Recently, GAIB's CEO @ konyk001 conducted an interview. After reading through it, I feel that it is much more readable than the official document. Let me record my feelings: Kony used to work as a VC and saw too many ups and downs of on chain projects. Later, he ended up doing GAIB himself, without restarting a chain or building an ecosystem. Just do one thing: financialize and tokenize the GPU assets of cloud service providers, package them into standardized on chain assets, connect to DeFi markets, and solve financing problems. It is to turn computing power into a tradable financial asset structure. Enable ordinary investors to invest in AI infrastructure. To be honest, are you saying it's innovative? somewhat. After all, Kony himself said that this is not pure financial innovation, but a value release. But I think this is more like a "realistic technology attitude" in the era of AI, the big foam narrative. Because what needs to be solved is the urgent issue of market resource allocation, rather than the problem imagined in the grand narrative. In the interview, he also mentioned one point: 'AI won't wait for you to be ready, it will only run forward wildly.'. '' Because if we really want to build an AI model, what runs behind it is not PPT, but cards, electricity, and computer rooms. Not enough computing power, everything is just empty talk. Over the past few years, large cloud vendors have taken on the vast majority of the demand for AI infrastructure; But since last year, more regional small and medium-sized cloud service providers have emerged, hoping to grab a corner of the market and have indeed succeeded. The problem is that everyone is short of money. Kony stated that GAIB is the financing engine for these new computing powers. This statement has a strong sense of propriety and does not exert too much force. And, like him, I firmly believe that GPUs will become the core commodity and highly liquid asset of the AI era, like oil and gold. I hope Gaib can turn GPUs into an economic unit that everyone can participate in. Not every project lives on heat. Some projects should rely on clear structure, real supply and demand, and someone using and working on them. Some projects are suitable for completing one round of interaction before leaving; But some projects are suitable to stay and take a look when they are not yet lively. I just want to see if it can really find a meaningful path in computational finance. Even if it's not possible, it's still a segment that I have participated in seriously.