AXON, TOST, and IOT projects all performed quite well today, and there is a common point worth noting — they all have actual hardware support. This is the key.

Unlike pure software or AI applications, hardware projects serve a real user base, and these user needs cannot be replaced by large language models. No matter how powerful LLMs are, they cannot solve problems in the physical world. This practicality and risk resistance are precisely why such projects can maintain resilience amid market fluctuations. In the current AI boom, projects with hardware foundations and real application scenarios are actually more scarce and competitive.