Silicon Valley's center San Francisco experienced a rare major power outage yesterday, and an unexpected event revealed the true differences in AI autonomous driving systems.

At the street corner, a fleet of a well-known driverless solution is stuck in the dark, its hazard lights flashing incessantly, like a child in distress. Meanwhile, another company's autonomous driving system is smoothly navigating the same street, relying on its own judgment to keep operating.

This detail is worth pondering: emergencies expose not only the underlying technical capabilities but also reflect the robustness design of the entire system. What qualities should a good AI driver possess? Is it absolute adherence to established rules, or the ability to make autonomous decisions in abnormal situations?

When infrastructure fails, true intelligence is reflected in the ability to break dependencies and self-adapt. This insight goes far beyond the field of autonomous driving for the entire AI industry chain.