I/O 2018: Google Bets On Artificial Intelligence

TechI/O 2018: Google Bets On Artificial Intelligence

I/O 2018: Google Bets On Artificial Intelligence

Google is betting on artificial intelligence (AI) to fuel its growth as the search giant unveiled a host of announcements at its much-hyped annual developer conference this week.

Amid intensifying competition, at the opening of Google I/O 2018 the company mapped out its software plans for the year ahead—nearly every one of them included a heightened focus on AI.

Machine Learning Leading the Way

Google’s CEO Sundar Pichai and other top executives revealed the plans from the event at the company’s Mountain View headquarters in Silicon Valley—including the next version of Google’s Android operating system, a new and improved Google News and added elements to its personal assistant.

The most eye-catching changes set out by Pichai overwhelmingly focused on how machine learning—a form of AI—can help Google harness the vast reams of data it collects to personalise devices and services for its customers. Here are the key take-aways from the event:

Android P – The Next Operating System

That’s the codename for Google’s next operating system for mobiles and tablets. The full name won’t be revealed until later this year. But there’s plenty of other nuggets of knowledge that investors can take away from the Android P announcement.

The big updates were all about personalisation. Google is introducing adaptive battery and brightness settings, working with the DeepMind AI business to build the machine learning techniques that will enable more personalisation.

Google is also working on how Android can predict what smartphone users will do next—before they even know themselves. For example: Anroid P will know your location and time, and make recommendations for things to do based on your habits.

Autocomplete Emails

Sticking with the prediction theme, Google also unveiled at I/O 2018 a revamped Gmail email system called “smart compose”. It’s essentially an autocomplete for emails, using AI to suggest phrases as users write emails.

The update is an extension of Google’s pre-existing Smart Reply feature, which suggests entire but brief replies to emails. The smart compose will instead make suggestions for complete sentences as you type them. Email addicts have a reason to be cheerful, then.

AI-Powered Personal Assistant

The personal assistant market is growing rapidly, and Google faces competition from Amazon’s Alexa and Apple’s Siri digital assistants. Again, Google is betting on AI to improve its own bot, Google Assistant. Scott Huffman, VP of engineering at Google, said the firm was adding six new voices to its assistant, as well as enabling follow-up questions to be asked and a children-friendly setting that encourages manners.

Google said its first home assistants, which will come equipped with screens, would be available from July. They are being developed by Lenovo, LG and JBL, not internally by Google.

AI is also powering Google’s smart assistant. A system called “Duplex” will bring together all of the company’s AI efforts—for example its AI voice could phone a restaurant and book appointments. There is no release date for this feature yet, however.

The “Android of Things”

The Internet of Things is already on the up, with billions of connected devices online, everything from “smart fridges” that can order milk when supplies run low to bins designed to combat climate change through simpler recycling.

Google wants a slice of the pie, as it has rolled out the Android Things 1.0. But instead of helping you order milk or recycle, its designed to improve security, which is a huge concern for the IoT.

Android Things will ensure that developers and devices receive Google security updates automatically for three years. It’s closed source, meaning that developers can’t, well, develop it before using it on their IoT devices.

No More Fake News?

Google received its fair share of the blame for the spread of fake news across the web recently. Google News itself provided a platform for many fake news stories rather than real breaking news events. So Google told the I/O conference that it has been working to combat hoaxes.

Foe one, Google is making News more personal by providing “briefings”, including stories that individual users are interested in, based on their search history—using AI.

Google CEO Sundar Pichai speaks on stage during the annual Google IO developers conference.
Google CEO Sundar Pichai speaks on stage during the annual Google IO developers conference. REUTERS/ Stephen Lam

Google is also providing a wider range of news sources for users, amid criticism over partisanship. Certain news events will have multiple articles on the subject, include fact-checking and news from a wider variety of publications. The idea is to encourage people to consume a wider range of viewpoints.

Augmented Reality Maps

Google also showed the I/O audience a version of Google Maps that automatically loads Streetview when a smartphone camera is opened. Google calls this a visual positioning system and it can recognise buildings and landmarks in front of the camera.

AI is also being used here. Google said it will use a system called Google Lens that can use a smartphone camera to scan objects. For example, it can highlight words written on a page and them into digital text. Pretty neat.

This website uses cookies to ensure best possible user experience. Read more