General Electric builds jet engines and wind turbines and medical gear. But the 124-year-old industrial giant is also transforming itself for the digital age. It’s fashioning software that pulls data from all this hardware, hoping to gain an insight into industrial operations that was never possible in the past. The problem is that analyzing all this data is difficult, and the talent needed to make it happen is scarce. So GE is going shopping.
The company just paid an undisclosed amount to acquire a Berkeley-based machine learning startup called Wise.io. “There’s thirty of them,” GE CEO Jeff Immelt says gleefully of the Wise.io team, which is heavy with astrophysicists. “You match them with aviation data, and they’re killer.”
That’s great for GE and Immelt—and for their customers. But what if you’re a small software company trying to inject some AI into your operation? Wise.io was on a mission to “democratize AI”—to creating tools anyone could use to build machine learning services—but it’s now disappearing into GE. And at a time when machine learning is the prime way of staying competitive in the software world, that’s a notable thing. In the battle for scarce AI talent, companies like GE have an overwhelming advantage.
The cost of acquiring a top AI researcher is comparable to the cost of acquiring an NFL quarterback.
Not everyone can go out and grab thirty AI-happy astrophysicists. And if you can’t do that, the talent pool becomes very small very quickly, since these machine learning techniques are so new and so different from standard software development. Even the big players talk about the tiny talent pool: Microsoft research chief Peter Lee says the cost of acquiring a top AI researcher is comparable to the cost of acquiring an NFL quarterback.
Over the past few years, other heavyweights have snapped up so many other AI startups you’ve never heard of. Twitter bought Mad Bits, Whetlab, and Magic Pony. Apple bagged Turi and Tuplejump. Salesforce acquired MetaMind and Prediction I/O. Intel acquired Nervana. And that’s just a partial list. And it’s not just software and Internet companies doing the buying. It’s also giants like Samsung and GE that are incorporating AI into physical things. As soon as startups spring up to meet the AI need, they get sucked up into the maws of the hungriest, richest corporations.
“These big companies are establishing their advantage,” says Chris Nicholson, the founder of Skymind, a still-independent deep learning startup. “Technology is always a battlefield, and they’re tilting it with these acquisitions.”
A Shrinking Pool
Nicole Shanahan is seeing this firsthand. Her startup, ClearAccessIP, is building software that seeks to automatically analyze patent portfolios. The company uses the latest in machine learning techniques to mine relevant information for patent transactions. That means Shanahan needs data scientists, people with experience in deep neural networks and software like Google’s open source TensorFlow engine that drives this increasingly important form of AI. She’s trying to hire four data scientists who specialize in machine learning, and she can’t land even one. Last week, she put an offer out to a candidate in Canada, but he declined, saying he couldn’t make the move to the States. “It really slows us down,” Shanahan says. “We may end up hiring him as a contractor and letting him work from there.”
Can’t she just hire ordinary coders? Not really. Building this machine learning technology is quite different from standard software engineering. It’s less about coding and more about coaxing results from vast pools of data.
The irony is that Shanahan’s domestic partner happens to be Sergey Brin, the co-founder of Google, one of the main companies soaking up so much of the available deep learning talent. In 2013, Google acquired DNNresearch, nabbing one of the primary forces behind this technology’s recent rise: University of Toronto professor Geoff Hinton. Then it snapped up DeepMind, the company that recently shocked the AI world by building system that could crack the ancient game of Go. These acquisitions were the start of the industry’s buying spree.
But it’s not just that big players like Google are buying up so much of the talent. It’s that these big players are nabbing so many startups that were trying to make AI development tools accessible to all kinds of businesses—tools that would let businesses take advantage of AI without needing as much in-house talent. Like Wise.io, Metamind aimed to provide tools that anyone could use to build machine learning services, but it has now disappeared into Salesforce. Nervana, another option, is now part of Intel. So, in a way, the world needs even more AI talent than it otherwise would—though there are signs that some of the big players will pick up the mantle of democratized AI.
In the wake of the US presidential election, it may be even harder for small players to find talent. Last week, University of Montreal professor Yoshua Bengio, another key player in the deep learning movement invited AI researchers and data scientists to Quebec, where he is trying to bootstrap an entire community of companies around deep learning research. “In the depressing aftermath of the US elections, I would like to point out that interesting things are happening in the great Canadian North, with a very different kind of government,” he wrote on Facebook. Yann LeCun, the French born head of AI at Facebook, reposted the invitation, no small thing in the tiny, close-knit global community of AI researchers.
“Montral looks like a good landing site if you are an AI researcher looking for a job in North America and you belong to one (or several) categories of now-undesirable people for the newly-elected US government,” LeCun said.
That potential brain drain is not a big problem for big companies like Facebook and Google and Microsoft and, yes, GE. As Immelt points out, they’re international companies that can employ researchers practically anywhere. In his post, LeCun points out that Facebook now operates an AI lab in Paris. But that dispersal of talent could add one more layer of difficulty for smaller players here in the States. If researchers flee the country and others decline to come, the pool of available talent shrinks even further, receding beyond US borders and even further into the biggest companies.
For smaller companies, the hope is that some of that expertise will trickle down in the years to come.
Correction: This story has been updated to clarify the mission of ClearAccessIP.