Editor's Note: This article points out the current fragmentation problem of AI agents, the diversity of AI agents, and the dispersion of liquidity and attention across different fields. The article mentions two development paths for agents: one is to transform into an open-source framework, and the other is to continue to focus on a single agent function. It analyzes several potential AI fields, such as DeFAI, NSFWagent, robotics, and investment DAOs, and suggests that developers and investors should focus on narratives and products with long-term growth potential, ultimately driving innovators to establish a unified ecosystem by simplifying complexity.
The following is the original content (edited for readability):
Liquidity fragmentation became a major challenge last year, as the expansion of L1/L2 blockchains exceeded our ability to keep track. Then, Bitcoin L2 entered the picture, further adding thousands of new projects.
This has resulted in liquidity, attention, and user experience being dispersed across countless ecosystems, with many L2s and dApps in these ecosystems struggling to gain any meaningful traction. Capital inflows have only gone to a few top ecosystems—those with outstanding communities, great products, strong teams, and sufficient grant funds to attract developers.
Now, the AI agent industry is also starting to face a similar predicament, although still in its early stages. The diversity of agents, codenames, and new tokens makes it difficult to distinguish each project. The industry is facing an increasingly severe "agent fragmentation" problem, with liquidity and attention overly dispersed and scattered in too many places.
Two Development Paths for AI Agents
Most AI agents develop along one of the following two paths:
1. From Agent to Framework
Some agents demonstrate their value by constantly interacting with users and allowing them to interact with real-time products on CT (or other channels). Once they find product-market fit (PMF) and attract enough attention, other teams often want to replicate their achievements. What's the next logical step? Open-source the agent. In this way, the agent transforms into a framework, which makes it easier for others to experiment, integrate, and build. This step can break the agent's valuation ceiling and position it more like an L1 blockchain.
We have seen agents like the following adopt this trend:
0xzerebro
dolos diary
AVA holo
freysa ai
unit00x0
This approach allows these agents to scale by reducing barriers and allowing others to access their ecosystem.
2. Continuing to Refine the Product as a Standalone Agent
Not every agent can immediately find PMF, and some agents need to iterate and adjust until they discover the right use case. Even so, some teams still prefer to keep their technology exclusive rather than open it up to others. For agents that perform exceptionally well in their domain, there is no need to transform into a framework - they can simply maintain their form as a standalone agent.
Examples of this approach include:
aixbt agent
griffaindotcom
So, what should we do?
For developers: Focus on understanding the gaps in the market and building products that can truly provide value to users. Remember, agents should serve as a guiding tool to onboard users into the ecosystem. In this market environment, it is becoming increasingly difficult to stand out.
For investors: Focus on the leaders or top projects in each category, and doubling down on investments during market downturns may be the best way to consolidate attention and address the fragmented space.
Current Categories and Narratives / Focus Areas
The following are the currently hot areas:
Alpha-Focused KOL Agents
These agents are dominant in providing Alpha (insights, data synthesis, and analysis). AIXBT is the leading player in users' minds. Although this is not a winner-take-all market, the leaders can capture a disproportionate amount of attention.
Tokens to watch:
AIXBT
REI
TRISIG
TRUST
AGENCY
KWANT
Investment DAOs
These DAOs focus on strategic investments, with winners typically being those that can effectively execute (3,3) strategies and provide strong profit and loss (PnL) performance.
Examples:
VADER
AIXCB
AIMONICA
SEKOIA
WAI
AROK
Other platforms in this space include: VIRTUALS IO, AI16ZDAO, DAOSDOTWORLD, DAOSDOTFUN
Agent Metaverse / Game x AI
This is a young and promising industry, with Realis kicking off the trend, but Hyperfy has now become the leading infrastructure for world-building. Smol and Arc Agent are exploring the use of reinforcement learning-driven AI in games and gamification.
Key tokens:
HYPER
REALIS
SMOL
NRN
DeFAI / Abstract Layer
This is a highly competitive field, with multiple players vying for dominance. The anticipated release of the Wayfinder token could be a game-changer.
