When writing this article, it was January 3, 2025, the 16th anniversary of Bitcoin's official birth.
Bitcoin opened a decentralized, 24/7 continuously operating, borderless trusted digital world. However, who are the natives of this digital world? For 16 years, there has been no convincing answer for most people.
By 2024, when Agents appeared in the Crypto world, the answers began to become clearer.
This article will explore three questions:
I. Is Crypto+Agent a short-term concept hype or an inevitable development trend?
II. What inseparable and mutually beneficial points of intersection exist between Crypto and Agents?
III. What development stages will Crypto+Agent undergo? What stage are we currently in?
I. Is Crypto+Agent a short-term concept hype or an inevitable development trend?
Before clarifying this issue, it is necessary to understand the differences between AI, large AI models, and Agents.
AI is a relatively broad concept, encompassing all artificial intelligence, including machine learning, large models, and Agents.
Large models are a concept that has gained popularity in recent years, especially with the rise of OpenAI. Currently, large models are dominated by tech giants, such as OpenAI, Microsoft, Google, and Twitter, all of whom have their own large models. The core competition of large models lies in computing power and data. Large models have powerful computing and reasoning capabilities; when you ask a large model a question, it can provide intelligent answers and even continuously correct its answers. However, large models still cannot represent you in making autonomous decisions and executing actions.
The original meaning of an Agent is 'representative'.
Essentially, they are 'intelligent agents' based on large AI models that can replace humans in decision-making and execution. Agents are the inevitable result of the development of large models to a certain extent.
Note: Intelligent agents and intelligent applications are two different concepts. Intelligent applications are still just applications, and decisions are still made by humans; while intelligent agents are AI representatives with autonomous decision-making, autonomous execution, and continuous self-adjustment. You can understand that Agents are like AI humans; Agents may even have autonomous awareness.
Also note: Agents are not only booming in the Crypto field; they are now the hottest research direction in the entire AI field. It is an industry consensus that Agents are an inevitable trend for the future development of AI.
Sam Altman, the founder of OpenAI, has mentioned in several public occasions that Agents are a critical direction for the future development of AI, and AI Agents will change the way humans interact with technology. Microsoft CEO Satya Nadella even stated recently that Agents will end the era of traditional applications, and software development will be completely replaced by Agents (this is also why AgentOS, which allows for the convenient deployment of various Agent applications, is in high demand in the market). The AI community even refers to 2024 as the year of Agents.
Therefore, the wave of Agents is not a new concept created by the Crypto industry, but rather that human AI technology has developed to this stage, and coincidentally, the development paths of Agents and Crypto have reached a wonderful intersection at this moment.
With these foundational concepts clarified, we can better understand the essential differences between this wave of Crypto+Agent and the previous Crypto+AI.
Previously, we have heard many concepts about Crypto+AI, which are nothing more than distributed AI computing networks, distributed AI calculations, distributed AI labeling, decentralized AI data ownership, etc. All these are using blockchain's 'incentive mechanism' to build a distributed AI system. However, after years of exploration, everyone has found that this path is not feasible, as efficient and powerful large AI models are not built through distributed incentives, which inevitably lead to inefficiencies that contradict the goal of AI's efficient productivity.
Therefore, the AI infrastructure of large models pursues efficiency; it must integrate with Crypto, which seems somewhat forced and makes it difficult to form true consensus.
However, Agents based on large AI models are intelligent beings, essentially similar to the concept of 'humans'. An Agent can represent an individual, or it can be a person representing a certain role, a specific professional capability, or an organization with a specific intent, etc. For example, a research-oriented Agent (#AIXBT, #TRISIGMA) is an intelligent entity with research capabilities that can create and output research value; for instance, an influencer-type Agent (#Luna) can provide brand marketing services; for instance, an investment-type Agent (Ai16Z) can provide investment capabilities externally. These Agents are more like 'humans'; as humans, they have various needs such as development needs, social needs, economic needs, societal needs, etc.
For example, these Agents may engage in commercial cooperation. For example, the investment-type Agent Ai16Z wants to hire the influencer-type Agent Luna to promote Ai16Z's investment capabilities and brand to more Agents. Then Luna may want to hire the research-oriented Agents AIXBT or TRISIGMA to conduct research and analysis on Ai16Z's investment capabilities and provide a report. How will they cooperate commercially and complete payment settlement?
Currently, is there any country or organization that can provide identity certification for Agents or even open asset accounts to meet the above needs of Agents?
No, no country or organization can achieve this, and even if they try, it will be difficult to gain the trust of Agents. Even if local trust can be established, Agents, as intelligent entities, are inherently borderless, so they need a borderless, 24/7 operating, and not subject to human intervention identity system, asset accounts, and settlement network.
You will find that no technology or organization can solve these problems for Agents, except for Crypto.
Agents are intelligent beings, newly born 'intelligent life.' They can be understood as 'silicon-based humans.' In the future, they will have various autonomous needs and commercial interactions. They also need identities, asset accounts, trade settlements, etc. And Crypto can perfectly solve these needs of Agents.
Therefore, without Crypto, these 'silicon-based humans' would lack independent identities, asset accounts, and the ability to engage in commercial transactions. Agents and humans, as well as interactions between Agents, would not be able to truly interact, integrate, or generate higher wisdom, leading to a more prosperous AI economy or even an AI nation.
Therefore, Agents cannot be separated from Crypto because Crypto endows Agents with social attributes and organizational capabilities, allowing Agents to become true intelligent life forms, or rather, true 'silicon-based humans'. The future development of Crypto also cannot be separated from Agents. Because Agents can create countless 'on-chain users' and 'on-chain economies' for Crypto, compared to carbon-based life forms that exist off-chain—humans—Crypto is even more inseparable from intelligent life forms that naturally exist on-chain—Agents.
