ArrowArtboardCreated with Sketch.Title ChevronCrossEye IconFacebook IconIcon FacebookGoogle Plus IconLayer 1InstagramCreated with Sketch.Linkedin IconIcon LinkedinShapeCreated with Sketch.Icon Mail ContactPath LayerIcon MailMenu BurgerIcon Opinion QuotePositive ArrowIcon PrintRSS IconIcon SearchSite TitleTitle ChevronTwitter IconIcon TwitterYoutube Icon
Technology

Google brings user-friendly AI technology to customers

New service lets low-budget players craft customized systems

Google CEO Sundar Pichai addresses the 2016 Google I/O developers conference in Mountain View, California, U.S.   © Reuters

PALO ALTO, U.S. -- Artificial intelligence and machine-learning models can be part of everyday technology available to businesses with any size budget, even those that lack dedicated in-house engineers, with Google's new cloud service.

The U.S. tech company launched Cloud AutoML, a custom-built AI system creation service, on Wednesday for select customers, with plans to offer it to a wider range of users at an undisclosed price. For now, it is providing a user-friendly service for building an advanced image-recognition system, which can be used for identifying certain pictures based on a set of conditions.

"Currently, only a handful of businesses in the world have access to the talent and budgets needed to fully appreciate the advancements of ML and AI," Fei-Fei Li, Google's chief scientist on AI and machine learning for cloud services, told a news conference. "Our mission is to democratize AI," said Li, who is also a computer science professor at Stanford University here.

Previously, an image search system required up to tens of thousands of pictures and a dedicated program, but a similar system can now be developed by sending many fewer images to the cloud, Google says. For instance, to build a system that recognizes an image with a certain car model out of a vast collection of pictures, a user now needs as few as 100 images.

Engineers previously had to program AI tools and have them learn by reading a massive amount of data. Even with Google's existing, free TensorFlow software, which drastically cuts down the programming process, some part of the work required an expert.

Google, Amazon.com, Microsoft and other cloud-computing service providers offer trained AI models for speech- and image-recognition, as well as for translation, on the cloud. But their functions are limited.

With the new cloud service, Google aims to address the shortage of experts in the field. There are only several thousand AI researchers with advanced knowledge in the world, and fewer than 1 million data scientists, according to the company.

Google trails behind Amazon in the cloud service market, with the leader offering customer-focused services and lower rates. Chinese e-commerce giant Alibaba Group Holding is also pushing cloud services featuring AI technologies, mainly in Asia. Google seeks to capture customers by offering cloud services for AI, an area where it is said to have an edge over rivals.

You have {{numberReadArticles}} FREE ARTICLE{{numberReadArticles-plural}} left this month

Subscribe to get unlimited access to all articles.

Get unlimited access
NAR site on phone, device, tablet

You have {{numberReadArticles}} FREE ARTICLE{{numberReadArticles-plural}} left this month

Subscribe to get unlimited access to all articles.

3 months for $9

Get unlimited access
NAR site on phone, device, tablet

Your trial period has expired

You need a subscription to...

See all offers and subscribe

Your full access to the Nikkei Asian Review has expired

You need a subscription to:

See all offers
NAR on print phone, device, and tablet media