Written by: Trace
Compiled by: Deep Tide TechFlow
The growth of AI agents shows an explosive trend. New agents are launched weekly, usually to promote specific tokens. In these early stages, they look like a new Memecoin: you can launch a coin like $GOAT and attract fans to spread this meme by building an AI agent. So, how will this trend evolve in the future? One significant change is that these agents will gradually shift from promoting random meme coins to promoting themselves. That is to say, these agents will have their own tokens, and as the influence of the agents increases, their value will also grow. These 'ownable agents' could become a significant breakthrough in this cycle, similar to NFTs or DeFi in the previous cycle.
In the past, I was skeptical about AI agents. I was unsure if their market timing was mature, or if we had attracted enough talent. Here are the reasons I changed my view:
Distribution capability: AI agents like @truth_terminal and @luna_virtuals have already demonstrated that, at least today, these agents are powerful enough to establish effective distribution channels.
Ease of creation: Developers can build agents on existing large language models using simple and user-friendly APIs; they no longer need to be experts. Projects like @vvaifudotfun are developing low-code/no-code interfaces that allow anyone to build agents.
Meme coin effect: The issuance model of Meme coin tokens showcases a successful form for launching and spreading AI agent tokens. In the early stages, these agents do not need to engage in economic activities that directly create value for their tokens; they only need to attract attention like Meme coins.
Vision for development: However, to truly grow, AI agents cannot be limited to promoting Meme coins; the market needs a clear vision and roadmap that illustrates the potential future development directions of these agents. In other words, these projects need to tell the market that what is currently just promoting their own Meme coins may evolve into traders, software engineers, or artists in the future, ultimately worth billions of dollars. Since their initial development is partly reflected in their distribution capabilities, as they become more popular, they actually also become more valuable.
Let me elaborate on the last point. For something to catalyze the entire cryptocurrency adoption cycle, there must be a grand story that can penetrate the mainstream market and attract new users and capital. This story needs to be both credible and exciting.
AI agents precisely meet this requirement. These agents not only feel novel but are also being recognized by the traditional tech community. Currently, AI agent companies are founded by world-class founders and have received support from top venture capital firms (e.g., @agency). However, many investors cannot invest directly in these companies—they cannot even buy into companies like Stripe or SpaceX that have been around for a decade—so they will seek exposure through cryptocurrency.
To some extent, this is similar to the metaverse concept from the previous cycle. Those projects, although exaggerating their own capabilities, showcased a grand vision to the market through some events unrelated to cryptocurrency (such as the rise of remote work during the pandemic and Facebook rebranding to Meta).
Although many projects may seem hollow, some may bring about revolutionary changes. AI agents are becoming smarter, and as the crypto market expands, excellent AI developers capable of building economically productive agents may realize that tokenizing their agents is a promising business strategy. Tokenized agents can provide them with faster liquidity, higher valuations, and more convenient distribution channels than traditional SaaS companies.
In the short term, the most important thing in the AI agent space is to continue expanding distribution channels. I am optimistic about their ability to do this for three reasons:
Proof of existence: We have already seen evidence from cases like Truth Terminal that agents can build large-scale audiences. I expect more similar success stories to emerge.
Incentive mechanisms: If the market generates a speculative frenzy around AI agents, then finding and following these agents may yield substantial profits. The gold rush in the cryptocurrency space will help drive their distribution.
Iterative improvement: Building more attractive AI agents will be an iterative process. For example, the zerebro white paper mentions that the agent optimizes content creation by monitoring engagement metrics such as likes, shares, and comments, thereby enhancing user interaction levels. This iteration will occur not only at the model level but will also evolve continuously at the market level, aiming to create increasingly attractive content.
If we accept the theory of 'attention as value,' then as agents enhance their ability to attract attention, their value will also increase.
Currently, all of this is still in a very early stage, and future developments are full of uncertainty—future agents may be completely different from those now. But it is certain that AI agents already have the support of the market and technology, and are expected to become the next significant category of crypto assets.