AI in Crypto: Feather Ruffling or Rebirth from the Ashes post-Meme Mania?

Original Title: "AI in Crypto: Is it a Muddle or a Rebirth after the Meme Frenzy? (Part One Part Two) By Wlabs"

Original Source: Melon Lab Wlabs

Part One

Introduction

Since ChatGPT made its debut at the end of 2022, the AI sector has been a hot topic in the crypto world. The nomads of WEB3 have always embraced the concept that "any concept can be hyped," let alone an AI with infinite narrative context and application capabilities in the future. Therefore, in the crypto community, the AI concept initially skyrocketed in a "Meme Frenzy" fashion for a while, and then some projects began to explore its actual application value: What new practical applications can crypto bring to the thriving AI sector?

This research article will narrate and analyze the current evolutionary path of AI in the Web3 field, from the early hype wave to the emerging application-oriented projects, and combine cases and data to help readers grasp the industry context and future trends. Let's start by throwing out some immature conclusions:

1. The AI meme phase is already a thing of the past, and what needed to be cut and what needed to be earned are all preserved as eternal memory fragments;

2. Some foundational WEB3 AI projects have always emphasized the benefits of "decentralization" for AI security, which users are not particularly interested in. Users care about "whether the token can make money" + "whether the product is user-friendly";

3. If you want to ambush crypto projects related to AI, the focus should shift to pure application-oriented AI projects or platform-oriented AI projects (which can consolidate many user-friendly tools or agents), which may be a longer-term wealth hot spot after the AI Meme.

Differences in the Development Path of AI in Web2 and Web3

AI in the Web2 World

In the Web2 world, AI is mainly driven by tech giants and research institutions, with a relatively steady and centralized development path. Large companies (such as OpenAI, Google) train closed black-box models, with algorithms and data not made public, and users can only use their results, lacking transparency. This centralized control leads to non-auditable AI decisions, with issues of bias and unclear responsibility. Overall, AI innovation in Web2 focuses on improving the performance of basic models and landing commercial applications, but the decision-making process is opaque to the public. It is this opacity that has led to the rise of new AI projects like Deepseek in 2025, which appear to be open source but are actually "fishing in troubled waters."

In addition to the opacity issue, large AI models in WEB2 also face two other pain points: insufficient user experience in different product forms and inadequate accuracy in specialized fields. For example, when it comes to creating a PowerPoint presentation, an image, or a video, users still tend to look for AI products with low entry barriers and better user experience to use, and are willing to pay for them. Currently, many AI projects are exploring no-code AI products to lower the user entry barrier.

Another example is that many WEB3 users should have experienced a sense of powerlessness when using ChatGPT or DeepSeek to obtain information about a specific crypto project or token. Large model data still cannot accurately cover detailed information in any sub-industry in the world. Therefore, another development direction for many AI products is to delve into and be precise in data and analysis in a specific sub-industry.

AI in the Web3 World

The WEB3 world is centered around the crypto industry, embodying a broader concept that integrates technology, culture, and community. Compared to WEB2, WEB3 strives for a more open and community-driven approach. Leveraging blockchain's decentralized architecture, Web3 AI projects usually claim to emphasize open-source code, community governance, and transparency to break the monopoly of traditional AI by a few companies in a distributed manner. For example, some projects explore using blockchain to verify AI decisions (zero-knowledge proofs ensure model outputs are trusted) or have AI models audited by DAOs to reduce bias.

Ideally, Web3 AI aims for "open AI," allowing model parameters and decision logic to be audited by the community, while incentivizing developers and users through token mechanisms. However, in practice, the development of Web3 AI is still constrained by technical and resource limitations: building decentralized AI infrastructure is extremely challenging (training large models requires massive computational power and data, yet none of the WEB3 projects have the funding capacity to match even a fraction of OpenAI's resources). Some projects claiming to be Web3 AI still rely on centralized models or services in practice, only incorporating some blockchain elements at the application layer. These Web3 AI projects are considered reliable and excellent, at least sufficiently developed in real-world applications; while the vast majority of Web3 AI projects are either pure memes or memes disguised as genuine AI.

