Horizon Robotics CEO: Car Chips, Mind-Off Driving and China's Speed

Kai says Horizon Robotics remains overwhelmingly focused on passenger vehicles—probably 95%, nearly 100%, of its business—and its chips already run in roughly one out of three cars in China. That market is moving quickly: Chinese automakers launched more than 580 models across roughly 60 to 70 brands in the first half of the year, while car-platform production cycles take about two years in China versus typically five years for global OEMs.

Kai says automotive chips are difficult for new entrants because they must meet demanding safety, quality, qualification, and testing requirements, and a new generation typically takes three to four years to develop and prepare for automakers. Horizon pursues innovation in areas such as chip architecture, compilers, and algorithms, but relies on capable partners and established practices for safety-critical automotive work.

Kai says better self-driving depends on advancing onboard computing, high-quality data that captures meaningful corner cases, and larger models—not on improving chips or models alone. The guest predicts all passenger vehicles in China will be hands-free by the end of 2028, while also predicting that coordinated self-driving cars could improve traffic efficiency tenfold if they stop competing for road space.

Kai describes China’s supplier base, partner availability, and close geographic proximity as factors that support fast collaboration and innovation. In that context, he argues that Chinese manufacturing could help U.S. robotics companies innovate faster, favoring cross-border cooperation over a zero-sum view of the robotics race.