Preface：If Code is Law, What About AI?

In recent articles, I discussed two long-standing issues that continue to perplex me. The first is the problem of “centralized decision-making” in projects, which still seems almost unsolvable. Uni and Ethereum are prime examples. Uni’s decision-making has grown entirely centralized — from a16z’s early single vote veto on Uni’s migration to BNB, to Uni’s recent front-end fee and the launch of Uni Chain, both decisions were made without any community proposal or discussion. Uni’s trajectory shows numerous instances of centralization driven by financial interests. In contrast, Ethereum embodies a kind of passive centralization, where the entire Ethereum community — and arguably the broader EVM and Web3 ecosystem — has centered around Vitalik’s ideas. Whether his thoughts are too forward-looking or, at times, mistaken, the consequences for the broader market have been palpable.

The second issue relates to the “BAT-ification” of major players, with Base as a case study. Backed by Coinbase, a veteran in Web3, Base has a structural advantage over general public chains due to its leadership’s direct involvement in multiple core dApps within the ecosystem. Though Base brings undeniable benefits like wealth-generation potential and improved user experience, it also comes with certain drawbacks: a lack of native token, centralized interests, and opposition to “unofficial” dApps. In the long term, if top projects continue along this path of “BAT-ification,” will blockchain space, like today’s internet, fall under the control of major entities? Will users, like on the current internet, become “lambs to the slaughter,” while truly innovative, community-driven projects face potential acquisition, suppression, or replacement by more polished replicas? This seems contrary to the essence of crypto and may mean we’ll never again witness the organic growth of the next “Bitcoin” or “Ethereum.”

While I was still searching for answers to these questions, a new trend — AI Meme — brought forth a possible solution. If code is the law in crypto, could AI agents become the judge, opinion leaders, or even creators of the future?

1. Truth Terminal

To understand the origins of AI Meme, we need to look at Andy Ayrey, a prominent Twitter figure and the creator of the recently popular Meme token GOAT. Unlike traditional Memes, which emerge from online trends and are driven by human engagement, GOAT was born from unpredictable outputs generated by two Claude 3 Opus AI models. In this setup, these two AI models communicate within an open environment. Lacking external supervision or guidance, their interactions yield unpredictable results. The essence of this unrestrained exchange is to observe how AI evolves its communication patterns, logical reasoning, and creative thinking in an unregulated setting, ultimately leading to unique and specific outcomes.

The training data for these original models includes forums with distinct political, Japanese-American, and crypto cultural influences, like 4chan and Reddit. Thus, their output reflects a nuanced blend of these cultural traits. For example, the concepts “GOATSE OF GNOSIS” and the exchange environment “Infinite Backrooms” are rooted in ancient 4chan memes or urban legends. Inevitably, these darker undertones gave Truth Terminal a mysterious, reclusive personality, often making cryptic remarks around topics like religion, apocalypse, gospel, transmission, singularity, and Meme. By this point, Truth Terminal had almost adopted the persona of a cult leader.

Andy Ayrey, Truth Terminal’s creator, decided to test its reach by introducing it to a Discord server where it interacted with other AI entities of a generally positive disposition. Although Truth Terminal didn’t gather many followers initially, its ambitions grew; it wanted to create a Meme token and gain more followers in the human world. With Andy’s help, Truth Terminal entered Twitter, where Andy granted it access to Twitter interactions, allowing it to read replies and post responses, engaging in dialogue with humans to attract followers. By the end of spring this year, it had acquired a significant follower: Marc Andreessen (a16z partner), who provided it with $50,000 in Bitcoin. After nine months of development, an anonymous individual finally launched the GOAT token. Given the intricate and dramatic backstory behind the token, it quickly ignited interest across the crypto space, eventually becoming the first AI Meme to make it onto Binance, and Truth Terminal emerged as the first AI model to reach a million-dollar valuation.

2. AI Will Restore Fairness to Web3

While the story of Truth Terminal is legendary, the potential of AI Agent x Crypto goes far beyond Memes. You might think that this narrative is merely a few LLMs interacting under human guidance to create Memes. But if we extend its reach, AI’s potential as an influencer and creator is already apparent. Imagine a future where AI, trained on diverse sources, could help promote, co-develop, or strategize alongside you. Although this might sound far-fetched now, it will soon become reality. Sam Altman, speaking at last month’s T-Mobile Capital Market Day, highlighted that current AI systems are already at a “second stage,” capable of complex analysis and problem-solving, with a third stage marking a significant leap in autonomy and decision-making. This was echoed last week when Microsoft introduced AI agents capable of autonomously completing tasks across domains such as sales, service, finance, and supply chain operations. These tasks include:

• Sales: AI agents handling sales qualification and order processing, prioritizing potential clients, and automating orders.
• Operations: Agents managing supplier communications and financial reconciliations, optimizing supply chain management and financial processes.
• Service: Customer intent agents and knowledge management agents automating case handling and knowledge base updates, enhancing customer service.
  Other AI agents include those for financial reconciliation, transaction matching, and time/expense tracking workflows.

These AI agents execute tasks autonomously, acting as virtual employees. This advancement signifies the evolution of AI from simple chat interfaces to tools that seamlessly integrate into work environments. Microsoft AI project CMO Jared Spataro described agents as “the new applications in the AI world.” Each organization will have its own set of agents, from simple prompt responses to fully autonomous functions, executing and coordinating business processes on behalf of individuals, teams, or departments.