Additionally, differences in funding and participation models also impact the development paths of the two. Web2 AI typically progresses through research investment and product profitability drivers, with relatively stable cycles. On the other hand, Web3 AI combines the speculative nature of the crypto market, often experiencing "boom and bust" cycles driven by market sentiments: funds rush in during concept hype, driving up token prices and valuations, but cool off swiftly when project hype and funds decline rapidly. This cycle makes the development path of Web3 AI more volatile and narrative-driven. For example, a conceptually lacking AI project may experience a surge in token prices due to market sentiment, while conversely, even technical progress struggles to attract attention during a market downturn.

When it comes to the main narrative of WEB3 AI, the "decentralized AI network," we are currently maintaining a "low-key and cautious anticipation." What if it actually becomes a reality? After all, WEB3 still has groundbreaking entities like BTC and ETH. However, at this stage, everyone still needs to pragmatically envision some immediately implementable scenarios. For example, embedding some AI Agents in current WEB3 projects to enhance the projects' efficiency; or combining AI with other emerging technologies to generate new ideas applicable to the crypto industry, even if they are just attention-grabbing concepts; or purely AI products serving the WEB3 industry, providing services that resonate with the WEB3 community, whether in terms of data accuracy or alignment with the work habits of WEB3 organizations or individuals.

To be continued, the next article will primarily focus on the five waves of WEB3 AI trends and review and comment on some products within them (such as Fetch.AI, TURBO, GOAT, AI16Z, Joinable AI, MyShell, and more).

Reference Article:

[ Web3 AI vs. Web2 AI: Why Open-Source and Transparency Will Win ]

Original Article Link

(Part Two)

Introduction

Every time we delve into deep industry research, the crypto market always experiences a major bull run. BTC has surpassed $100k, and ETH has surprisingly hit $2500. Is this the so-called law of hard work paying off? Just kidding, it's actually because the US and the UK reached a trade agreement, and the initial talks between the US and China went well (after Trump tweeted about it first, followed by official confirmations). This still indicates that the entire crypto industry remains a high-risk financial speculation market, moving along with the economic environment, without yet forming an independent market like gold and oil. This strengthens our resolve to find breakthroughs that can have long-term development in the crypto field, with AI + WEB3 being the race we currently value most. Continuing from our previous train of thought, this article will review the five waves of AI development in WEB3 over the past few years. Let's start by summarizing it in a table. Of course, we will focus on the current fifth wave:

A Look Back at the Five Waves of Web3 AI Projects

Since the intersection of blockchain and AI, the crypto space has experienced five major waves of AI concept frenzies. Each wave of frenzy has been led by iconic projects or events, showcasing the evolution from pure conceptual hype to gradual technical exploration.

First Wave (Early Exploration Phase) — Emergence of AI Concept Coins like Fetch.AI: Old Wine in New Bottles

As early as 2018-2019, there were projects attempting to combine blockchain and AI concepts, but with very limited impact. The real trigger for the first wave of frenzy was the market response following the launch of ChatGPT in 2023. OpenAI's ChatGPT gained popularity at the end of 2022, showcasing the power of AI in entering mainstream civilian domains. It felt somewhat similar to the internet boom that followed the popularization of home computers over 20 years ago, as well as the frenzy of mobile internet products triggered by the booming sales of smartphones about 10 years ago. The emergence of ChatGPT made ordinary users truly feel that it could enhance their work efficiency.

It should be noted that a prominent feature of the crypto community is "anything hot can be hyped," so ChatGPT sparked endless imagination in the crypto market about the prospects of "AI + blockchain." Starting in January 2023, a group of early AI concept tokens saw skyrocketing prices, initiating the first wave of bullish trends. A typical representative was Fetch.ai (FET): as an early Binance Launchpad project, claiming to build a decentralized AI agent network, the FET token's price surged rapidly in January 2023 (multiplying several times in a month), with a market capitalization exceeding $200 million at one point. During the same period, AI-related coins such as SingularityNET (AGIX) and Ocean Protocol (OCEAN) also led the market.

The hallmark of this phase was: old projects with new narratives — many AI concept coins issued between 2017 and 2019 were revived on the back of ChatGPT, becoming the focus of capital chasing. However, at that time, many of these projects were still in their early stages, with limited real-world applications. The price surges were mainly driven by market sentiment and the allure of the AI mega-trend.

