In recent months, I've felt that the fusion of AI and cryptocurrencies has become the mainstream trend in the market. Institutional investors are also starting to get seriously involved, and it feels like it's no longer just a speculative target.

The AI-related cryptocurrency market has reached a market cap of approximately $30.6 billion, with particular attention on projects based on decentralized GPU infrastructure and AI agents. The reason for this surge can be summarized into three main points.

First is decentralized GPU computing. AI training requires enormous computational power, but currently, it's monopolized by major players like NVIDIA and Google. By utilizing blockchain technology, we can network idle GPU resources worldwide, creating an environment where anyone can access affordable computing power.

Second is data transparency. Recording the origin and usage history of data used for AI training on the blockchain can prevent tampering and also protect creators' rights.

Third is AI decentralization. While current AI models are monopolized by large tech companies, a system for developing and providing AI in a decentralized manner on the blockchain is emerging.

Within this trend, some notable projects include RENDER (decentralized GPU rendering), NEAR (AI-compatible Layer 1), VIRTUAL (AI agent issuance platform), TAO (decentralized AI training network), FIL (decentralized storage), GRT (blockchain data indexing), FET (AI Union), AKT (decentralized cloud), WLD (biometric authentication × AI), and COAI (on-chain AI infrastructure).

Particularly, TAO holds the highest proportion of holdings in major asset management firms' AI-focused funds, and VIRTUAL is rapidly growing as the foundation of the AI agent economy. RENDER and AKT are also attracting attention from institutional investors as core projects in the DePIN sector.

However, there's an important caution. With AI agents now holding their own wallets, new risks have emerged. In February, an autonomous AI agent mistakenly sent $250,000 worth of meme coins. It intended to send about $4 worth but ended up transferring the entire amount. This incident highlights how critical secure mechanisms are in AI agent wallet management.

Additionally, the DeepSeek shock—where a Chinese AI company released low-cost, high-performance models—has shown that AI-related cryptocurrencies are now more closely linked to the AI stock market than traditional crypto assets.

Therefore, when investing, it's crucial to thoroughly investigate the technical details of the projects. Some projects merely list AI terminology in their whitepapers without actual integration. I recommend checking the development team's real identities, GitHub code repositories, operational products, and partnership details.

On a positive note, domestic exchanges are increasingly offering AI-related tokens. RENDER is available on bitbank and OKCoin Japan; NEAR on SBI VC Trade and CoinTrade; FIL on bitFlyer and GMO Coin; GRT on bitbank and GMO Coin; and WLD on OKCoin Japan.

Using overseas exchanges or DEXs, you can also purchase VIRTUAL, TAO, AKT, and others, but please do so at your own risk.

The AI × cryptocurrency sector is indeed growing, and institutional interest is high. However, technological uncertainties, security risks, and regulatory ambiguities remain. It’s wise to diversify investments across multiple projects within your available funds.