Google I/O 2018 Summary — Don’t be evil… be AI first
Matt Payne, from Tribal’s Tech Engineering Practice, gives you the highlights of the Google I/O event and what it might mean for your brand in the future.
Bringing together developers from all around the globe, Google walked us through the next generation of tech at the Google I/O conference. The one thing everyone needs to know, if you didn’t already, is that Google is an AI company. Nearly all updates/services Google discussed had AI elements within them.
CEO, Sundar Pichai, kicked things off and got straight down to business by addressing some important bug fixes: restoring the laws of physics by fixing the floating foam in the beer mug emoji and moving the cheese from the bottom of the bun to sit on top of the meat in the cheeseburger. Hopefully, that keeps them from being ridiculed by Late Night TV shows.
After getting those blunders out of the way, Pichai then went to reveal some big announcements coming out this year — including John Legend lending his voice to Google Assistant, Google Lens using AR, Gmail writing emails for you, Google Duplex using natural language to make calls for you, Google Maps becoming more social, and of course the announcement of the new Android P — just to name a few.
With AI being the common thread throughout the entire conference, we’ve summarised some of the key highlights and what this could mean for your brand below.
Google Assistant — six new voices, continuous conversations and natural language.
Deep Mind focused on bringing natural language into Google Assistant using Wavenet — a generative model for raw audio. It is rolling out six new voices that mimic the pitch and pace of human language. For brands developing any Actions for the Google Home, you can select which voice you would like to use. This is the first step in having voices auto-generated based on user preferences. Also, with the help of Google, the one and only John Legend’s voice is coming to an assistant near you later this year.
More natural back and forth discussions are also coming. With only one invocation name “Okay Google,” it can now hold a discussion for longer without having to activate it every time. Even being able to detect when a user is speaking to it or someone else. Natural discussion with voice is an area that Google is really owning. No other voice AIs has come close to matching what they have to offer.
With all of these new introductions to voice, one thing is clear, voice development will be the new browser annoyance for developers. Having to build a single core, that needs to be continuously updated with new natural language variations, smart displays and interactions for different voice devices will be key to getting your Brand’s Action / Skill working correctly for all users (much like having to get your site looking right on all the variations of browsers).
A big improvement has also been made to Google Routines. Fully customisable so you can play music, control your smart home, and get your daily briefing. Previously limited to certain features and phrases, you can now group commands together into a single trigger phase.
Duplex. Smart AI conversations and making calls for you.
Google’s latest bit of technology named “Duplex” allows for the next gen of AI assistance. Even bringing pauses, um’s and ah’s within the speech. All to give a more lifelike discussion between you and the machine. It can even handle ambiguity, for example, it can call a restaurant to make reservations on your behalf, then be told that they don’t take reservations and let you know that it hasn’t been able to make one.
This is all part of an experiment Google will release this summer to a limited number of users. You will be able to have a discussion with the AI assistant and be able to book a reservation, a hair appointment or even check holiday hours. All of which happens in the backend via the AI with it even making the calls for you. To protect your privacy however, the calls will not be coming from you, but will be handled by Google.
Google Lens.
A new beta version of Google Lens will be built into the camera app itself (it currently resides within Google Photos). The beta will be released within the next few weeks. It will be built on 10 different Android devices, starting with the Pixel 2 and the LG 7 ThinQ. The new Google lens includes features such as:
- Smart text — Allowing you to copy and paste analog examples into the digital word.
- Style match — Buying and finding reviews online based on the image that you have captured.
- Real-time results — Users no longer have to specifically select an item and wait for results to load, Google will pro-actively serve you information and results are anchored to the object as you move your camera’s viewfinder (as shown below)
A huge opportunity with this update lies in retail. With Google’s state-of-the-art real-time processing, it is able to pull in reviews, instructions, and places for you to buy items. Imagine pointing a camera at new piece of furniture and then it tells you how to put it together. Or pointing the lens at a blouse and it shows you where to buy it or pull up relevant reviews. The tech is not just able to recognise each item and brand, but can also provide shoppable links for everything as well.
A great mashup of Lens and Google Maps is also coming. This will see AR being added to Street View and help you to navigate in real-time. It also will surface reviews, and point out common places of interest such as restaurants, parks, etc.
Android P — the next gen for Android.
Whilst a developer preview was released in March this year, at Google I/O we saw the announcement of the official update. So along with a fresh new look, there are additional features such as:
- Adaptive battery — Using AI to help optimise phone performance and power consumption.
- Adaptive brightness — Using AI to help control the brightness of your screen.
- App actions — Helping to predict your next task.
- Digital wellbeing — A dashboard which helps you limit the amount of time you spend on your phone, on certain apps and also helps auto-adjust the brightness; as well as new features such as wind down mode.
Gmail — going social and writing emails for you.
Gmail can draft emails for you. By itself. Well almost. Google’s Smart Reply feature is expanding to include Smart Compose. This uses AI to help draft emails from scratch. Currently, the feature only makes recommendations. Although, after looking at the breadth of Google’s AI vision from across the event, it is obvious that in the near future we will relinquish all control to the AI.
Whilst the Google I/O event is still ongoing and other updates have been released, it is obvious that Google plans to bring the AI fight to the other tech giants — introducing it into all aspects of digital and analog life. However, in order to make it more mainstream, they will still have to work on winning over more consumers, proving the usefulness of its products to their everyday life.
In summary, partnerships will be key for brands, from everything from automotive to TVs and fridges. Brands need to start thinking about how they can survive in an AI world, how they should adapt their services and products, how can they utilise voice and machine learning to help drive product improvements, social interactions, and customer service. As Google is apt a proving, simplicity is key.
