Learn Satoshi Nakamoto's skill of "self-examination" to create an intelligent encryption system. There are two types of autonomy: -1. Passive autonomy, i.e. tools/platforms, such as blockchain platforms with formal consistency and deterministic consensus. -2. Active autonomy, i.e. intelligence, like Bitcoin miners actively computing nonce autonomously. The autonomous updating of feedback neural weights in neural networks like GPT large models. In Sinton's words, machines are self examining and iterating without human involvement, and humans are not clear about the details of how machines do it, that is, the process of independent calculation and thinking of machines cannot be reproduced in human form. It is unclear whether the nonce of Bitcoin is calculated by miners using paper sheets, guessed by them, or by quantum computers. Just like the weight adjustment of neural networks, humans cannot formalize how to deduce and calculate step by step. Only the experienced machine knows all the experiences on its own, presenting us with descriptive symbols (the result of scrutiny), and cannot and does not need to convey the process of scrutiny to us. Both company management and smart cryptocurrency techniques can be mapped using this logic. Active autonomy is necessary to generate intelligence. Socrates' self-examination is the best means of generating this wisdom. The process of self-examination is a way to enhance one's intellectual abilities, but it does not need to be presented synchronously. Instead, only the results of self-examination need to be provided, which can be described in descriptive language.