Google just placed yet another bet on the idea that artificial intelligence will remake the world—and throw off wild profits.
The company disclosed today that it has created a new venture fund dedicated to investing in AI and machine learning companies. The initiative's first public investment: lead investor in a $10.5 million funding round for Seattle startup Algorithmia, which has built a kind of app store for algorithms. The service aims to make it easier for any company to use machine learning.
Google wouldn’t make anyone available to talk about its new fund, which doesn't yet have a name, or disclose its size. The company will say it is led by veteran executive Anna Patterson, vice president of engineering for AI. And the fund's existence and lead role in Algorithmia's latest round fit with what you might call Google’s shovel strategy for artificial intelligence: The company believes there are piles of money to be made by giving other companies the tools to strike gold using the technology.
Google and Algorithmia both rather cutely refer to the effort to sell AI shovels as “democratizing” the technology. More plainly, they’re pursuing a business opportunity created by the recent, rapid advancement in the power of machine learning. Lots of companies with ideas about how to use the technology in their businesses can’t access the expertise needed to do so, which is in short supply.
Algorithmia is trying to fill that need with a marketplace that is something like an Etsy for algorithms. Developers can upload their creations—say, a machine-learning model that identifies a car’s model year from a photo—and get paid if another person or company uses them. The company’s system converts uploaded algorithms into cloud services that can be easily integrated into an app, website, or other service.
There are currently more than 3,500 algorithms listed in Algorithmia's marketplace, for everything from detecting nudity to parsing sentences. A private version of the platform is used by large companies in finance, pharma, and other industries to help developers on staff share and deploy their algorithms. “There’s a lot of infrastructure needed to make a reality out of this promise that AI is going to affect every business process out there,” says Diego Oppenheimer, CEO and founder of Algorithmia. “We’re the AI plumbers—it’s the stuff people don’t want to do, but it’s a big market opportunity.”
Google isn’t the first to spin up a dedicated AI fund or even to plow some of it into a startup working on AI plumbing. Microsoft announced it had created an AI venture fund last December, and in May invested in a startup called Bonsai that helps companies build with machine learning. IBM and Amazon have also created AI-related funds—IBM's for startups building on top of the Watson suite of developer tools and Amazon's for the Alexa voice operated assistant.
But just because a large company throws money at an emerging new technology doesn’t guarantee that such companies will stay ahead of the curve. Steve Blank, a veteran Silicon Valley entrepreneur and adjunct professor at Stanford, says that it makes sense for Google and others to fund outsiders. After all, no one company can hope to monopolize all the creativity and talent in a given area. But he says the history of the tech industry shows that dominant companies like Microsoft of the ’90s and Google of today eventually get out-maneuvered by startups more able to adapt to new ideas. (There's even a name for the phenomenon: the innovator's dilemma.)
“Machine intelligence is not like inventing a better sword, it’s like inventing gunpowder,” Blank says. “Smart startups will figure out how to attack large tech companies with it just as large tech companies used it to attack the companies before them.” In other words, tech giants handing out cash to enable creativity and innovation around machine learning could be handing their future competitors a loaded gun.
