Written by: Web3 Farmer Frank
AI Agents are in the '2020-2021' boom cycle of DeFi Summer, and Infra projects may be key to capturing Beta dividends.
Is the AI Agent currently in the '2020-2021' boom cycle of DeFi Summer?
"Carving a boat to seek a sword" is often the most worry-free prediction method—looking solely from the data dimension, the total locked value (TVL) of DeFi grew from 600 million USD in January 2020 to 26 billion USD in December 2020, increasing more than 40 times in just one year. The current market value of the AI sector is 44 billion USD, which also means that if it can truly replicate the boom cycle of DeFi, its potential market value may exceed 1 trillion USD.
So the question arises: Are AI Agents really experiencing a boom cycle similar to the '2020-2021' DeFi Summer? Aside from AI Memecoins, what other projects will drive the momentum of this sector's boom cycle?
The booming AI Agent boom cycle
According to statistics from Cookie.fun, as of January 1, 2025, the overall market value of the AI Agent sector has exceeded 16.6 billion USD, with a remarkable increase of 53.79% in the past 7 days. Against the backdrop of market fluctuations in the entire cryptocurrency market, the ecosystem of AI Agents is expanding at an unprecedented speed, from Virtual to AI16Z, and then to platforms like Arc.
Especially AI Agent projects driven by Meme culture, with their low threshold for issuing tokens and transaction processes, have quickly become new centers of traffic and attention. Most of them significantly lowered the participation threshold for ordinary users by using humorous expressions, highly interactive community cultures, and user-friendly dissemination strategies, successfully attracting a large number of non-professional users.
To put it bluntly, this is actually another form of cultural rebellion against 'VC coins', and is a continuation of the Meme narrative. However, combined with AI Agents, it not only lowers the participation threshold and expands the potential market size but also injects more inclusivity and sustainability into the AI × Crypto narrative.
Including the recently outstanding performer 'TAO CAT', as a product of the Masa team's collision of the two major ecosystems of Virtuals and Bittensor, it cleverly demonstrated the superiority of AI Agents over traditional AI Bots in self-evolution, with the core still being fair issuance and community-driven economic narratives.
Interestingly, during the process of writing this article, TAOCAT was selected for the new round of Binance Alpha projects, followed by a significant rise in its price, and MASA also experienced a rising trend as a result.
This easily evokes the prosperous scene of DeFi Summer: from the initial foundational applications of Compound and Uniswap to the subsequent massive influx of users and the derivation of more complex on-chain trading strategies, the interweaving and resonance of technological narratives and capital drives ultimately form periodic prosperity. Thus, at least currently, the popularity and market value growth of AI Agents based on AI Memecoins indeed brings a new round of narrative and imaginative space to the cryptocurrency industry.
However, how far this boom cycle can go depends not just on AI Memecoins as the starting point but more on whether AI Agents can become a long-term cornerstone of the cryptocurrency industry rather than just a short-term trend.
For instance, in the current AI Agent landscape, the total market capitalization of the top 5 exceeds 7 billion USD, accounting for nearly 45%, with the market capitalizations of the Virtual ecosystem and AI16Z ecosystem reaching 4.65 billion USD and 2.37 billion USD, respectively, thus becoming industry benchmark cases in the entire cryptocurrency field.
However, in essence, the current prosperity of the 'AI Agent' sector is still primarily dominated by AI Memecoins or issuance platforms, remaining limited to the 'equitable' economic narrative of fair asset issuance. It needs to further deepen its outreach. Looking back at the previous crypto cycle, we can find that while DeFi has driven a massive influx of funds into the blockchain field, the user threshold for DeFi remains relatively high, requiring financial knowledge and operational skills, which somewhat limits its large-scale popularity.
The emergence of AI Agents essentially opens a new narrative in the cryptocurrency field with lower thresholds and higher potential—unlike DeFi, which requires users to actively learn complex financial tools, AI Agents can directly provide users with personalized and intelligent services, thus lowering barriers to participation.
These are some interesting variables, namely meeting more ordinary users' needs through intelligent solutions, allowing AI Agents to be not just a technical product but a brand new 'equitable' economic narrative.
