Google Israel R&D center punches above its weight

Prof. Yossi Matias Photo: Eli Yizhar
Prof. Yossi Matias Photo: Eli Yizhar

Google Israel R&D chief Yossi Matias: Capabilities in Israel have made it more significant than its size.

Most of us know quite a bit about Google Israel, which handles the business activity of global giant Google in Israel. About the Google R&D Center in Israel, which operates separately from Google Israel, on the other hand, most of us know much less, perhaps because it takes care to operate under the radar, even though it has 700 employees.

At a meeting organized last week by "Globes" in cooperation with LeumiTech, the spokesman of the Google R&D center was its CEO - Google VP engineering Prof. Yossi Matias. He talked about the role of the Israeli center for Google, its development over the past decade, and shared his opinion of Google's vision, the local high-tech industry, and the growing role of technology in our lives.

Matias, a professor of mathematics and computer science, started in the high-tech industry as a researcher in Bell Laboratories. He spent a short time as an entrepreneur (he founded Zapper, and was chief scientist at HyperRoll, acquired by Oracle). In the summer of 2006, 11 years ago, he accepted the task of building from scratch Google's R&D center in Israel, at a time when Google was much smaller than it is now. Google, in case you have forgotten, the technology giant that created the search engine people just cannot do without, is now worth $652 billion, and for a long time has been much more than just a search engine. Establishing the center in Israel - and in other countries outside the US - is decisive proof of this.

"I've been at Google for over 10 years, and I'm still experiencing surprises. I've never been in one place for so long," Matias says with a smile. "When Google chose to build an R&D center here, they explained to me that there were two reasons. First of all, Google is a company based on engineering talent, so it looks for it wherever it's likely to be found. Secondly, Google is a global company, and most of its users are outside the US, so it's important for it to have a range of perspectives, cultures, and capabilities. I also asked what they wanted me to build in Israel, and they answered, 'You will figure it out,' meaning that you will decide. That's typical of Google: every individual influences its vision."

Matias says that the beginning was like a startup. "I lived in the US, and when I used to come to Israel for a few days, I would interview people in cafes and look at offices for rent. 10 years later, a team of 700 people has been assembled and gradually proved itself, taking on itself substantial tasks. It's a center that has grown very quickly, in comparison with other R&D centers."

"Globes": Based on the number of people you employ, where is the Israeli center in comparison with Google's other R&D centers around the world?

Matias: "The centers in New York, Seattle, and Zurich have over 1,000 people each, so you could say that Israel is in the second tier in size, but size is not the only criterion. The contribution of a center is a function of its capabilities, and capabilities in Israel have made it more significant than its size, because very significant work for the global company has been done in Israel."

Give us some examples.

"A large proportion of the work naturally involves searches, and that's the reason why my title is VP engineering in searches for the international company. The search team in Israel is probably the largest outside of California (where Google's headquarters are, T.T.). We have a collection of teams in Israel, in other words, a collection of projects, each of which is like a startup in itself. For example, the automatic supplements in the search engine use technology developed in Israel. The technology behind searches in sports and weather was developed in Israel. In general, the search is constantly reinventing itself."

What is the next big challenge in searches?

"The challenge hasn't changed: to organize the results so that they match what the user needs in the best form and time. It's expectations that are changing, meaning the user's expectations, and mobile devices have made this clear. For example, if I search on mobile for information about a movie, I want to be able to buy a ticket with two or three clicks. On mobile, the challenge is more interesting and tougher. As we get used to technology, we reach a point at which we want and expect more from it. I no longer want to waste 10 minutes in ordering a ticket to a movie I want to see."

Isn't that a challenge that involves more artificial intelligence?

"Yes, but not only that. It's how we connect all the players in this search ecosystem. Part of it's artificial intelligence, and part of it's machine learning - two things that enable us to do more things in a smarter way. When that happens, it's like magic."

How do you learn what the user expects?

"It's a very complicated process that includes research, behavioral analysis, and so forth. Part of this process is expressed in the use of assistance, like Amazon's Alexa. In other words, we can talk to technology, and not necessarily via the smartphone. Sometimes we need information proactively, meaning that the technology guesses what I need to know, for example, when we should leave home for a morning meeting. There's no reason not to get such help.

