Briefly
- Elon Musk says Nvidia’s autonomous driving software program won’t strain Tesla for 5 to 6 years or longer.
- Nvidia unveiled Alpamayo, an open-source AI mannequin household for self-driving techniques, at CES 2026.
- Musk argues legacy automakers face lengthy delays in integrating cameras and AI {hardware} at scale.
Tesla CEO Elon Musk stated Nvidia’s newest autonomous driving software program won’t pose critical aggressive strain on Tesla for a number of years.
Nvidia confirmed off its new self-driving know-how at CES 2026 on Monday.
The software program facilities round Alpamayo, an open-source household of AI fashions designed to deal with complicated city driving utilizing camera-based video enter. The corporate demonstrated the system navigating a Mercedes automotive via metropolis streets in Las Vegas.
However Musk stated the software program stays 5 to 6 years from posing an actual menace to Tesla, citing the lengthy hole between partial autonomy and a safer-than-human driving, totally self-driving car, in addition to sluggish {hardware} deployment by automakers.
“The precise time from when [a self-driving car] type of works to the place it’s a lot safer than a human is a number of years,” Musk wrote. He added that legacy automakers face an extra delay due to the time required to design and combine cameras and AI computer systems into manufacturing automobiles at scale.
Regardless of Musk’s feedback, Nvidia CEO Jensen Huang praised Tesla’s self-driving know-how as “essentially the most superior AV stack on this planet.”
“I believe Elon’s method is about as state-of-the-art as anyone is aware of of autonomous driving and robotics,” Huang informed Bloomberg. “It’s a stack that’s laborious to criticize. I wouldn’t criticize it. I might simply encourage them to proceed to do what they’re doing.”
Throughout a Keynote speech at CES, Huang stated the chip producer’s work on self-driving vehicles goes again practically a decade.
“We began engaged on self-driving vehicles eight years in the past, and the rationale for that’s as a result of we reasoned early on that deep studying and synthetic intelligence have been going to reinvent your complete computing stack,” he stated. “And if we have been ever going to know the way to navigate and the way to information the business in the direction of this new future, we now have to get good at constructing your complete stack.”
Progress stalls
Nonetheless, advances in autonomous driving haven’t decreased the challenges for the budding business.
Waymo, which operates totally driverless robotaxis in a number of U.S. cities, lately issued a voluntary software program recall in December after automobiles didn’t cease for varsity buses.
That very same month, the corporate additionally briefly suspended service in San Francisco after an influence outage prompted automobiles to stall at intersections and block visitors.
In the course of the outage, Musk stated on X that Tesla’s restricted robotaxi service, which operates with a human security monitor, was unaffected.
Musk first hinted on the thought of self-driving vehicles in 2013, with the primary model of Autopilot launching two years later.
Tesla’s benefit lies in its present fleet and vision-only system, with automobiles already transport standardized cameras and onboard AI {hardware}.
Beneath its “Tesla Imaginative and prescient” method, the corporate depends totally on cameras reasonably than lidar and has eliminated radar and ultrasonic sensors from many automobiles and markets.
Tesla’s autonomous driving ambitions have drawn scrutiny, with critics questioning the security and reliability of its Autopilot and Full Self-Driving options following a collection of high-profile crashes, a few of which resulted in fatalities and federal investigations.
Usually Clever E-newsletter
A weekly AI journey narrated by Gen, a generative AI mannequin.