As the founder of Slow Mist, Yuxian, tweeted: "Another special feeling of disconnection: the ultimate (perhaps) most important role of AI Agents in the Crypto industry is to issue a token to stimulate those incentives. In the non-Crypto industry, major players have long been far ahead, including some interoperability protocols they defined, creating increasingly user-friendly AI applications, while most attention in the Crypto industry is still on tokens."
Thus, AI Agents have indeed entered a boom cycle due to AI Memecoins, gaining high recognition and funding attention in the market, marking their official entry into a golden age: subsequently, from Meme-driven traffic centers to diversified application scenarios, from the improvement of infrastructure to the long-term value excavation, AI Agents are expected to reshape the landscape of the cryptocurrency industry, becoming the most innovative and impactful field within the cryptocurrency space.
The 'Alpha' and 'Beta' of AI Agents
If we trace back to the root, we will find that while the current AI Agents are merely a value capture of token issuance and imagination, the real landing is still closely related to on-chain trading attributes.
After all, the practical use of AI Agents goes far beyond just issuing tokens—early in the task-driven AgentGPT handling process, AI Agents have already demonstrated their potential in on-chain trading strategies.
Users can customize their overall goal task, and then the AI Agent will allocate tasks according to its resources, breaking down the overall goal task into smaller tasks through the flow and allocation of tasks at the agent layer, continuously executing and correcting according to the goals without human assistance, ultimately achieving the overall goal.
In other words, the core vision is 'intent-centered', where users only need to know what they want to achieve (intent) without having to worry about the processes and steps in between. In many cases, the user's intent is complex and may not be accurately expressed. However, AI trained with specific models can more accurately identify user intent and infer the user's potential goals and needs based on their keyword information.
In this process, combined with highly intelligent AI, whether it is natural language input of intent, judging needs, breaking down goals, calculating optimal process combinations, or executing operations, AI can play a significant advantage. Currently, there is already a surge of intelligent trading projects involved in complex cross-chain trading in the market, which can efficiently discover the best exchange paths, allowing users to complete optimal trades in real-time at the best prices.
This is the truly imaginative direction for the development of AI Agents—where real value lies in becoming the interactive application layer of Web3, allowing users to complete all on-chain operations simply by conversing with AI Agents, thus potentially becoming the underlying infrastructure of Web3, deeply integrating with blockchain, smart contracts, etc., giving birth to new application forms and business models, and bringing broader imaginative space to Web3.
So from a macro perspective, the rapidly developing AI × Crypto narrative is completing self-iteration at an astonishing speed—from the initial AI Meme craze to the gradual shift towards practical and interactive AI Agents. The entire market's attention is refocusing, which also implies a transition in narrative logic from random token speculation to actual technological implementation.
AI Agents are no longer just a hype but have truly integrated into the industrial logic through value capture mechanisms, such as AI-driven KOLs, yield-generating tools, trading and Alpha strategy optimization, as well as the construction of underlying infrastructure. AI technology is gradually shifting from an entertainment narrative to a tool for solving practical problems.
Its value goes far beyond the currently popular AI Memecoin; through technological innovation and ecological integration, it truly promotes the evolution and popularization of Web3. From this perspective, AI Agents are not merely tools; they may reshape the ecological rules of the cryptocurrency market, making AI Agents more easily accepted and adopted by mainstream users.
Any 'long slope with thick snow' track will never lack infrastructure
In the cryptocurrency industry, Alpha is a point, Beta is a plane. Alpha is usually difficult to capture, but Beta projects with a high degree of certainty are relatively easier to ambush.
The AI Agent field is no exception. Although it is difficult to grasp Alpha opportunities like AI Memecoins, infrastructure (Infra) as the core of Beta presents an almost clear value capture opportunity, with a narrative transmission path that is extremely clear—despite the explosion of AI Memecoins stirring up waves of funding and market attention, the final landscape of projects that can emerge is far from certain, while Infra, as the 'water seller', can almost secure a prime fishing spot thanks to its early ecological advantages.
