China’s self-driving truck leaders say AI success won’t speed up rollout – here’s why

BEIJING — While AI updates make headlines every few weeks, those advances aren’t enough to get self-driving vehicles on the road more quickly.

That’s according to Chinese autonomous trucking companies, who say improvements in big language models, from Anthropic’s Cloud to China’s DeepSeq, have little impact on vehicle deployment timelines.

“Having the world’s best linguistics[expert]doesn’t mean he’s a good driver,” Pony.ai CEO James Peng told reporters last week. “AI is a very broad term. They’re completely different things. Absolutely… zero relevance.”

“When we work on language, when we play sports, when we drive, we all use different skills,” he said.

Autonomous driving uses artificial intelligence to mimic a human driver with a combination of sensors, chips, and algorithms. But the real-world training data needed is very different from what powers large language models, called world models, like OpenAI’s ChatGPT.

CEO Julian Ma told CNBC that Inceptio, a self-driving truck startup, is still sticking to its timeline for a mid-2028 commercialization milestone, unaffected by broader advances in AI.

They expect Inceptio to have collected 5 billion kilometers (3.1 billion miles) of truck driving data in China by the third or fourth quarter of 2028 – enough to allow fully autonomous heavy-duty trucks to operate on public roads.

With 5 billion kilometers of collected driving data, AI can turn it into a world model with 50 billion kilometers of experience – which is then enough to allow a heavy-duty truck to drive completely on its own, Ma said. He hopes that in some parts of the country, trucks can start operating without any humans.

Achieving that goal in about two years is already fast enough, he said, adding that for driverless trucks to become a widespread reality, they will need partnerships with manufacturers and regulatory approval in addition to the technology.

Autonomous vehicles rely heavily on data about driving on the roads. Like robotaxi companies, self-driving truck operators run manned tests to safely gather training data.

According to ARK Invest, Incepio has recorded the most commercial autonomous truck miles in the industry to date, exceeding US rivals. big ideas 2026 report In January. At that time, the company had driven 250 million miles – much more than fellow Chinese autonomous driving company pony.aiWhich came in second place with 4.2 million miles.

According to the report, US-based rivals Aurora, Kodiaq and Gatik join the top five with a combined 8.9 million miles.

Inceptio’s Ma said in late April that the company’s trucks had driven 700 million kilometers (434.96 million miles), and was targeting 1 billion kilometers (621.4 million miles) by the end of the year. He said the company can use AI to focus on which specific scenarios to collect test data.

At the Beijing Auto Show, Pony.ai also announced an upgrade to its PonyWorld 2.0 AI model to improve its ability to collect specific data and train models more efficiently. The company, which already operates robotaxis in China and other countries, unveiled a fully driverless light-duty truck that it developed with the battery giant. CATL.

While China has 5-year development plans that increasingly emphasize technological goals, Ma said it is often companies that take the lead in driving innovation.

“We have to make it happen,” he said, “before regulators see the technology in action and are convinced enough to provide policy support.”

But it’s clear there’s still a long way to go before we see driverless trucks and cars driving across the country.

“Automobiles are actually the most challenging area for AI, and to some extent surpasses the difficulty of embodied AI, because it involves safety,” Ma said. Embodied AI includes humanoid robots.