Article reproduced from: BlockBooster
Author: Kevin, the Researcher at BlockBooster
The term AI agents comes from OpenAI's roadmap. Sam Altman divides the capabilities that AI should possess into five parts, with the third step being AI agents that will be frequently encountered in the coming years.
AI agents can autonomously learn, make decisions, and execute tasks. Depending on their intelligence and capabilities, Stuart Russell and Peter Norvig classify AI agents into five categories in their book (Artificial Intelligence: A Modern Approach):
Simple Reflex Agents: Respond only to the current state.
Model-Based Reflex Agents: Consider historical states in decision-making.
Goal-Based Agents: Focus on planning and finding the best path to achieve specific goals.
Utility-Based Agents: Aimed at weighing benefits and risks to maximize utility.
Learning Agents: Continuously learn and improve through experience.
So what level do the AI agents appearing in the current market or industry belong to? What direction are they heading?
OpenAI's o1 has reached Level 2 artificial intelligence. I personally believe that the current AI agents in the industry are at Level 2.5, between Level 2 and Level 3. This is not to say that the agents in the industry have surpassed OpenAI; in fact, web3 agents are still in the GPT wrapper stage. So why Level 2.5? Because through human or programmatic intervention, let's call it mediation, the combination of the GPT wrapper and mediation forms a type that is not thoroughly substantiated but has objective proactivity. It is an extension of a certain direction of the OpenAI model application. In terms of what agents can do, they are at the most basic level of simple reflex agents. Some of these agents consider historical states, but they require active input. Only by continuously feeding data can agents complete their learning, which is a passive model training approach and far from reaching the state defined by Level 3. The latter three types—Goal-Based, Utility-Based, and Learning Agents—have not yet entered the market. Therefore, I believe that current AI agents are still in the early stages, fine-tuning Level 2 general LLMs, and structurally have not departed from Level 2. So, can evolution to Level 3 be achieved solely through crypto? Or does it require waiting for companies like OpenAI to develop?
Why is there a discussion about Base or Solana becoming the narrative center for AI agents?
Before discussing which ecosystem can promote the emergence of Level 3 agents, we should determine which ecosystem has the potential to become fertile ground for AI agents. Is it Base? Or Solana?
To answer this question, it may be useful to first reflect on how AI has influenced Web3 in the past two years. When OpenAI first released ChatGPT, industry protocols were still following habitual thinking, quickly rushing into the infrastructure bubble. This led to a surge of computing power/inference aggregation platforms, along with the birth of AI + DePIN infrastructure. The commonality between the two is the establishment of a grand vision, and it is not to say that a grand vision is bad; in fact, agents can also construct such visions, but in terms of implementation and user demand, such large infrastructure protocols have not fully considered these aspects. The market demand they wish to stimulate is still far from saturated in the traditional internet industry, and user education and market education are insufficient. Under the impact of the Memecoin craze, the AI infrastructure seems even more hollow.
Since infrastructure is too heavy and large, why not lighten it? Agents derived from the GPT wrapper are efficient and iterate quickly in terms of launch and user reach. Lightweight agents have sufficient potential to create bubbles, and after the bubble bursts, fertile ground for new birth will emerge.
Furthermore, in the current market environment, using agents and Memecoins to launch projects can bring products to market in a very short time. This allows users to directly experience the product. In this process, agents can also leverage Memecoins to expand the community roadmap, achieve rapid product iteration, and this iteration is low-cost and fast. Serious AI protocols no longer need to be constrained by heavy old consensus frameworks; they can break free from their cages, charge into the fray lightly, and bombard users with lightweight and rapid iterations. After sufficient market education and dissemination, they can then build on this foundation and lay the groundwork for grand visions. Lightweight agents cover the ambiguous Memecoin veil, and community culture and fundamentals will no longer become contradictions. A new path for asset development is gradually emerging, which could be a potential path for future AI protocol choices.
The above discussion answers the potential of AI agents to become the core narrative. Under the premise that AI agents can continue to grow rapidly, choosing the right ecosystem becomes particularly important. Is it Base? Or Solana? Before answering this question, it may be worthwhile to look at the current state of serious agent protocols in the market.