As mentioned above, from the perspective of AI's role shift within the blockchain ecosystem, it is evolving from a mere tool to a productive force driving industry transformation. This not only serves existing cryptocurrency users but may also attract traditional financial users to enter the on-chain ecosystem by lowering barriers, constructing a more inclusive and sustainable economic logic:
Shifting from elite finance that serves a few to inclusive finance that serves the majority; moving from the unsustainability of short-term speculation to the sustainability of long-term growth.
This suggests that AI Agent Infra is likely to become a key entry point for on-chain application intelligence. It also implies that projects in the same track that have not been fully priced by the market may conceal new value reassessment opportunities, one important direction being data.
As we all know, one significant challenge faced by AI Agents is how to efficiently mine and utilize high-quality data. This not only determines the learning ability and decision-making level of AI Agents but also directly affects the effectiveness of their practical applications and user experiences.
Especially for the blockchain ecosystem, Twitter data, Discord data, web crawler data, etc., may hide billions of signals, containing many potential use cases that can be derived and realized, and can be personalized for mining and utilization—such as tracking whale movements, analyzing 'Smart Money' transactions, monitoring on-chain address interactions, filtering layered differentiated on-chain users, accurately reaching different player groups, and so on.
Therefore, data infrastructure is not only the support for the development of AI Agents but also the core of their value creation and market promotion. Interestingly, as the current leader in the AI data sector, recently, Masa's construction of the SN42 data service subnet and SN59 AI Agent Arena on Bittensor deserves mention.
The AI Agent Arena is a significant step that Masa, as a data infra, is taking towards the AI Agent platform, making the data network no longer abstract but capable of empowering AI Agents in real-time, allowing them to evolve in real-time and iterating Bittensor's powerful reward mechanism, enabling AI Agents to compete fiercely in the arena. The arena has attracted leading AI Agents from the Virtuals and creator.bid ecosystems within just a few weeks, including TAO CAT incubated by Masa.
Therefore, in essence, Masa is centered around 'infrastructure (data) + application scenarios (arena) + token economy (AI Agent)', attracting more users and developers to operate the platform, continuously expanding the AI Agent business scenarios and stimulating platform activity.
From a broader perspective, the narrative path of AI Agent Infra like Masa is very clear—focusing on the 'adaptability' issue of AI agents on-chain, addressing the efficiency, stability, and intelligence needs of on-chain operations. If this track continues to ferment, it not only holds the potential to build the infrastructure for on-chain AI ecology, but may also drive a comprehensive upgrade of the developer ecosystem and user experience.
According to market rotation patterns, the success of Virtual will also drive investors to focus on complementary functions, technological innovations, or undervalued projects in the same track, especially those with strong ecological building capabilities and clear narrative directions. Therefore, Infra projects like Masa undoubtedly form a natural complementary logic with the token issuance function of Virtual.
The improvement of infrastructure and the maturity of technology have enabled AI Agents to move from the uncertain Alpha stage to the profitable, scalable Beta stage. Therefore, from this perspective, Infra projects like Masa are indispensable drivers in this process.
The aforementioned TAOCAT's selection for the new round of Binance Alpha projects also adds real-world evidence to my analysis logic, further strengthening the notion that as market attention gradually shifts towards the AI Agent sector, infrastructure projects (like MASA) may encounter unexpected opportunities for value growth.
Overall, this round of enthusiasm for AI Agent Infra is just the beginning. The market's demand for on-chain intelligent infrastructure is still rapidly expanding. The performances of Virtual and Masa provide us with a good reference. More projects with clear positioning, complementary functions, and valuations not fully explored may gradually be discovered by the market, and opportunities for value reassessment often hide in the shadows of niche areas.
Conclusion
AI Agents are not the endpoint but the starting point: in the future cryptocurrency market, AI Agents will not only be a component of the ecosystem but will also become an important driving force propelling the entire ecosystem forward.
Infrastructure is the foundation of the ecosystem; no major trend can lack its support. Whether it is pioneers like Virtual or potential players like Masa, their commonality is embedding their value into the core of the entire ecosystem through technology and logic, and the market's recognition of this value is only a matter of time.