Second Wave — This Round of the "Crypto Bull Market" Gave Rise to AI Meme Coins

If we define the continuous rise of BTC since the end of 2023 as the core theme of this round's "crypto bull market," then there were only two tracks that caused a stir during this period: one was the inscriptions and runes classified under the BTC ecosystem, which brought a wave of heat through innovative asset issuance; the other was the Meme track.

Some meme projects have begun to incorporate the AI concept, and as Generative AI continues to rise in popularity, the Crypto community has witnessed the emergence of an AI Meme Coin craze. A typical example is $TURBO — touted as the world's first meme coin created by AI. Digital artists used ChatGPT to help design and release the TURBO token, starting with a mere $69. Surprisingly, this frog-themed meme coin "orchestrated" by artificial intelligence experienced a rapid surge shortly after launch, attracting a large number of retail investors. TURBO rose to fame with the unique story of "AI creating the next great meme coin," and after landing on several mainstream exchanges, its price skyrocketed by hundreds of times within a few weeks, briefly joining the ranks of popular meme coins.

The key feature of this wave is: AI as a gimmick integrated into crypto subculture. Numerous new projects claim to combine AI tools such as ChatGPT, Midjourney, etc., to generate coin names, imagery, or storylines, introducing various AI elements into meme tokens. Inspired by the AI trend, investors rushed into these high-risk tokens with a gambling mentality. However, in essence, these meme coins lack AI technological substance; they are merely a community game capitalizing on the AI concept. For example, TURBO itself has no practical use, and the "AI-driven" aspect is only a marketing gimmick, with its value mainly supported by community consensus and speculative psychology. The second wave of enthusiasm gradually waned in the latter half of 2024, with many AI meme coins plummeting to near-zero, and the market began to calm down.

$TURBO Price Movement, Source: CoinMarketCap

The Third Wave — AI Agent Memes and the GOAT Myth

From the latter half of 2024 to early 2025, the crypto industry began to witness a new height of AI and blockchain narrative fusion: the AI Agents concept emerged, shifting the hype from simple memes to narrative-driven automated agents. The hallmark event driving this wave was the explosion of Goatseus Maximus ($GOAT).

$GOAT was initially released as a meme coin on the Solana blockchain, but unlike before, it was wrapped in an "AI Agent Genesis" story: an AI agent named "Truth GPT" (crafted by developer Andy Ayrey) aspired to create a viral meme coin, even attracting the attention and sponsorship of renowned venture capitalist Marc Andreessen (Andreessen personally provided $50,000 in seed funding to this AI). With such a storyline, GOAT instantly ignited market sentiment upon launch — its price surged over 120 times on the first day (a 24-hour increase of 12,000%), reaching a high of around $0.4664 in a few weeks! This rags-to-riches myth made GOAT a hot topic in the crypto community at the time, hailed as the "AI-empowered super meme coin."

The GOAT craze revealed the power of the AI Agent Narrative in the crypto space: a team or community envisions a story of an "autonomous AI" leading a project, giving the token a sci-fi-like gimmick. For example, claiming the AI agent autonomously manages funds, carries out tasks without human intervention, thus imparting a sense of mystery and futurism to the token. During this period, besides GOAT, various other AI agent-themed concepts emerged, such as TruthGPT, various AI trading Agent DAOs claiming to automatically trade coins, many of which also saw significant price surges in the short term.

However, most projects remain in the narrative stage, merely storytelling. The so-called AI agents are often just scripted pre-set logic, far from true autonomous intelligent applications, let alone discussing whether the application's effects are good or bad. This craze peaked around January 2025 and then collapsed—external events (such as Trump announcing the launch of his personal meme coin $TRUMP in January 2025, drawing market funds) caused speculative capital to flee, and coins like GOAT saw rapid price drops. The entire AI Agent sector went from being red hot a few months ago to hitting rock bottom.

$GOAT Price Movement, Source: CoinMarketCap

Fourth Wave—The Rise of a16z-aligned AI Ecosystem + AI Agent Foundation Concepts

After the hype of the previous waves receded, a group of AI projects more focused on technology and ecosystem building emerged by the end of 2024. In contrast to the earlier phases primarily driven by community hype, this stage saw the emergence of AI platform projects with a prominent background or a serious vision. Examples such as ai16z and Virtuals were seen as representatives of the "a16z-aligned AI ecosystem" (dramatically mimicking the well-known venture capital firm a16z). These projects' commonality is: building an AI Agent infrastructure, aiming to become the "app store" for AI agents in the Web3 space.