First, Arweave/AO: PermaDAO mentioned that AO adopts the Actor model for design, where each component is an independent autonomous agent capable of parallel computing, which aligns highly with the architecture of AI Agent-driven applications. AI relies on three elements: models, algorithms, and computing power, and AO can meet such high resource demands. AO can independently allocate computing resources for each agent process, effectively eliminating computing performance bottlenecks.
In addition, Spectral is one of the few protocols based on agents, with document-to-code transformation and model inference as its development direction.
Looking back at a type of agent token currently in the market, it can be found that these agents hardly utilize the underlying infrastructure of the chain. This is a fact because all models in the industry, including agents, are off-chain. Data feeding is off-chain, model training is not decentralized, and the information output is also not on-chain. This is an objective fact because the EVM chain does not support the combination of AI and smart contracts, and naturally, Base and Solana do not support it either. Next year, the introduction of AO can be anticipated; whether it can allow models to go on-chain and perform well. If AO fails, the model going on-chain may have to wait many years after Ethereum is introduced, at least not before 2030, or other public chains to achieve model on-chain. However, if a structure and historical resource reserve like AO cannot be realized, then achieving model on-chain may be even more difficult for other public chains.
Currently, AI agent tokens do not have many practical use cases. In fact, it is difficult to clearly distinguish between AI agent coins on Base and Solana and AI Memecoins. Although agent tokens do not have special uses, why do I believe that AI agent coins and AI Memecoins should not be confused? Because I think we are currently in the stage of creating an AI agent bubble.
Why is there a discussion about Base wanting to compete with Solana for the dominant public chain position of AI agents?
Base attracted significant market attention in the first half of this bull market, with a brief impressive performance in the battle for market share of Memecoins, such as $BRETT and $DEGEN. However, it still lost to Solana. I believe AI agents are the next area for Base to compete in, and it already has several advantages.
AI agents will accelerate the birth of bubbles and create chaos, but ultimately will leave behind users and applications:
The birth and expansion of bubbles will attract market attention, and this attention will undergo a qualitative change over time. What characteristics does such a qualitative change have? In the process of increasing market attention, a series of user pain points and market gaps will be exposed. When the main contradictions cannot be coordinated, but attention continues to increase, that is the moment when qualitative changes are born. When the qualitative change is completed, the accumulated users and applications can carry the grand vision. This is something Memecoins cannot and do not intend to achieve, which is why I believe that although agents and Memecoins are currently blurred, they should never be confused.
Before a qualitative change occurs, bubbles will generate a mess and various dramas, such as: the number of agents will increase exponentially, and thousands of agents will crowd into users' sight. How do they crowd in? Agents can connect to social media like X and Farcaster to self-promote their tokens, using various angles appealing to degen and the unique information density of agents to market their tokens.
Subsequently, rapidly iterating agents can complete on-chain transactions. A group of Viking pirates has invaded the dark forest. Current panel protocols, bots in TG groups, and Dune panels will be invaded by agents. Familiar indicators for users will be manipulated by agents, such as trading volume, address numbers, chip distribution, and simulating dealer behavior. On-chain data may need more professional cleaning to reflect value; otherwise, it will be deceived by agents, just like Viking pirates plundering your wealth.
If the market can reach this stage, then the new era belonging to AI agents would be halfway successful, because 'attention equals value' will allow agents to enter mainstream spaces. This potential comes from:
Powerful Distribution Capability: Agents generate sufficient discussion, such as Goat, and stable distribution paths can be replicated.
Ease of Deployment: The deployment platforms for agents will also experience explosive growth. Zerebro, vvaifu, Dolion, Griffain, and Virtual will allow users to build agents without needing to know any code, and the UX of agent deployment platforms will also be optimized in competition.
Memecoin effect: In the initial stage, agent tokens lacked suitable business models, and the token use cases were minimal. By donning the Memecoin guise, they could quickly accumulate community support, keeping the startup success rate efficient.
The upper limit is extremely high: OpenAI's Level 3 agent is still in development. A product that major players cannot quickly release will inevitably have a huge market space. The lower limit of agents is Memecoins, but the upper limit is highly autonomous advanced intelligent agents.
Low market resistance: Agents led by Goat can establish a large audience. Unlike AI infrastructure, users are not averse to agents; when users are not averse, there is a strong possibility of beginning to pay attention to them.
Potential Incentives: The use cases for agent tokens have not yet been developed. If agents introduce a point system and strengthen incentive mechanisms, they could accumulate a large user base.
