Tel Aviv hosting Google’s new start-up accelerator
Google – which can hire the cream of the crop from anywhere in the world – has picked Tel Aviv to be the site of a new accelerator that will assist start-ups that specialize in machine learning, artificial intelligence and data science.
The accelerator began operating in the beginning of March as part of Google’s Launchpad startup program. It’s the first time Google has taken its machine-learning program outside the United States.
“Ask yourself, how can Google take its first steps in Israel?” asked Roy Glasberg, founder and general manager of the Google Global Acceleration programs and the Google AI Studio. “Silicon Valley is the leading, most mature tech ecosystem in the world. But after five years of traveling the world, I came back to Israel, which is our best sandbox for experimentation.”
Glasberg met at Google’s sprawling tech campus near the Tel Aviv Hashalom railway station, along with Nir Chinsky, head of Google Cloud MEA & CEE, which is responsible for business strategy and growth across East and Central Europe, the Middle East and Africa. Both raved about the accelerator program as they showed off their skyscraper office suite.
Almost every major tech company in the world has expressed interest in opening a start-up accelerator in Israel, from Microsoft to Facebook to Amazon.
“In Israel, you have a couple of things that enable us to be the sandbox,” Glasberg said. “We’re not just fish but we’re a fish tank. Everything is concentrated in one small area, and the ability to find mentors and professionals and companies, it’s all confined to a very small area, where everyone is collaborating and working together.”
As part of the new machine-learning accelerator, Google will work with several start-ups at a time. The accelerator is an attempt by Google to harness the innovative drive of small start-ups and be able to scale that technology on a larger scale.
“IT’S NO longer about building an app or a service,” Glasberg said. “It’s about scaling, it’s about building automated, smart and embedded machine learning that enables you to write and perform computing at a pace that has never been seen before, one that was limited to companies to huge resources, huge budgets, engineering teams of thousands of people.”
Google provides an opensource, machine-learning platform for the start-ups, along with cloud services for them.
Eventually, AI and machine-learning services will be available to any startup or sole developer, not just for companies focusing on those fields.
“AI and machine learning used to be an area that you needed a PhD in algorithms to deal with that,” Chinsky said. “That was the first barrier to companies; build the models, train it with data, and you start to see the results .... What Google is doing, it’s making those technologies not only accessible for PhDs.”
Part of the drive to set up the Tel Aviv accelerator is the company’s effort to move from a mobile-first mentality to AI-first.
And similar to how the Internet transformed how brick-and-mortar shops conduct business, AI and machine intelligence could portend a similar revolution, say the Google Israel executives – it will be an enabling force that one cannot do without.
One local start-up that Google is working with is Jerusalem-based BrainQ Technologies. The company is working on a product that could enable paralyzed patients to work, treating neurological disorders in innovative ways.
“We’re looking for startups that have a very critical mission that will have a very big impact. And BrainQ wants to enable people who went through a stroke to regain brain and body functionality. They do that by regaining AI and machine-learning capability,” Chinsky added.
Separately, Google works with some 100 to 150 startups through different Israeli accelerators. Many of those companies work in healthcare, fin-tech and agro-tech, among other industries.
In total, some 2,000 startups seek to work with Google. That gets whittled down to less than 10% of all interested firms, a selection process conducted by outside accelerators that mirrors Ivy League admissions rates.
“We went to San Francisco and built a program based on the knowledge and expertise we got in Israel,” Chinsky said. “And now we’re bringing it back to Israel.”
The machine-learning accelerator program was initially launched with Peter Norvig in San Francisco last October.
Google is also expanding the accelerator program to multiple locales around the world, including Lagos, Nigeria.
The Tel Aviv accelerator doesn’t take equity from start-ups, but asks the companies to work with Google on specific challenges. The selected start-ups work with Google for half-a-year.