"At first, such capabilities surprised us. Later, we got used to them, and now we complain when they don't work. That's the nature of technology - you get used to it very quickly, until you complain when it doesn't work like it should.

"On the subject of artificial intelligence and machine learning, what's important in both of them is the interaction between a person and technology. We're in an intermediate era, in which people still have trouble with technology, like the third age.

"Artificial intelligence creates more natural interaction between us and technology. For example, a large proportion of Google searches on mobile are done by voice, meaning that instead of typing, people ask the smartphone questions aloud. For me, artificial intelligence leads to ambient intelligence (electronic environments that are sensitive and respond to human beings, T.T.) - intelligence that detracts from the friction created between a person and an electronic product as a result of the latter's use. It saves a lot of headaches. For example, telling the home WiFi network to connect to the smartphone, or telling the television to change channels without using the control."

Google is not a serial acquirer of Israeli companies - at least, not on the same scale as Cisco Systems or Oracle. Its biggest acquisition in Israel was Waze - over $1 billion - and anything further about that is superfluous. Commenting on Google's acquisitions policy, Matias says, "We're in an exciting era, with a combination of cloud computing, which enables large teams in large enterprises to develop new things more quickly, so sometimes there are more in-house things. It's not necessarily a strategic question of whether to acquire to develop yourself; it's more a desire to build large teams that can drive the technology forward. When it's necessary and there's an opportunity, we bring a team in from outside through an acquisition."

Although Matias is a professor of mathematics and computer science by training, he admits that his job does not really leave him any time for academic pursuits, such as lecturing.

What do you think about the trend in the number of students signing up to study the exact sciences, such as mathematics, physics, and so forth? Is it going up or down, or staying the same?

"It has become more popular to take an interest in the exact sciences. The opportunity now for young people to develop themselves and do significant things is historically unprecedented. By learning the right things, every young person is capable of solving big problems in just a few years. How did that happen? Technology has reached a state in which almost anyone can solve problems - through cloud computing, which is the ability to utilize the functionality present on server. For example, Google has an open code library with a very popular machine learning program open to everyone. For example, we have found that machine learning greatly improves Google Translate. In other words, anyone who wants to solve some problem can use some kind of software."

How easy or difficult is it for you to compete for Israeli brains?

"On the one hand, we are in big demand, and it's very rare for someone with an opportunity to work with us not to do so. On the other hand, it sometimes happens that there is a mismatch, due to contrasting goals. When I interview a candidate, I ask him, 'Why did you come to Google?' The answers I usually get are that people want to work in a place that has an impact, in other words, to do things with a positive effect on a lot of people. Google touches billions of people every day, and that's a really good feeling."

The way you see it, looking ahead, is the supply of brains the biggest problem of Israeli high tech?

"The problems can be divided into two: The startup nation model is being copied in more and more countries, so it has to be preserved. In addition, especially where the supply of brains is concerned, there's no doubt that we have to take care of our future technological reserve. The educational system has to be changed, so that it at least maintains what we have. Boys and girls have to be encouraged from a young age to think entrepreneurially - to learn how to learn."

Entrepreneurial or business thinking?

"Entrepreneurial. Business thinking will come later by itself. The engine behind innovation is entrepreneurial thinking that asks: how do I solve a problem, and how do I use technology to solve the problem? In addition, it's important that we adjust ourselves to changes and develop accordingly. We shouldn't be afraid of changes. Change is part of the magic."

Finally, in response to a question from the audience about how the way and the degree to which the human mind degenerates when technology solves more and more problems, Matias answered, "Technology raises the level with which the human mind can cope. You could say that technology doesn't make the human mind idle; it challenges it. Rather than making the human mind degenerate, it empowers it."

Published by Globes [online], Israel Business News - www.globes-online.com - on May 15, 2017

© Copyright of Globes Publisher Itonut (1983) Ltd. 2017

Prof. Yossi Matias Photo: Eli Yizhar
Prof. Yossi Matias Photo: Eli Yizhar
Twitter Facebook Linkedin RSS Newsletters גלובס Israel Business Conference 2018