2025 is the 'Agent Year.'
The release of Gemini 2.0 has fired the key shot towards the new world of AI Agents.
Recently, Google announced its most powerful AI large model, Gemini 2.0, with a clear positioning as 'AI model for the agentic era.' New features include native image generation and audio output in multi-modal outputs, and it supports native calls to Google Search, Maps, Lens, and other tools.
Google claims that 2025 is the 'Agent Year,' and the financial, payment, data aspects involved with Agents are a perfect match for blockchain. This is a revolution more exciting than DeFi.
AI agents will be the focus of mainstream development in the coming years. The agents in the crypto circle are the most natural integration with AI, and agents can directly decentralize using currency.
You can understand it this way:
$goat=BTC pure meme.
$virtual =eth [ Its$game= L2 ]
$ai 16z =sol [Its$ELIZA= JUP or ray, inaccurate but easy to understand.] Both have an explosive number of developers and are active on GitHub.
The AI agent track will see projects worth 10 billion to 50 billion emerge; currently, it looks like:
(1) agent platforms or frameworks, virtual games, ai16z, ELIZA, etc.
(2) Leading meme category, $goat - this one is particularly interesting, Binance does not list it on spot, but if you believe that 'trend is the ultimate force,' then just stick to doing the right thing and wait.
In any case, please lock onto the AI Agent track and welcome the arrival of this epic bull market's main wave.
Just sharing my personal investment logic for reference:
1. AI MEME, only considering leading coins like $GOAT with narrative origins, lacks logic. An early meme in a track may primarily have speculative attributes, but its role in educating the market and cultivating user awareness cannot be ignored, just like DOGE's contribution to the popularization of cryptocurrency.
2. AI Agent infrastructure is important; in the long run, it is underestimated. For example, early frameworks like $ELIZA, which have a rich GitHub open-source library and a particularly wide application coverage, are still very early and could be siphoned off by stronger frameworks at any time. However, the absolute advantages of open-source sharing and first-mover advantage form a barrier. Better frameworks are likely to emerge on top of it, which may influence its growth height but cannot eliminate it. If we treat ELIZA as the EVM of the AI Agent era, there will inevitably be new frameworks like EVM++, EVM Compatible, etc. to fill the gaps.
3. The issuance, custody, and trading of AI Agent assets are still stuck at the narrative level of meme-based asset issuance. One should focus on platform value and not be misled by 'junk assets' on the platform. The logic is simple: AI Agents are essentially a process of realizing value based on Crypto incentive mechanisms in AI scenarios. Early infrastructure, applications, protocol frameworks, etc., are still immature; how can there be so many AIXBT-level super applications emerging? What can the convenience of issuing coins bring? A realization? In the end, it's just creating a bubble trap for platforms like VIRTUAL, VVAIFU, and others. Therefore, don't chase after AI Agent super applications too much unless there is an absolute information gap. Otherwise, when you see the market cap surge, the cost-effectiveness will not be high; locking in innovative platform infrastructure is different.
4. There is still a vast blank space in the infrastructure construction of AI Agents. For example, early frameworks like ELIZA only solved the issues of LLM large models and AI Agent dialogue mechanisms, and were applied in scenarios like Twitter, Discord, etc. However, more still involves matching web2 information. When will there be an ELIZA++ version that can connect the ELIZA framework with web3 trading scenarios based on public chain paradigms? It is worth looking forward to; for example, although platforms like Virtual and VVAIFU can quickly issue assets, they still lack greatly in subsequent decentralized custody and AMM-style intent-based trading. When will the previously heated discussion on AI Agent + TEE private key management mature? Can asset issuance platforms extend new gameplay in trading, with more interesting trading methods and incentive models emerging?
Moreover, can there be chains that establish interaction standards for AI Agents, becoming the Cosmos of the AI Agent era, providing standard interoperability capabilities? Or chains that build permanent memory modular storage for AI Agents, providing necessary DA capabilities? There are also AI Agent visualization tools, execution engines, modular combination packaging services, etc. With these technologies and market-driven supporting infrastructures, will you still hold onto a pile of incomprehensible MEME anxieties?
As for the past various layer1s, high-performance chains, cross-chain infrastructures, chain abstractions, DA composite chains, layer2s, Restaking, etc., where is the way out for perfect infrastructure? Besides the original infralay application PMF delivery realization issues, a breakthrough point is the new concept of 'riding the AI Agent narrative.'
So, any chain that can provide some communication protocols, DA, application development, etc. for AI Agents is worth paying attention to. Of course, there will definitely be chains that purely ride the concept, but they are better than being stagnant and relying on old resources.
More practical strategies, specially prepared for spot enthusiasts and bull market hoarders! Click on my avatar to follow and get my bull market strategy deployment, shared for free with all fans. You need references, I need fans, to be the most generous bull market blogger on the internet!