Elon Musk believes he has the ability to transform Tesla from a pioneering maker of electric vehicles into a market leader in the booming field of artificial intelligence.
Key to his efforts is “Full Self-Driving” (FSD), a feature that remains in beta after launch three yearseleven different iterations and over 500 million miles of real-world testing.
Yet despite delays in software verification, eventual success would mean Musk finally has the foundational technology he believes is needed to make Tesla the most valuable company in the world.
Those plans may now be in jeopardy after Tesla confirmed in a quarterly filing with the Securities and Exchange Commission that its controversial self-driving software is the subject of a U.S. Justice Department investigation. Published on Monday.
“The company has received information requests, including subpoenas, from the Department of Justice. These include requests for documents related to Tesla’s Autopilot and FSD features.”
Previous wording in its various 10-Q risk reports referred only to requests and never to formal subpoenas.
Tesla added that before making the usual legal disclosures, to its knowledge, no government agency has concluded that any wrongdoing occurred during any ongoing investigation: “If the government decides to take enforcement action, could have a material adverse effect on our affairs.”
Tesla did not disclose the nature of the information sought by the Justice Department.Neither party can be contacted via wealth for further comment.
Autopilot is the most basic form of the software and comes standard on all Tesla vehicles. FSD is currently a $12,000 optional add-on, after a price reduction of $3,000.
Under Musk’s leadership, Tesla has pursued a path to autonomous driving technology that is unique in the industry. While other companies such as General Motors’ Cruise and Alphabet’s Waymo are cautiously trying to limit robotaxis to specific areas (often meticulously mapped to the inch and outfitted with expensive laser scanners), Musk is trying to use robots that only feed in camera data. Complex onboard computers.
The idea behind this approach is that if proven reliable, Tesla would enjoy unparalleled scalability to roll out the technology faster than any competitor.
Invest billions in Nvidia AI chips
Musk did not instruct the engineering team around Ashok Elluswamy to complete development before making the software commercial, but to make it available to paying customers. However, it is stipulated that customers are aware that they are being asked to validate the technology.
Musk is already far behind his original ambitious plan to have 1 million self-driving Teslas of his existing “S3XY” models on the road by the end of 2020. That led to some angry buyers, such as Apple co-founder Steve Wozniak, who accused Musk of dishonesty.
Still, the CEO said last week, after Tesla completed assembling a cluster of 10,000 GPUs powered by Nvidia’s latest H100 processors, that he remains as convinced as ever that artificial intelligence “has the potential to make Tesla the company it is today” The most valuable company in the world”. Used for neural network training.
In addition, the company has invested an additional $1 billion in the Dojo supercomputer, which uses custom-designed chips to speed up artificial intelligence training on visual data. Musk said he could license its use to other companies and believes it could one day be worth hundreds of billions of dollars.
Tesla shares rose 1.5% on Monday, paring some losses after its worst trading week so far this year on investor pessimism and lower-than-expected third-quarter sales and profits.
Svlook
