US Top News and Analysis | China’s self-driving truck leaders say AI breakthroughs won’t accelerate rollout — here’s why
AI generated summary, Read the full article for complete information.
Chinese autonomous trucking firms say that recent leaps in large‑language‑model AI are not speeding up the deployment of driverless trucks. Pony.ai CEO James Peng warned that expertise in linguistics does not translate to driving ability, and Inceptio’s Julian Ma reaffirmed the company’s target of a mid‑2028 commercial rollout, aiming to accumulate roughly 5 billion kilometres of real‑world driving data by the third or fourth quarter of that year. Inceptio has already logged about 700 million kilometres and plans to reach 1 billion kilometres by year‑end, using its own AI to prioritize data‑collection scenarios, while Pony.ai has introduced an upgraded PonyWorld 2.0 model and unveiled a fully driverless light‑duty truck developed with battery maker CATL. The firms stress that autonomous‑driving systems rely on massive, vehicle‑specific sensor data and “world models,” not the textual knowledge powering LLMs, and they still face significant regulatory hurdles, with China recently suspending new autonomous‑driving licences after robotaxi incidents and U.S. operators such as Waymo experiencing setbacks from power outages.