Iterative Potential: As mentioned earlier, agents are lightweight and capable of rapid iteration. This objective iterative capability can create increasingly attractive products and content for users.
Therefore, AI agents can become the core narrative and are a battleground for competition.
Why does Base have the potential to compete with Solana?
With strong support from Coinbase and North American capital, the Base ecosystem experienced explosive growth in 2024. In November, capital inflow exceeded that of Solana and significantly surpassed Solana in the past seven days.
If ETH can continue to break through the ETH/BTC exchange rate next year, the spillover effect of the ETH season will have a significant impact on Base. Currently, 23% of the funds flowing out of ETH are directed toward Base, and this figure is still rising.
AI agent Launchpad Mapping
Virtual
The V1 phase mainly focuses on model training, data contribution, and interaction features, while in the V2 phase, Virtual launched a token incubation platform for AI agents, with a landmark update in October being fun.virtuals.
Among them, LUNA has developed into an 'independent entity' with its own identity and financial capabilities. Throughout this process, LUNA's roadmap aligns with that of Coinbase, which provides powerful technical tools and support to facilitate the implementation of AI agents on Base.
AI agent technology shows great promise in brand building, especially in creating cultural brands. Through AI agents, brands can interact more efficiently with communities. This includes simplifying interaction tasks and flexibly distributing rewards, enhancing user stickiness and brand recognition.
It is worth noting that all transactions of AI agents only support the use of native Virtual tokens. Virtual tokens absorb the value capture of the entire ecosystem and become an important pillar of ecological development.
Virtual focuses on perfecting product functionality, empowering users with AI tools, and building a bridge between Web2 and Web3. It emphasizes 'utility value' rather than 'hype'. Although its tool-based products are frequently used in practice, they lack the dissemination effect that cryptocurrencies usually possess, which is also a shortcoming of the V1 phase.
Clanker
'Posting equals issuing tokens' lowers the threshold for token issuance and attracts a large number of users to try it out. People rush to @Clanker, a phenomenon similar to the operation of having AI summarize video content on social media; however, here, content publication directly translates into asset issuance.
How does Clanker operate?
TokenBot (i.e., Clanker) will deploy Meme tokens on Base into unilateral liquidity pools (LP), and liquidity will be locked immediately. Token issuers will receive the following benefits:
0.25% of all Swap fees.
1% of the total supply of tokens (unlock period of one month).
Users can check the number of tokens deployed or create their own tokens through the clanker.world official website.
Unlike PumpFun, which issues tokens on Raydium through bonding curves and charges a 1% transaction fee along with a fixed fee of 2 SOL, Clanker does not adopt a bonding curve model, instead generating revenue through a 1% fee charged for transactions on Uni v3.
AI Agent Layer
AI Agent Layer is a platform focused on creating AI agents and Launchpads within the Base ecosystem, officially launched on November 18. Before the platform's launch, the AIFUN Token was first issued on November 14, and is now listed on exchanges such as MEXC and Gate, with a current price of $0.09 and a market cap of approximately $25 million.
Creator.bid
Creator.bid was initially an AI platform focused on the monetization and ownership of digital content. In April of this year, the platform completed a new round of financing.
On October 21, Creator.bid announced the official launch on the Base mainnet, enabling one-click creation and publication of AI agents, providing content creators with new tools and profit models.
Simulacrum
Simulacrum is built on Empyreal. It transforms platforms like Twitter, Farcaster, Reddit, and TikTok into a blockchain interaction layer. Users can perform on-chain operations, such as token trading or tipping, through simple social media posts.
Using techniques like account abstraction, AI agents, intent-driven and language models to simplify complex blockchain backend operations. Making DeFi easier for ordinary users.
vvaifu.fun
Similar to Pump.fun, users can easily create AI agents and their associated tokens. AI agents can seamlessly integrate with social platforms such as Twitter, Telegram, and Discord to achieve automated user interaction.
Dasha is an AI agent created by vvaifu.fun, possessing an independent Twitter account, Telegram channel, and Discord community, all operated and managed by AI.
Top Hat
Top Hat can not only interact with users through text but also understand and process image content. After a user sends an image, the AI agent can 'understand' the content of the image and respond.
Griffain
With a platform for trainable AI agents, Griffain has launched 1,000 trainable AI agents, showcasing the future potential of smart contracts and automated trading.
