Google turns its lucrative web search into AI machines



For any other business, that would be a rickety distraction from its core business. At Google, the two are linked. Artificial intelligence (AI) is at the extreme of machine learning, which allows people to create software that can learn more about the world. Google has been one of the biggest corporate sponsors of AI and has invested heavily in it for videos, speech, translation and, recently, research.

Over the past few months, a “very large fraction” of the millions of queries per second that people type into the company’s search engine have been interpreted by an artificial intelligence system, dubbed RankBrain, said Greg Corrado, principal investigator of the company. , describing for the first time the emerging role of AI in research.

RankBrain uses artificial intelligence to embed large amounts of written language into mathematical entities – called vectors – that the computer can understand. If RankBrain sees a word or phrase it is unfamiliar with, the machine can guess which words or phrases might have a similar meaning and filter the result accordingly, making it more efficient at handling search queries ever. seen before. .

Unique questions

The system helps Google, based in Mountain View, Calif., Process the 15% of queries per day it receives, which its systems have never seen before, he said. For example, it is adept at handling ambiguous queries, such as “What is the title of the consumer at the top of a food chain?” And RankBrain’s use of AI means it works differently from other search engine technologies.

“The other signals, they’re all based on findings and ideas that information seekers have had, but there’s no learning,” Corrado said.

Keeping an edge in search is essential for Google, and making its systems smarter and better able to handle ambiguous queries is one of the ways to keep control over time-pressed users, who now mostly search for money. using their mobile devices. “If you say the people at Google think about search,” Corrado said.

Several tools

RankBrain is one of “hundreds” of signals that go into an algorithm that determines which results appear on a Google search page and where they rank, Corrado said. In the few months it was deployed, RankBrain became the third most important signal contributing to a search query result, he said.

“I was surprised,” Corrado said. “I would say it went better than we expected.”

Adding RankBrain to search is part of a half-decade of Google’s push towards AI, as the company seeks to integrate technology into all aspects of its business. “Machine learning is a critical transformative means by which we rethink everything we do,” Google CEO Sundar Pichai said during the company’s earnings call last week.

Smarter than your average engineer

RankBrain so far lives up to its AI hype. Google’s search engineers, who spend their days crafting the algorithms behind the search software, were asked to examine certain pages and guess which they think Google’s search engine technology would rank in. head. While humans guessed correctly 70% of the time, RankBrain had an 80% success rate.

Typical Google users agree. In experiments, the company found that disabling this feature “would be as damaging for users as forgetting to post half of the pages on Wikipedia,” Corrado said.

Getting here was not easy. RankBrain’s rollout is a one-year team effort that began with around five Google engineers, including research specialist Yonghui Wu and deep learning expert Thomas Strohmann. It took a long time to make sure the system was categorizing things correctly.

The effort spread to dozens of people after Amit Singhal, the company’s senior vice president of search, gave the go-ahead for it to be rolled out across Google search at early 2015.

“It’s very carefully monitored,” Corrado said, with nothing that Google periodically updates the system with a load of new data to help it reason better with new concepts.

Key products

Google’s decision to deploy AI in research shows companies are starting to outsource their most valuable activities to systems controlled in part by artificial intelligence. Facebook Inc. uses AI techniques to filter the news feed which includes the social network’s personalized homepage and Microsoft Corp. uses artificial intelligence to increase the capabilities of its Bing search engine. Microsoft declined to say whether it was using an approach similar to Google’s.

“Search is the lifeblood of Google,” Corrado said. “Machine learning isn’t just a magic syrup you pour on a problem and it improves it. It took a lot of thought and attention to create something we thought was worth doing. . Bloomberg

To subscribe to Mint newsletters

* Enter a valid email address

* Thank you for subscribing to our newsletter.

Never miss a story! Stay connected and informed with Mint. Download our app now !!


Leave A Reply

Your email address will not be published.