An Interesting Bet
Our VLA 2.0 is set to launch next quarter. As the first version, it must progress steadily and reliably while also making a stunning debut, which puts immense pressure on us. Therefore, after testing FSD in Silicon Valley in mid-2024, I came back recently and yesterday tested and compared the latest FSD V14.2 and Robotaxi.
Overall, FSD has evolved from the silky L2 advanced driver‑assistance experience of a year ago to a reassuring, quasi‑L4 software stage. Over about 5 hours of test drives in Silicon Valley and San Francisco, standard vehicles and Robotaxi, both running the same model, behaved consistently and steadily. There are still some imperfections, but it's a major leap over last year's level. Our colleagues will share related videos later. This test drive further convinced me that the future will share the same autonomous driving system and hardware architecture between private vehicles and robotaxis. Everyone will be able to skip the L3 stage and to own an L4-capable car directly.
Due to time constraints, our first version will not achieve all the capabilities of the current FSD V14.2. But over dinner with the team today, I made a bet with the autonomous driving team: if by 2026-08-30, our VLA in China reaches the overall performance level of FSD V14.2 in Silicon Valley, I will build a distinctive Chinese-flavored cafeteria in Silicon Valley modeled after XPENG current headquarters canteen. If we fail, our Head of Autonomous Driving, Dr. Xianming, has promised to streak across the Golden Gate Bridge. Our team believes they will win the cafeteria.
Who do you think will win?
$XPEV