Human individual cognition and culture are an organic whole that promote each other, and their development mode is similar to the mutual promotion of computers and the Internet. The current artificial intelligence is based on deterministic systems, but to achieve higher-level intelligence, it needs to draw on the emergence of human intelligence. From the perspective of computational complexity theory's P/NP problem, individual cognitive processes are like solving NP hard problems, full of exploration; The formation and verification of cultural consensus is relatively easy, similar to the P problem. The development of human intelligence and culture presents a shift from the complexity of individual solving to the simplicity of group verification, manifested as non-linear emergence of intelligence. This understanding provides a new approach for building future artificial intelligence, which is to design a hybrid architecture consisting of distributed formal systems oriented towards verification (similar to P problems) and solution (similar to NP problems). Verification oriented systems ensure reliability, while solution oriented systems encourage exploration and innovation. The Bitcoin designed by Satoshi Nakamoto is considered an early example of this architecture. The UTXO system and proof of work mechanism represent distributed formal systems with different properties, which interact through consensus mechanisms and exhibit characteristics such as decentralization and anti censorship, indicating a paradigm shift in automation systems from determinism to self-organizing intelligence. Understanding and drawing on the collaborative evolution model of human intelligence and culture, especially the complexity from the P/NP perspective, will help break through the limitations of current artificial intelligence and move towards a more advanced and autonomous general artificial intelligence.