Recently, I've been seeing a lot of talk about AI and cryptocurrencies, but actually, the market size is growing incredibly large. In terms of market capitalization, it's expanded to about $30.6 billion, nearly 4.7 trillion yen in Japanese currency. Major asset management firms like Grayscale are creating AI-focused funds, and a16z has listed AI agents as a key trend for 2026, so institutional investors' interest is rapidly increasing.

There are three main reasons why AI × blockchain is gaining attention. First, decentralized GPU computing. Currently, GPU processing power is concentrated in large companies like NVIDIA, Google, and Amazon, but if blockchain can network idle GPUs worldwide, anyone could access the necessary computing resources. Second, data transparency. Recording the origin and usage history of AI training data on the blockchain can prevent tampering and protect creators' rights. Third, AI decentralization. While big tech companies currently monopolize AI models, a system is emerging where development and provision happen in a decentralized manner on the blockchain.

There are actually quite a few stocks in this field, some of which can be bought on Japanese exchanges. Render is available on several domestic exchanges including Gate, and NEAR recently trades around $1.39 with a market cap of about 1.8 trillion yen. You can also buy FIL, GRT, and WLD domestically. Meanwhile, stocks like VIRTUAL and TAO are accessed through overseas exchanges, but among AI agent-related tokens, they have large market caps. Especially TAO, which holds the highest proportion in Grayscale's AI fund, is attracting a lot of attention.

However, there are some points to be cautious about. Looking at the DeepSeek shock in January 2025, AI-related cryptocurrencies tend to move closely with the overall AI industry. When DeepSeek released low-cost, high-performance AI, AI-related stocks were sold off sharply. There was also an incident where an AI agent accidentally sent a meme coin worth about $250k to the wrong address. This shows there are still technical and operational challenges.

The topic of AI agents is also interesting. Traditional AI, like ChatGPT, just responds to questions, but AI agents can autonomously execute tasks. For example, they can automatically buy when Bitcoin drops below $100k. Once they hold a cryptocurrency wallet, they can conduct economic activities without human intervention. In fact, a CEO of a major exchange announced successful cryptocurrency trading between AI bots.

For investment, it's safest to start with stocks available on domestic exchanges. Render is a leading example of decentralized GPU rendering, and NEAR is gaining attention as a fast, low-cost layer 1 blockchain. FIL is for decentralized storage, GRT indexes blockchain data, and WLD combines biometric authentication with AI—each has different use cases. However, it's crucial to thoroughly investigate each project's technical details. Some projects only list AI terminology in their white papers without actual integration.

AI-related cryptocurrencies are likely to continue expanding, but they are more influenced by the trends in the AI industry than traditional cryptocurrencies. It’s wise to diversify investments across multiple tokens within your available funds.