The hidden barrier to AI implementation lies in "trustworthy data." When discussing AI, it’s impossible to avoid models and computing power, but the deeper you go, the more you realize that the real bottleneck isn't the number of parameters, but the availability of real-world data that can be practically applied. Messari's "Crypto Theses 2026" early insight: scenarios like robotics and autonomous driving will increasingly depend on real data, which is the core logic repeatedly emphasized in the DePAI track.