$AI16z was a phenomenally popular AI meme coin that emerged in November 2024, its name referencing the prominent venture capital firm a16z. The project essentially positioned itself as a Solana-based decentralized AI smart trading fund, i.e., an "AI investment DAO," leveraging an AI agent to access on-chain and off-chain market data, analyze community sentiment, and execute trades automatically. This novel concept instantly captured market attention upon its launch: within a day of going live at the end of October 2024, its market cap surged from zero to around $80 million, igniting the enthusiasm of speculators. Surprisingly, a16z co-founder Marc Andreessen even endorsed it on Twitter, personally retweeting the introduction and link to AI16z's concept. With endorsements from notable figures and the "AI investment + Meme" story, AI16z became the hottest crypto topic of that month.

There are also many industry doubts about its hype being higher than its actual implementation, but it is undeniable that it has opened up people's imagination about the "AI + DAO Autonomous Investment" model. In order to further enhance the token's value, the ai16z team proposed in December 2024 to evolve the project into an "AI Layer1" economic plan: to launch a Launchpad dedicated to serving AI agents, making the ai16z token the base currency for intelligent agent transactions and fee mediation, and designing node incentives and value buyback mechanisms, etc. In this phase, we can see the project team trying to transition from hype to building a sustainable ecosystem, supporting token value through technical strength and economic models, rather than just storytelling.

At the same time, the Virtuals protocol, as another representative project, has been very successful in ecosystem promotion. Their team has built a compelling narrative, attracting many developers to create AI agent projects on their platform, and also attracting a large number of retail investors. At its peak, the Fully Diluted Valuation (FDV) of the top projects in the Virtuals ecosystem once exceeded $5 billion, and the FDV of the token corresponding to the Eliza framework also reached about half of that, at the billion-dollar level. Other standout projects (such as AIXBT) saw peak market values reach hundreds of millions of dollars. During this period, capital gave "AI agent platform/framework" projects a very high valuation. However, at that time, the user base and actual revenue were far behind the valuation—for example, despite Virtuals claiming ecosystem prosperity, statistics showed that its actual daily revenue after going live was once less than $500, forming a stark contrast with its sky-high valuation—as the market turned cold in early 2025, the token values of these projects generally halved or even collapsed, proving that a grand vision and narrative alone were not enough to support the price.

Therefore, as of now, the representatives of the fourth wave have not truly become an "app store" for AI Agents, but instead more resemble platforms like Pump.fun that focus on AI Agent Meme. While everyone waits to see if they can make a comeback, the fifth wave is surging.

The Fifth Wave — A "Blooming" Closer to Real-World Applications

After a brief lull, the WEB3 AI sector warmed up again in early Q2 2025, and this time, the trend was already significantly moving from the virtual to the real: focusing on next-generation AI agent real platforms and tools; new AI ecosystem chains tailored specifically for AI projects; leveraging decentralized user data for value storage; and even integrating WEB3 AI with other high-tech fields like robotics. These all present a "blooming" trend, and we look forward to the day when these practical projects can truly bloom. Below, we list some examples of the fifth wave for your reference:

Zerebro

As an AI Avatar project in the Solana ecosystem, Zerebro aims to provide a platform for decentralized content generation and agent deployment, envisioning the creation of a "consumer-grade AI Agent Launchpad." The team announced the development of a platform called "Zentients" by the end of 2024, with the idea of using the ZEREBRO token as a medium for creating AI Agents and providing value feedback. The scenario depicted by Zerebro is that anyone can easily generate their own AI Agent on the platform (e.g., email assistant, coding helper, trading bot, etc.) and issue corresponding agent-specific tokens; the ZEREBRO token then serves as a universal payment and governance tool for all agent services, enabling value accrual. This model is seen as the AI Agent version of a real application store, with promising prospects.

