Tesla Autopilot vision team lead Andrej Karpathy formally left the company. His action came after his four-month sabbatical leave fueled speculations about whether he would return. He joined the company as Tesla's AI director. He is known for deep learning and computer vision.

Tesla
(Photo : Milan Csizmadia/Unsplash)
Tesla

Tesla's Andrej Karpathy's Resignation Details

Karpathy's decision coincides with Tesla's disclosure in a California regulatory filing that it was laying off 229 data annotation personnel as part of the company's bigger Autopilot team. The company also closed its office in San Mateo, California.

Karpathy claimed to be on a sabbatical in March to relax and travel, but he had previously stated that he intended to return to Tesla this month.

Karpathy posted on Twitter how grateful he was to be a part of the team over the last five years. He 

Karpathy said he had no concrete plans for what he might do next, adding that he looks to spend more time revisiting his long-term passions around technical work in AI, open source, and education. Yet, Techcrunch said that it had a source telling that the ex-Tesla director was thinking of some venture investments.

Karpathy was an OpenAI researcher until he joined Tesla in 2017. He has an extensive experience in AI-related subjects and developed one of Stanford University's most prestigious deep learning courses.

His employment at Tesla, where he worked on the computer vision technology that supports the Autopilot advanced driver assistance system, was related to his dissertation work. Karpathy's dissertation was on developing a system that would allow a neural network to recognize numerous distinct and precise elements inside an image, label them using natural language, and then report back to the user.

Tesla's CEO Elon Musk replied to Karpathy on Twitter suggesting that the split was amicable.

ALSO READ: Tesla Crash in Texas Highlights Risks of Driverless Cars

Tesla Autopilot Investigation

In The Verge's report, The federal government is widening its investigation into more than a dozen collisions between Tesla vehicles and emergency vehicles while using Autopilot. The investigation is currently classified as an "Engineering Analysis," the second and last stage of a study before a potential recall.

The National Highway Traffic Safety Administration (NHTSA) website documents indicate that the next phase will involve additional testing and crash analysis to determine the extent to which Tesla's Autopilot and related systems may exacerbate behavioral safety issues or human factors issues by lessening the effectiveness of the driver's supervision.

The government's investigation involves 16 accidents in which Tesla drivers were using Autopilot when they struck stationary emergency vehicles, causing 15 injuries and one fatality.

Most of the occurrences happened at night, and the scene control equipment, such as cones, warning lights, flares, and an illuminated arrow board, was disregarded by the software. 

RELATED ARTICLES:AI: The Present (and Future) of Vehicles

Check out more news and information on Technology in Science Times.