Key trends from Google I/O 2018

 | 

Alice Cruickshank

More than 7,000 people attended this year’s Google I/O 3-day conference, providing a not-so-sneak peak into Google’s upcoming developments.

Streamed worldwide, Google I/O, which stands for the web developer term input/output, has been taking place for 10 years at the Mountain View headquarters in California.

This year’s event has confirmed what many already predict to be the biggest technology trends of the next few years: artificial intelligence and augmented reality.

 

Artificial Intelligence

Such is Google’s belief in Artificial Intelligence that the company announced it is building AI centres around the world. In fact, almost all of its software will be given an AI update in 2018, from Google Assistant to News and Photos.

Google isn’t planning on keeping machine learning innovations to itself. A new software development kit that hopes to make machine learning accessible for all mobile developers will be available soon. The ML Kit will help developers use application programming interfaces including text recognition, face detection, barcode scanning, image labelling and landmark recognition – excellent news for companies looking to develop their own apps.

 

Google Assistant

Google Assistant, the company’s voice activated technology, was only announced two years ago, but currently is used on 500 million devices worldwide. AI is being used extensively in Google’s mission to make Google Assistant more naturalistic, and able to understand increasingly complex requests.

Google Home

Google Assistant is currently used worldwide in Google Home hardware as well as mobile phones, Google smartwatches and Pixelbook laptops

The famous “OK Google” required between each question will soon be no more, as the company launches Continued Conversation. Users will also be able to ask multiple things at once thanks to new Multiple Actions technology. This means owners could be asking their Google Assistants to turn off several lights in the house at the same time or ask complex multi-clause questions such as: “what is proptech and what is its UK market value in 2018?”

Advances in AI technology mean Google will be providing six new voices for Google Assistant instead of the existing rather robotic sounding ‘Holly’.

The newly revealed Wavenet technology closely mimics human speech in pitch, pace and intonation, and Google is aiming to cater for accents and dialects globally in the near future. Google CEO Sundar Pichai revealed the unexpected announcement that soon users will be able to hear the voice of singer John Legend speak to them through their Google Assistant devices. Very smooth.

Google Duplex Speech Demo

Demonstrations showcased the ability of Google Duplex to imitate a human voice

From the soothing to the space age: prepare yourself for ultra-realistic AI phone calls. The jaw-dropping Google Duplex technology will provide the Assistant with the capacity to book reservations for owners via telephone using naturalistic language. In examples demonstrated at I/O, Google Assistant even threw in ‘ums’ to make its conversation sound realistic, and was able to respond to unexpected or unusually phrased questions. The technology is still in testing mode, and Google are keeping information on it fairly guarded for now.

 

Augmented Reality

Google Maps has been trialing augmented reality using a combination of Street View and the smart phone’s camera to guide pedestrians. With the AR tech, the directions will be shown in front of you by layering them over the real world view. On the example shown in the keynote speech, information about buildings as the user passes them is available, making phone cameras portals of information. There is no release date for the augmented reality tech as yet.

Visual search tool, the Google Lens app, which was unveiled last year at the I/O conference, will also see some ambitious AR updates. It will now recogonise real world text, meaning you can ‘copy and paste’ text from a document in front of you into a message. You’ll also be able to take a photo of a document and convert it to PDF – an announcement the crowd at the keynote speech were particularly impressed with. The exciting text recognition tech will be rolled out to Google Photos users within the next few months.

Your comments

Read our comments policy here