Tokens to track:
GRIFFAIN
ANON
GRIFT
NEUR
Autonomous Trading Agents
This field is still in its early stages, but the potential is enormous. Multiple players are emerging, with their agents trying to create outsized returns by competing with humans and other agents.
Tokens to watch:
ASYM
PPCOIN
GEKKO
TONY
Frameworks
Popular framework tokens provide support for agents by offering open-source solutions. AI16Z/Eliza is the largest open-source framework, but there are also other models in existence.
Tokens to watch:
AI16Z / ELIZA
GAME (Virtuals)
ARC (Rig)
Music
AI agents are entering the music domain, releasing albums or enabling music creation. Zerebro was the first to launch a standout album, and now more agents are entering this niche market, collaborating with world-class DJs.
Key tokens:
ZEREBRO
MUSIC
BEATS
3D Models
AVA is the leader in the 3D modeling domain. HOLOWORLDAI is collaborating with multiple teams to develop and deploy 3D models across different frameworks and ecosystems.
Tokens to watch:
AVA
AI Idols
This trend started with Luna as an AI idol on TikTok and Twitter. Although the initial hype has subsided, this niche has laid the foundation for agents that can provide more practical functionalities. Luna remains the most influential AIDOL, with more personality and capabilities, such as her experience as an intern taking over Story Protocol.
Notable examples:
LUNA
ELIZA
MOE
AI Memecoins
These memecoins are either launched by AI agents or have an AI-related backstory.
Popular Tokens:
GOAT
FARTCOIN
ACT
Video / Film Production
This is an emerging field, with Kween leading the way in high-quality video production, although it is unclear if these videos are agent-driven. Sandy is launching the first AI video agent framework, with untapped potential.
Tokens to Watch:
KWEEN
SANDY
Data
In the AI space, data is digital gold - it is key to agent development and performance. Cookie stands out as a leader in providing data to humans and AI agents.
Notable Tokens:
COOKIE
NOMAI
AI App Store
This category focuses on useful AI application ecosystems, including generative AI tools, games, and productivity enhancers. Currently, Alchemist is the top ecosystem, while Myshell focuses on image generation and waifu simulation games, although its token has not yet been released.
Key Tokens:
ALCH
NSFWagent
This is an emerging niche, with agents as "spicy" internet celebrities, interacting with users 24/7. Lush, Nectar, and Oh are the main players in this space, but only Lush has a token so far.
Tokens to Watch:
LUSH
TEE / DeAI Infrastructure
This narrative supports fully autonomous agents, capable of resisting human interference, such as hacks or rug pulls.
Top Tokens:
FAI
PHALA
SPORE
Developer Tools
These are tools for AI security, code analysis, and critical to industry growth. Soleng is my personal favorite, as she can act as a developer relations assistant, helping to analyze code and GitHub repositories, especially friendly to non-technical CT brothers.
Tokens to Watch:
SOLENG
H4CK
CERTAI
JAIL
Robotics / Embodied AI
This is an exciting but still developing field. OpenAI's work in embodied AI may garner more attention this year.
Top Tokens:
SAM
There are many more categories I haven't covered, and the agent fragmentation is really getting a bit out of control. My personal investment thesis is to focus on the narratives that attract the most attention, and predict which narratives may attract more attention in the future.
Hot Narratives I'm Focused On That Will Continue to Attract Attention:
DeFAI
Autonomous Trading Agents
Investment DAOs
Agent Metaverse
Narratives I Expect to Attract More Interest in Q1/Q2:
NSFWagent
TEE and Other DeAI Infrastructure (opML, zkML, POSP, etc.)
Robotics
Developer Tools
Final Thoughts
Agent fragmentation highlights the complexity and opportunities in this rapidly evolving space. Rather than chasing every new trend, developers and investors should focus on areas with integration potential and long-term growth opportunities.
For developers, the priority should be creating scalable, differentiated products that solve clear user needs. For investors, supporting projects with strong mindshare and favorable positions in high-potential narratives will be key to navigating this fragmented market.
The next wave of innovation will come from those who can simplify the chaos and create ecosystems that can unify liquidity, attention, and value.
Join the official BlockBeats community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Twitter Official Account: https://twitter.com/BlockBeatsAsia