Inspired by this vision, the price of the ZEREBRO token saw a sustained uptrend in early 2025. Especially after a period of consolidation, in April 2025, ZEREBRO surged 400% in just two days, attracting significant market attention. During the same period, some other related projects also experienced explosions: for example, the $AIOS token, claiming to be an "Agent Operating System," rebounded tenfold, revitalizing the previously dormant Crypto AI sector. Investors' imagination is shifting from a single AI Agent towards the underlying support platforms: the market is starting to pursue projects that provide standard interfaces, communication protocols, and deployment environments for AI agents, hoping they will become bridges connecting Web2 AI technology and Web3 value capture. While Zerebro and others have planned rich features, the actual products have not yet fully materialized during the hype phase, and many application scenarios are still in testing or demo stages.

Just last week, Zerebro once again set the market on fire, but in a bizarre and somewhat laughable manner: Zerebro's co-founder  Jeffy Yu posted a "suicide" video online, followed by an obituary platform publishing a related obituary, which was then featured in the San Francisco Chronicle. This sequence of events triggered a new meme token LLJEFFY's price to skyrocket to over 100 million U, only to plummet back to near zero in an instant. Some key opinion leaders (KOLs) noticed on-chain data showing financial activity in wallets associated with Yu after his alleged death, leading to the conclusion that this was a new exit scam. If the "fake death event" is indeed true, then Zerebro is likely one of the more elaborately played Meme coins, and everyone can collectively give it the finger. This shameless operation is a disgrace to the crypto world.

Metis AI Chain Hyperion + LazAI

Metis is a well-established ETH Layer2 project, and perhaps everyone's deepest impression of it is: it is the only project with ETH founder Vitalik Buterin's mother and aunt as founding partners. Since 2021, it has not missed the bull runs of recent years. It surged by a hundredfold at the beginning of 2022 during the bull market, saw a 10x increase in early 2024, and is currently back to its all-time low. The reason why Metis intersects with AI is because on May 7th, it officially announced that its "AI Chain Hyperion" testnet would launch on May 14th. Due to the low profile of the Metis team, we had to gather a lot of information to understand what AI Chain Hyperion and AI Chain LazAI are all about (yes, they will go all-in on AI in the future, planning to launch two AI chains).

The first appearance of the AI Chain Hyperion was to address the problem encountered by all current L1 and L2 solutions: WEB3 AI projects require a larger amount of computation to meet the demands of frequent read-write operations and state computations while still striving to maintain decentralization, a feature that the previous EVM environment could not support. Metis team's newly released Hyperion solution can be understood as: reshaping a new chain that is most suitable for WEB3 AI projects or gaming projects (both requiring high-performance computing). Through the Metis SDK and Alith AI Agent framework, it aims to solve the issues faced by development teams. Hyperion still uses $Metis as the sole token, which makes us more inclined to believe that the Metis team truly intends to make a tangible impact in the AI sector, rather than just shouting high-sounding slogans. This is perhaps the reason why the recent Metis token value has been showing signs of movement.

If the first AI Chain Hyperion was just a tentative move by the Metis team in the field of AI, as many have also heard other public chains announce their goal of creating a high-performance AI public chain, then the mysterious second AI Chain LazAI is their formal entry into the AI sector. LazAI will be a brand-new Layer1, aiming to address the valuation mechanism of the data contribution made by individuals (including individuals and institutions) in AI. At first glance, it sounds complicated, but consider this example, and you'll understand: everyone knows that tech giants like Alibaba or ByteDance are profiting by selling your personal data privacy to businesses, and you have no way to stop them unless you stop online shopping and watching short videos.

However, LazAI leverages the decentralized nature of blockchain, where individual data still belongs to the individual. When using AI, you typically need to input a large amount of individual data, which may be very valuable to certain projects or businesses. Users can then access the value of their own data, which inherently belongs to them, and this value can be reflected through tokens, among other means. Moreover, LazAI can rely on algorithms to differentiate the varying levels of value in different data, thereby minimizing the pain points in the crypto industry, such as studios or brush parties forcibly injecting useless data. Therefore, the combination of "decentralized categorization storage of personal data + data valuation mechanism" will be the key feature of LazAI. This is also the direction we believe the entire industry should strive for: how to use the inherent features of WEB3 to change some of the shortcomings of the WEB2 world.

It must be said that only a project team like Metis, with its prominent background and deep roots, dares to embark on such a grand narrative like Hyperion + LazAI. And only then can we believe that they may truly be working towards this goal. To be continued.

Original Article Link

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia