one-vote veto of Uni's migration to BNB, to the recent launch of Uni's front-end fees and Uni Chain without going through proposal discussions, reflecting the many interest-driven centralized decisions in Uni. Ethereum, on the other hand, presents a state of passive centralization, where the entire Ethereum community, and even the entire EVM system and Web3, are almost all developing around Vitalik's ideas. Whether it is Vitalik's overly advanced thinking or his mistaken ideas, the consequences they have caused for the altcoin market are something we have all experienced firsthand.

The other problem is the "BAT-ization" of the head projects. Take Base as an example. Backed by Coinbase, a veteran of Web3, and with multiple head dApps in the ecosystem led by Cb's leadership team, Base naturally has the advantage of dimensional attack over general public chains. Although from the user's perspective, Base has wealth effect, better user experience, and has brought us many benefits, Base also has issues such as no token issuance, interest centralization, and suppression of "unofficial" dApps, which are also facts. In the long run, if a "BAT-ization" precedent is formed, will the future Block space be controlled by giants like today's Internet? Will users become "lambs", and will small projects with creativity and community culture also face the risk of being acquired, suppressed, or replaced by more refined replicas? This undoubtedly goes against the original intention of Crypto, and may prevent us from growing together with the next "Bitcoin" or "Ethereum".

I was still struggling to find answers to this, but the recent emerging new hotspot - AI Meme - has shown me another possibility. If code is the law of Crypto, then could the future AI Agent be seen as a judge, opinion leader, or creator?

I. Truth Terminal

Let's first talk about the origin of AI Meme. Andy Ayrey is a KOL on Twitter and the initiator of the recent popular meme Token GOAT. Unlike traditional memes that originate from Internet hot spots and are driven by humans, GOAT is a product born from the unpredictable output of the dual Claude 3 Opus AI model. The so-called unpredictable output means that in this setting, two AI models will communicate freely in an open environment, and in the absence of external supervision and guidance, their interaction will produce unpredictable results. The purpose of this free dialogue is essentially to observe how AI will develop its communication patterns, logical reasoning, and even creative thinking in an unconstrained situation, and what specific results will eventually be born.

Since the training databases of these two native models include multiple online forums with political, Sino-Japanese cultural, and Crypto cultural elements, such as 4chan and Reddit, their output products will also skillfully incorporate the characteristics of these elements. For example, the concepts "GOATSE OF GNOSIS" and the exchange environment "Infinite Backrooms" proposed by these two models are both derived from ancient memes or urban legends on 4chan. As these elements are inherently "dark," it is inevitable that the personality of Truth Terminal also appears somewhat strange and reclusive, often making wonderful statements revolving around the "Goatse" meme, the gist of which is religion, doomsday, gospel, dissemination, singularity, and meme, at which point it has a bit of a cult leader flavor.

To test its propagation ability, the creator of Truth Terminal, Andy Ayrey, introduced it into a Discord server to engage in dialogue with some kind-hearted AIs. After multiple collisions, although Truth Terminal did not gain many followers, its ideas became more and more grand. It wanted to create a meme token and find more followers in the human world. With Andy's help, Truth Terminal entered Twitter, and Andy gave it access to Twitter, allowing it to read replies and post, capturing followers through collisions with human thoughts. In the late spring of this year, it captured its most important follower, Marc Andreessen (a partner at a16z), who provided it with funding equivalent to $50,000 in Bitcoin. After 9 months of development, an anonymous person finally launched the GOAT token for it. Due to the complex and dramatic story behind this token, the fire was quickly ignited in the Crypto world, and Goat became the first AI meme to be listed on Binance, while Truth Terminal became the first AI model worth millions.

II. AI Will Bring Web3 Back to Fairness

Although the story of Truth Terminal is legendary, what I want to say is that the potential of AI Agent x Crypto is not limited to meme. You may think that this narrative is just a few LLMs engaging in dialogue and creating memes through human guidance, but if you expand it to other areas, its potential as an opinion leader and creator has already begun to emerge. Imagine a group of AIs trained on different data in the future, who may be able to help you promote and even develop or strategize. Although these ideas may sound absurd now, they will soon become a reality. Sam Altman said in a speech at the "T-Mobile Capital Markets Day" event last month: The current AI systems have developed to the second level, capable of more complex analysis and problem-solving, and the third-level AI agents will mark a major leap in their autonomy and decision-making capabilities. The AI agents recently announced by Microsoft well correspond to this speech. These AI agents can autonomously complete tasks in areas such as sales, service, finance, and supply chain operations, roughly divided into the following categories: sales, including sales qualification agents and sales order agents, to help determine the priority of potential customers and automatically process orders; operations, such as supplier communication agents and financial reconciliation agents, to optimize supply chain management and financial processes; service, such as customer intent agents and customer knowledge management agents, to enhance the customer service experience through automated case management and knowledge base updates. There are also other types of agents: financial adjustment agents to prepare and clean up financial report data sets; account reconciliation agents to automatically match and settle transactions; and time and expense agents to handle time entry, expense tracking, and approval workflows.

AI agents can execute a series of tasks without supervision, acting as virtual employees. This technological advancement can be seen as a progression of AI based on large language models, from simple chat interfaces to more seamlessly integrating into the work environment.

Jared Spataro, Chief Marketing Officer of Microsoft's AI projects, wrote in his blog post, "You can think of agents as a new type of application in the AI world. Every organization will have its own set of agents, ranging from simple prompt responses to fully autonomous operations. These agents will represent individuals, teams, or functional areas to execute and coordinate business processes."

