Automakers’ AI Startups Expand, Pulling Talent from Silicon Valley

Automakers’ AI Startups Expand, Pulling Talent from Silicon Valley

In a sign of how serious the US automobile industry is about beating Silicon Valley to marketable self-driving cars, several AI startups working on this technology have multiplied in size since being bought by legacy automakers over the past two years, Christina Rogers reports at the Wall Street Journal. Argo AI, an artificial intelligence startup founded in Pittsburgh by former top engineers from the self-driving vehicle divisions of Alphabet and Uber, had fewer than a dozen employees when Ford Motor Company bought a $1 billion majority stake in it early last year. Today, it has 330 employees, including a number of software engineers and robotics researchers formerly employed by major tech companies like Apple and Uber.

Argo attracted these employees with an equity offer for new hires, which big tech companies can’t offer, Chief Executive Bryan Salesky tells Rogers. This ensures that each employee is “able to benefit from the upside being created in a direct way”—a potentially massive payoff given that Argo is helping Ford prepare to bring a fully autonomous car to market in 2021 while also developing a system it can sell to other companies. Being backed by a major company, but not owned outright or micromanaged by that company, gives Argo the agility to continue operating like a tech startup, while also benefitting from Ford’s economies of scale to manufacture and market the products it designs.

General Motors also bought a self-driving car startup, Cruise Automation, as part of a series of high-tech investments in 2016 that signaled the company’s intent to develop autonomous vehicles and made it a more attractive employer for tech talent. San Francisco-based Cruise, which GM also spent $1 billion to acquire, has staffed up to 740 employees and got $2.25 billion investment from Japan’s SoftBank Group last month, Rogers adds. Japanese automakers like Toyota and Nissan are also investing in the development of robotics and autonomous driving technology.

A major challenge for these companies has been in competing for talent with Silicon Valley: Google’s parent company Alphabet, along with ridesharing apps Uber and Lyft, are also developing self-driving cars, while Apple and Microsoft have been hiring AI talent for a variety of other purposes. In limited supply, this talent is exceedingly valuable, with newly-minted PhDs in artificial intelligence able to demand salaries in the hundreds of thousands of dollars. Cash-flush tech companies like Google, Apple, and Microsoft are often the only employers able to pay those prices—although some observers suspect that at Alphabet’s self-driving car unit, Waymo, for example, such massive compensation has made key employees hard to hold onto. Argo’s equity-based compensation strategy—offering enormous rewards, but only once the company is successful down the line—might give it an ironic edge in retention over companies that offer AI experts massive piles of cash up front.

In any case, AI expertise can’t remain limited to a few rich companies for long, as Gartner research shows that many companies already have plans to implement AI in the near future. Given the relative scarcity of AI experts with advanced degrees in the field, some companies are looking at ways to develop at least basic AI and machine learning skills internally: LinkedIn, a Microsoft-owned company, predicts that AI will soon be part of everything it does, so it recently launched an AI Academy for its employees to gain some familiarity with the technology. Google and the online learning company Coursera, meanwhile, are partnering on a course series to teach machine learning models to developers. Last year, Stanford University professor and Coursera co-founder Andrew Ng has launched a website, Deeplearning.ai, that offers coders with no AI background training in how to use deep learning, the main technique behind the current generation of AI.