At the recent AI industry summit, a point really struck me—the concept of high-risk artificial intelligence.

Think about it: when AI is no longer just for chatting or writing copy, but truly involved in payment settlements, on-chain transaction execution, business negotiations, and asset management decisions—areas involving real money—the rules of the game will change completely.

At that point, whether AI is smart or not becomes secondary. The only two critical questions are: Is this thing reliable? Can it be hacked? Can it be trusted? Are there vulnerabilities to attacks?

This is exactly the direction I have been observing—when AI touches on blockchain, DeFi, and on-chain assets involving real value transfer, trust mechanisms and security defenses become the decisive factors. It’s not about how powerful the AI technology itself is, but whether the entire system can withstand the test of risks.