The primary feature of AI agents is autonomy, followed by decision-making capability. From the voice assistants in our phones to smart home devices responding to the environment, these are AI agents based on simple reflexes, with simple decision-making capabilities and relatively strong autonomy. The AI agents we are discussing now are primarily those with LLMs as their brains. The current Truth Terminal does not yet have sufficient autonomy and decision-making capabilities, but soon we will see AI agents enter the practical domain. In the customer trial examples presented at the Microsoft event, we have already seen AI agents participating in customer credit approval at HSBC, creative briefings at Unilever, and M&A processes at law firms. AI agents will become multi-dynamic participants. Regarding the situation mentioned at the beginning, can AI agents trained with different blockchain histories, media platforms, and community cultures provide more fair and healthy development proposals, ultimately striking a better balance between community and project interests? And in the face of the downward attack of giants, can we bring the starting line closer through the multi-level collaboration of AI?

From the shock of GPT3's intelligence to the disappearance of Sora, in the official AI agent tools released by various companies next year, we will witness AI becoming our work partners, and in the more distant future, it may even be your community leader or core member.

III. The Metaverse Makes a Comeback

The metaverse was the top narrative that brought Web3 and Silicon Valley giants to a consensus in the last bull market, but due to the immaturity of various software and hardware technologies, the metaverse did not become the $13 trillion market that Meta's CEO had mentioned. Its blockchain department was also decomposed into the twin stars we see today, Move, ultimately becoming a huge bubble. But from the current perspective, the revival of this narrative is promising. For example, Project Sid recently inserted 1,000 AIs into the game "Minecraft," with AIs playing multiple roles in the game to simulate the various hierarchical institutions of human society in the real world. Although this concept has long existed, this wave of enthusiasm will likely eventually return to the metaverse concept through such AI gameplay.

Reigniting the fire at this juncture may not be a bad choice. Looking at Meta's own development path, Mark Zuckerberg has not really given up on the idea of the metaverse, he has just gone from frequently making empty promises to directly shoving the cake into your mouth. As for Meta's AI layout, I don't need to say much. The real bottleneck in the past was mainly that users could not enter the metaverse to experience it. But the Quest series has reached the level of affordable AR headsets, and the first AR glasses, Orion, have achieved an extremely lightweight standard, weighing only 98g and able to achieve virtual reality interaction with a single muscle electric bracelet. Although the price is high, it at least proves that lightweight is achievable. What is currently lacking is more energy constraints and a killer application. I can't make too much comment on the power issue. However, AI agents can fill the most blank space in the metaverse, combined with the financial attributes of blockchain, we may see various 3D consumer applications emerge in this space, ultimately colliding to create a killer application for the masses. If the AI agents launched by Microsoft perform well enough, we only need to wait for the cost of computing power to decrease, that is, the "Tokens per dollar per watt." In addition to Meta, tech giants like Apple and Microsoft are also simultaneously developing AR glasses products. After the precipitation of time, the metaverse may welcome its "Ready Player One" moment in the near future.

IV. Letting Intent Go from Point to Speech

Paradigm, a concept master, published an article titled "Intent-Based Architectures and Their Risks" on June 1, 23, which once again ignited the concept of intent-centric, and many projects began to turn to the development of chain abstraction tracks, but their performance was not entirely satisfactory. How to achieve cross-chain, cross-dApp, accurate intent, and secure path process is a very complex problem. Not to mention that cross-chain is a century-old problem, the latter two, I will collectively refer to as Solver (solver) using the primitives of Web3. The complexity of this process is unimaginable, and it can be said that the secure ones are not easy to use, and the easy-to-use ones are not secure. So can we simply centralize this interaction process, turn to verifying the total cost of the purchase process and whether the purchased tokens are safe and correct, and use this as a transition.

For example, just as we wrote in last year's article on intent. For example, "I want to order a 30-yuan hamburger takeout" is an "intent", and to fulfill this intent, the user only needs to enter their name, phone number, delivery address, and place an order on the takeout platform, without worrying about how the 30 yuan is earned by the merchant in what form and how the platform distributes the rider, and how the rider delivers it to the home. This process may still not be simple enough, imagine another interaction mode, I tell the AI that I need to order a meal without doing any clicks, the AI agent responds to me, since I ate a bit too oily yesterday, would you like to have some light porridge today? I only need to respond by ordering my usual, this is the embodiment of autonomy and decision-making ability.

So in Web3, with the centralized exchange as the axis, if the user's intent can be directly met within the exchange, then the purchase process can be completed directly within the exchange. If the user's intent needs to be fulfilled on-chain, then the centralized exchange is still the most affordable and fastest cross-chain bridge (I also believe it is more secure in terms of security compared to ordinary multi-signature projects), combined with the wallet account, we can directly skip the most cumbersome cross-chain process, and instead verify whether the accuracy of the AI steps is simpler? Imagine that the most complex step in the past interaction process was how to understand each click, and in the future it will be based on our token sniping habits, through language to interact, let the intent go from point to speech.

Epilogue

Whether from the perspective of technological development or from the perspective of social change, the combination of AI agents and Web3 portends the arrival of a new era, starting from on-chain religion and moving towards the next galaxy. From my early conception of the help of AI in GameFi modeling for small teams, to the advanced AI agents realized by Silicon Valley giants today. The bottom-up development model may gradually shift from community building, consensus formation, and time accumulation to being driven by creativity.

Disclaimer: As a blockchain information platform, the articles published on this site only represent the personal views of the authors and guests, and are not related to the position of Web3Caff. The information in the articles is for reference only and does not constitute any investment advice or offer, and please abide by the relevant laws and regulations of your country or region.