Google puts AI expert in charge of its search algorithms
Google puts AI expert in charge of its search algorithms
By Richard Waters in San Francisco
February 3, 2016 8:14 pm
Google has put an artificial intelligence expert in
charge of its search algorithms, signalling a sea change in one of the core
technologies of the internet that may ultimately give intelligent machines the
job of finding and sorting information for humans.
The search company on Wednesday named John Giannandrea,
an expert in machine learning, to run the engineering group that devises its
ranking algorithms, the core technology for making sense of the deluge of
information on the internet.
He replaces Amit Singhal, an Indian engineer who oversaw
a rewrite of Google’s ranking technology after joining 15 years ago. He has
since headed its search engineering efforts as the company has consolidated its
grip on internet search for most of the world and wrestled with the new
approaches to search forced by things such as the rise of smartphone apps.
For Google, Mr Singhal’s departure is as significant as
the departure of design chief Sir Jony Ive would be for Apple, said Danny
Sullivan, editor of Search Engine Land and a long-time analyst of the search
industry.
He called Mr Singhal the “guiding force” behind the
Google search engine, which is still the main source of the company’s profits,
and an “orchestra leader” who had succeeded in maintaining the quality of the
search results as Google has grappled with the mass of new online information
and new ways of searching.
His departure coincides with a wave of investment in
artificial intelligence at Google, both to boost the overall quality of its
core products as well as support a push into newer fields such as driverless
cars and robotics.
The projects Mr Giannandrea has been in charge of at
Google include RankBrain, a neural network — modelled on how layers of neurons
in the brain interact — that tries to use machine intelligence to improve the
ranking of search results without the intervention of human programmers.
Hopes pinned on developing system into a powerful and
profitable set of services
Despite the growing importance of AI, machines are not
likely to seize control of Google’s search algorithms for the foreseeable
future, said Mr Sullivan. The technology has been put to use by Google to
better understand what searchers are looking for, he said, but it is still a
stretch for it to predict which piece of information people will find most
useful.
Nevertheless, the rise of an AI expert to head the
rankings group points to an underlying shift in the technology that could one
day see much of the job of search ranking left to machines, said Greg Sterling,
an analyst at the Local Search Association in the US.
The technology shift might one day leave Google’s search
engineers facing the risk of being put out of work by intelligent machines as
workers in other fields, Mr Sterling said, though he added that there was
always likely to be some level of human oversight.
“There will always be some sort of human oversight or
spot-checking,” Mr Sterling said. “There will always be someone in the control
room.”
Mr Singhal has long used the fictional intelligent
computer on Star Trek as the model for the perfect search engine. In a post on
Google+ announcing he was leaving the company, he said he had first dreamt of
the idea when growing up in the Himalayas. “My dream Star Trek computer is
becoming a reality, and it is far better than what I ever imagined,” he added.
Comments
Post a Comment