张宁 / Zhang Ning (Pony.ai)

Zhang Ning is the [[PonyAI|Pony.ai]] speaker in 没有方向盘的出行，走到哪一步了？ NVIDIA × 小马智行一次聊透智能驾驶. He explains Pony.ai’s Robotaxi business as part of a broader unmanned-transportation portfolio that also includes logistics, port trucks, bulk-commodity trucking, last-mile delivery, and compute-platform work.

His strongest contribution is the operating definition of Robotaxi. In Zhang’s account, L4 is not just a more advanced L2 feature set; it means the system owns the driving task and the operator must absorb the service responsibilities formerly handled by drivers. That links him to Autonomous Driving Responsibility Boundary, Robotaxi Fleet Operations, and Robotaxi Economics.

The episode also uses Zhang to connect World Models and Autonomous Driving Simulation to real driving. He argues that L4 has to reason about what other road users may do after the vehicle acts, so useful simulation cannot be only replay; it must support counterfactual world evolution and continuous Sim2Real improvement.

Connections

• [[PonyAI|Pony.ai]] - company represented in the episode.
• Robotaxi Economics - cost, density, lifecycle, and profitability frame.
• Robotaxi Fleet Operations - dispatch, charging, maintenance, incidents, and passenger handling.
• Autonomous Driving Responsibility Boundary - L2/L3/L4 responsibility distinction.
• Autonomous Driving Simulation and World Models - counterfactual traffic reasoning and validation layer.
• Robotaxi Local Acceptance and Autonomous Vehicle Regulatory Patchwork - public trust, regulation, and overseas rollout context.

Source Position

• The source treats Zhang as a Pony.ai operator explaining L4 deployment, not as a neutral industry analyst.
• His claims about Pony.ai’s fleet scale and gross margin are retained as source claims rather than independently verified facts.