Development pace in AI is infrastructure, the variable that determines how quickly a laboratory builds empirical knowledge of its own systems and how much that knowledge accumulates relative to competitors. Huawei's chair has called on China's artificial intelligence laboratories to accelerate their work, a direct counter to Silicon Valley's calls for restraint grounded in rising concerns that advanced AI poses existential risks to humanity.
The Silicon Valley argument for slowing down rests on the view that AI capabilities have advanced faster than humanity's ability to assess whether the technology is safe. That position has driven calls for a deliberate reduction in the pace at which AI is developed and deployed.
Huawei's chair has framed China's response as the inverse. The call, directed at Chinese labs, treats acceleration as the correct posture and positions the existential-risk argument as a constraint China should reject. The framing is directive rather than aspirational: China's labs must accelerate, not should or might. That language treats the pace question as settled.
The operational stakes of that difference are concrete. A laboratory running at higher velocity accumulates more training cycles, more model evaluations, and a broader base of observed system behavior. Those inputs compound. The lab that builds them faster has a larger empirical record to draw on when capability claims need to be tested or defended.
The existential-risk framework has been Silicon Valley's organizing argument for pulling back. Huawei's chair, speaking to China's AI laboratories, has countered it with a directive to move faster.