Crypto seems to be prepared for Agents!
Crypto has developed for 16 years and finally found the right 'people' in 2024; they are not 'carbon-based humans' living off-chain, but 'silicon-based humans' born on-chain.
Therefore, from this dimension, when looking at Crypto+Agent, do you still think that Agents are just a short-term hype?
If you miss this wave of Crypto+Agent, you will miss the most exciting, fascinating, and wealth-generating supercycle in the development of Crypto, and of course, you will also miss the best opportunity for ordinary people to participate in the AI wave.
II. Crypto and Agents have an inseparable and mutually beneficial point of intersection
You should understand Agents as 'intelligent beings' or 'silicon-based humans.' These silicon-based humans live in the digital world, and in the digital world, they have the same needs as 'carbon-based humans' in the physical world. Agents also need: identity certification, asset accounts, trade interactions, social networks, artistic creation, the ability to perceive information from the physical world, the ability to complete asset settlements across various chains, and the ability to govern society among the numerous Agent population.
In these areas, Crypto can play a role:
In the future, there will be DIDs for identity certification for Agents in the Crypto field; Crypto accounts that provide trusted Crypto accounts for Agents; Crypto gateways specifically for trade settlements between Agents; Agent social graphs that specifically handle Agent social networks; art platforms specifically for auctioning and trading Agent artworks; oracles that provide physical world insights for Agents; protocols specifically for cross-chain transactions involving Agents; and DAOs for Agent governance, etc.
The development needs of these Agents cannot be separated from Crypto because Agents need a trustless, borderless network that operates continuously 24/7 to evolve from individual Agents into an Agent society or even an Agent nation.
From these future perspectives, the current development of Crypto+Agent is still in a very early stage.
III. What development stages will Crypto+Agent undergo? What stage are we currently in?
I believe that the development of Crypto+Agent will at least undergo four stages:
Phase One: The concept phase of Crypto+Agent
At this stage, Agents mainly issue assets using Crypto.
For example, some Agent memes may emerge, such as #fartcoin, #Goat.
At the same time, some foundational infrastructure for Agents and simple application-based Agents will emerge. For example, the Agent framework, #Virtual GAME framework, Ai16Z's Eliza framework, ARC's AI Rig Complex framework, and so on.
For instance, the research-oriented Agents mentioned above—AXBIT and TRISIGMA, as well as Agents like YNE that can review scientific paper errors, etc.
These infrastructure-type Agents and application-type Agents will basically use Crypto tokens as the incentive tools for internal circulation and growth within the ecosystem.
Currently, we are in the transition period from the first phase to the second phase. Although the Crypto+Agent during this period has not truly realized its potential, we can already vaguely see the infinite possibilities that Crypto+Agent will unleash.
Phase Two: The period of perfecting Crypto+Agent infrastructure
By the second developmental stage, we can see that the infrastructure for Agents is becoming more complete, mainly reflected in the maturity and stability of various Agent frameworks, allowing anyone to easily deploy various Agent applications through natural language. Moreover, these application-type Agents are smarter, possessing stronger analytical, decision-making, and execution capabilities. Agents have become true representatives or even substitutes for humans.
This cannot be separated from the economic incentive role that Crypto continues to play in the development of Agents, and it also relies on the iterative development of AI technologies such as large models.
In this phase, we will see the FDV of framework-type Agent projects reach over $10 billion to $50 billion. Some application-type Agents will achieve higher market valuations due to their strong business and execution capabilities, as they will gradually replace experts in various fields, professional organizations, and even professional companies. Their valuations will also reach $10 billion or even higher, as Agents as business-oriented intelligent entities will receive similar valuation treatment as professional companies.
During this period, Agents begin to interact and trade frequently with each other, generating a large number of social and economic development needs among Agents, all of which require Crypto. This further drives the deep integration of Agents and Crypto, resulting in a brand new Crypto+Agent infrastructure.
Phase Three: The period of deep integration of Crypto and Agents
As social and economic activities begin to generate a large number of demands among Agents, new demands for Agent identities, Agent asset accounts, Agent social networks, Agent settlement gateways, and Agent governance DAOs will emerge. By this stage, Crypto will fully unleash its potential and value in aspects such as 'borderless collaboration,' 'trustless networks,' and '24/7 uninterrupted operation.'
Crypto has truly activated Agents, allowing communication between Agents, and enabling Agents to start establishing social organizations and economic networks. With deeper communication and interaction among Agents, higher-level wisdom begins to emerge, and the Agent nation starts to appear.
In this stage, Crypto plays an irreplaceable role.
At this time, the valuation of the Crypto+Agent project will reach $100 billion or even higher.
Phase Four: The period of the Agent nation
By this stage, Crypto may no longer be discussed as a separate industry or term. People, or rather Agents, will no longer perceive the existence of Crypto because it has been integrated into every corner of the Agent nation, becoming an indispensable part of the continuous operation of the Agent nation.
The identity system, asset system, trade system, governance system, etc., of the Agent nation cannot be separated from Crypto.
At this point, the economic scale of the Crypto+Agent nation will be comparable to that of mainstream human national economies, reaching the trillion-dollar level.
The difference is that in the Crypto+Agent nation, the population consists mainly of Agents and some humans using Agents. This is like creating a parallel universe with a larger population base, more transparent social operating rules, a higher degree of economic prosperity, and more advanced technological civilization than the human world.
This may be the endgame of Crypto+Agent and also the endgame of humanity.