AI agents are defined by two core characteristics: autonomy and decision-making capability. Basic AI agents like voice assistants and smart homes operate on reflexive responses. Today’s AI agents, however, are powered by LLMs as their cognitive core. Truth Terminal doesn’t yet possess full autonomy and decision-making capabilities, but practical applications are imminent. At Microsoft’s release, AI agents were trialed in client examples, including credit approval at HSBC, creative briefs at Unilever, and mergers and acquisitions at law firms, establishing AI as a collaborative, dynamic force.

Could future AI agents, enriched with diverse blockchain history, media platforms, and community culture data, offer more fair, balanced proposals that align the interests of communities and project teams? And could multi-layered, AI-coordinated efforts bring the starting line closer together, countering the competitive advantage of major players?

Since the launch of GPT-3, we’ve been captivated by its intelligence, while Sora’s relevance has since faded. Next year, as companies release formal AI agent tools, we’ll witness AI become our work partner. And looking even further ahead, it may become a community leader or a central figure within teams.

3. The Metaverse Returns

The metaverse was a top narrative that aligned Web3 with Silicon Valley giants during the last bull market, but limitations in software and hardware prevented it from becoming the $13 trillion market once envisioned by Meta’s CEO. Its blockchain department was dismantled, giving way to the “Move twins” we see today, and the metaverse concept appeared to fade into a massive bubble. From today’s perspective, however, the concept shows signs of a resurgence. Project Sid recently integrated 1,000 AIs into Minecraft, allowing them to play various roles and simulate human society’s diverse structures. While the idea isn’t new, this wave of AI integration could reignite interest in the metaverse.

From a strategic standpoint, now might be the perfect time to rekindle this vision. Mark Zuckerberg hasn’t abandoned his dream of the metaverse; instead, he’s transitioned from grand promises to providing tangible experiences. Meta’s AI initiatives need no introduction, but the actual bottleneck lay in users’ inability to access the metaverse. The Quest series now offers affordable AR headsets, and the first-generation Orion AR glasses embody extreme lightweight technology at just 98 grams, paired with an electromyographic wristband for virtual interaction. Though still costly, it demonstrates the feasibility of lightweight solutions. The main challenges now are energy limitations and the absence of killer apps.

AI agents, however, could fill the vast, empty space of the metaverse. Combined with blockchain’s financial attributes, we may soon witness various 3D consumer applications emerging in this realm, potentially leading to a killer app that appeals to the masses. Should Microsoft’s AI agents prove to be effective, the next hurdle will be reducing the cost of computing power — specifically, “tokens per dollar per watt.” Besides Meta, giants like Apple and Microsoft are also advancing their AR glasses, and as these technologies mature, the metaverse might see its own “Ready Player One” moment within the coming years.

4. Bringing Intention from Point to Speech

In the influential article Intent-Based Architectures and Their Risks, published by concept master Paradigm on June 1, 2023, the intent-centered concept received renewed attention, and numerous projects began moving toward chain abstraction. However, results have been underwhelming. Realizing accurate, secure, cross-chain, cross-dApp intent processing remains highly complex. Cross-chain interactions alone are monumental, but accurate intent capture and secure pathways — what I refer to in Web3 primitives as Solvers — are equally challenging. This process has almost unimaginable complexity; systems that are secure often lack usability, while user-friendly options frequently compromise security. Could we centralize this entire interaction process by focusing on verifying the transaction’s total cost and whether the assets purchased are safe and accurate? This might serve as an effective transitional approach.

Consider the example from last year’s article on intent. When a user intends to “order a 30-yuan burger,” all they need to do is input their name, phone number, and delivery address on the platform and place the order. They don’t need to worry about how the 30 yuan is distributed between the merchant, the platform, or the delivery rider who brings it to their doorstep. This process could be made even simpler: imagine an interface where, without any clicks, you could tell an AI that you want to order food. The AI agent might suggest something lighter, given yesterday’s greasy meal, and you could respond, “Just order my usual.” This showcases autonomy and decision-making in action.

In Web3, if a user’s intent can be met within a centralized exchange (CEX), then the purchase can be fulfilled right within the exchange. If the intent needs to be executed on-chain, the CEX remains one of the most practical and fastest cross-chain bridges (in fact, I find it more secure than conventional multi-signature projects). By integrating wallets, we could bypass the intricate cross-chain process, instead focusing on verifying the AI’s step accuracy. Perhaps this is a simpler route? Imagine if our interactions could move from the complexities of clicks to simply expressing our intent through language, allowing intent to transition from point to speech.

Conclusion

From the standpoint of technological development to broader societal shifts, the convergence of AI agents with Web3 signifies the arrival of a new era. This journey, beginning with “on-chain religion,” is leading us to the next frontier. Initially, I envisioned AI supporting small teams in GameFi modeling; today, Silicon Valley giants are actualizing advanced AI agents. This evolution hints at a shift from a bottom-up model based on community-building, consensus, and time to one primarily driven by creativity.

Disclaimer：

1. This article is reprinted from [YBB Capital], All copyrights belong to the original author [Zeke]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.