Tesla re-emphasized the technical details of FSD, claiming that he "does not use high-precision maps, but only relies on neural networks and a large amount of data."

CTOnews.com news on October 3, Tesla has repeatedly stressed that his FSD technology does not use high-precision maps and relies on "visual recognition" to achieve autopilot. A few days ago, his official account forwarded a short video released by Tesla Silicon Valley car owners' Club and reiterated the relevant technical details of FSD.

It is reported that the video released by Tesla Silicon Valley owners' Club shows FSD driving on a dirt road by the lake without any navigation. "the world is not aware of what is happening," the club wrote. "

CTOnews.com, a post related to ▲ Tu Yuan X platform, found that Tesla's official account gave a long response, claiming:

Tesla FSD does not rely on high-definition maps, which means Autopilot can be enabled in places that cars have never seen before.

Although FSD will consider navigating to the correct destination, if no route or map is available, the software will choose the most likely path.

This' most likely path'is determined by what most people do in any given scenario and is powered by the overall learning results of Tesla's millions of cars around the world.

And the 'concept of lanes' is loosely embedded in the FSD to enable cars to 'confidently drive on unmarked roads'.

Musk has repeatedly stressed the visual recognition ability of Tesla cars and attacked the use of lidar hardware and high-definition mapping software, claiming: "roads are designed for biological neural networks and eyes. So digital neural networks and cameras will work best."

At the 2023 World artificial Intelligence Conference in early July, Musk also said that Tesla was "very close" to achieving fully autonomous driving. "I have been wrong about this prediction in the past, but I think we are closer to it than ever before," he said. "