Jensen HuangNvidia's charismatic CEO, Huang, chose the Consumer Electronics Show (CES) to drop his bombshell: a new generation of open-source artificial intelligence models designed for autonomous driving. Unveiled this Monday, January 5, 2026, in Las Vegas, this technology promises to transform cars into rolling brains, capable of "reasoning" in the face of the unexpected rather than simply reacting mindlessly to pre-programmed scenarios. A leap forward that, according to Huang, evokes the impact of ChatGPT on conversational AI. But behind the enthusiasm, a question lingers: will this breakthrough finally realize the dream of the driverless car, or is it just another high-tech mirage to boost Nvidia's stock, already the world's most valuable company? The collaboration with Mercedes-Benz, launched in 2020, is bearing fruit. Huang, a showman in a leather jacket, presented these models as a "ChatGPT moment for physical AI." The idea: algorithms that analyze complex situations in real time, anticipate dangers, and make nuanced decisions. No more robot cars braking for a flying plastic bag. Here, we're talking about AI that "understands" the world, thanks to massive datasets and advanced simulations. Nvidia, the graphics chip giant, is extending its hardware empire into software, encroaching on the territory of Tesla and Waymo. A colossal market: automotive robotics, according to Huang, will be the next El Dorado, far surpassing the video games that made his company's fortune.
A frantic race towards autonomy
What makes this announcement so striking is the timing: CES 2026, the major consumer electronics trade show, arrives just as the promises of autonomous vehicles have been stalling for years. Tesla and its high-profile accidents, Waymo and its robot taxis limited to a few American cities – the industry promised the moon but delivered only prototypes. Nvidia is fighting back with open source: models accessible to all manufacturers to accelerate adoption. Mercedes, a pioneer in this partnership, will integrate these tools into its future models, aiming for commercialization by 2028. Huang sees this as a turning point: physical AI, applied to mobile robots, will unlock immense markets, from logistics to personal transportation. But skepticism lurks. The regulatory challenges (who is responsible in the event of an accident?), ethical ones (AI that "reasons" like a human, really?), and technical ones (the colossal need for computing power) remain insurmountable. Nvidia, valued at stratospheric heights, is riding this wave to justify its stock market bubble. Huang, a visionary or a purveyor of dreams? His speech, tinged with Californian optimism, evokes Disney's 70s self-driving Beetle #53 – stubborn and unpredictable. Except here, they're promising stability and reliability. It remains to be seen whether this AI will live up to expectations, or if it will end up in the ditch of unfulfilled tech promises. In Las Vegas, where neon lights mask doubts, Nvidia has reignited the race. For French drivers, accustomed to Parisian traffic jams, the self-thinking car sounds like a dream. But in 2026, with European regulators on high alert, the widespread arrival of these autonomous vehicles still seems a long way off. Huang, however, is betting on acceleration: the robotics of tomorrow starts today, and Nvidia wants to be its brain.
