Intense competition in artificial intelligence is fuelling a war for talent and putting a high premium on scarce AI expertise.
Lavish remuneration packages are on the table for experts who can help develop AI applications, from face-scanning smartphones and whizzy gadgets to advanced healthcare systems and autonomous vehicles. This war for talent is escalating due to several trends, such as competition between the automotive industry and Silicon Valley to build self-driving cars. The likes of Facebook and Google have an array of other ideas for AI, and seemingly inexhaustible resources to bring them to fruition, with the right talent on board.
They will have to get in line. Element AI, an independent lab in Montreal, estimates that there are fewer than 10,000 people in the world with the skills to conduct serious artificial intelligence research. “What we’re seeing is not necessarily good for society, but it is rational behavior by these companies”, said Andrew Moore, a former Google executive who is now dean of computer science at Carnegie Mellon University. “They are anxious to ensure that they’ve got this small cohort of people” who can work on this technology.
DeepMind Technologies, a British AI lab acquired by Google in 2014, has seen soaring costs related to employee compensation. Recently released annual financial accounts show the lab’s “staff costs” ballooned to $138 million last year – equivalent to $345,000 an employee – as it expanded from a team of 50 at the time of the acquisition to 400 employees. Such price tags make it extremely difficult for smaller companies to compete in this area of technology recruiting.
The latest research into AI centres on deep neural networks, which learn tasks independently by analysing data. Google, Facebook and a few others started to recruit the few researchers who specialize in it in 2013. As the New York Times explains, “Neural networks now help recognize faces in photos posted to Facebook, identify commands spoken into living-room digital assistants like the Amazon Echo and instantly translate foreign languages on Microsoft’s Skype phone service.” These techniques are now being applied to, among others, autonomous driving, hospital services that can identify diseases in medical scans, and next-generation digital assistants that understand spoken words rather than merely recognising them.
With a shortage of AI expertise on the talent market, companies are draining it from academia, leaving fewer who can teach the technology. Uber, for example, hired 40 people from Carnegie Mellon’s renowned AI program to work on its self-driving cars in 2015. Stanford University and the University of Washington have also lost professors to companies. In an effort to nurture home-grown AI engineers, Google, Facebook and others run classes for employees.
In order to compete for this scarce human capital, start-ups and smaller companies are hiring physicists and astronomers, who have the necessary math skills. Many in the US are also searching internationally, in Asia, Eastern Europe and other locations that generally have lower wages. “I can’t compete with Google, and I don’t want to,” said Chris Nicholson, Chief Executive and co-founder of Skymind, a San Francisco start-up that has hired engineers in eight countries. “So I offer very attractive salaries in countries that undervalue engineering talent.” Unsurprisingly, big companies are doing the same.