Will This Deep-Learning Expert Take Tesla's Autopilot to the Next Level?

Tesla just lost a major Autopilot executive. But the electric-car maker's new hire doesn't disappoint.

Daniel Sparks
Daniel Sparks
Jun 21, 2017 at 9:26PM

Well, that was quick. On Tuesday, Tesla (NASDAQ:TSLA) said its head of Autopilot software is leaving the company, just six months after he joined the electric-car maker. The departure from Tesla's Autopilot team is the second of two key losses from the important unit in the past seven months.

But Tesla's new hire for the Autopilot team doesn't disappoint. Here's a look at Tesla's recent departures and hires for the vital segment, as well why Autopilot's ongoing development is so critical for Tesla.

Tesla Model X interior with Autopilot activated

Image source: Tesla.

Comings and goings

In January, Tesla's hiring of Apple's Chris Lattner was considered a major win. At Apple, the executive was responsible for creating the tech giant's new programming language, called Swift. At Tesla, he took on the role of vice president of Autopilot. "We are very excited that Chris is joining Tesla to lead our Autopilot engineering team and accelerate the future of autonomous driving," Tesla said in a press release about the important hire.

But Lattner's appointment to Tesla's Autopilot team notably followed the loss of Tesla's director of Autopilot programs, Sterling Anderson, who was integral to the deployment of Autopilot at Tesla.

And now Tesla's Autopilot team is seeing turnover at the top again. Tesla confirmed Lattner's departure this week.

"Chris just wasn't the right fit for Tesla, and we've decided to make a change," Tesla said in a statement to The Motley Fool. "We wish him the best."

As Lattner departs, Tesla is bringing in Andrej Karpathy, a research scientist at OpenAI -- Tesla CEO Elon Musk's nonprofit side project. Karpathy holds a computer vision Ph.D. at Stanford University and knows how to work with deep neural nets -- an expertise required to work with the vast amount of data served by Tesla's Autopilot sensors built into its customers' vehicles. Evidence of his knowledge in deep neural networks: Karpathy taught Stanford's first deep-learning class at Stanford.

Karpathy's title suggests he is taking on an expanded role in comparison with Lattner's responsibilities. As director of AI and Autopilot Vision, Karpathy will report directly to Musk. Despite his expanded role, however, Tesla's Jim Keller has overall responsibility for Autopilot hardware and Software. Karpathy will work closely with Keller, Tesla says.

Autopilot's success is crucial

It would be difficult to overstate the importance of Autopilot for Tesla, which has been shipping every vehicle since October 2014 with Autopilot sensors. And every vehicle since October 2016, including Tesla's upcoming Model 3, is shipped with Tesla's second-generation hardware suite of sensors and computing power.

Related Articles

Tesla has promised an "Enhanced Autopilot" for its vehicles with updated hardware -- a version of Autopilot that will feature more capabilities, including automatically changing lanes on the highway without any driver input at all, automatically exiting the freeway, and more complex summoning navigation.

Tesla Model X interior

Tesla vehicle interior. Image source: author.

But the big test for Tesla's Autopilot will come later on. Eventually, Tesla says its second-generation Autopilot hardware, combined with extensive software validation and regulatory approval, will enable its vehicles to completely drive themselves.

Karpathy undoubtedly has a big challenge in front of him.