“We are on the cusp of one of the biggest technological advancements in recent times – machine intelligence,” Thomas Davies, Director of Google for Work EMEA, told delegates on the first day of UCExpo 2016.
Whereas over the last 20 years, developers have created code that always had an end point, machine intelligence is “open-ended”, which means that applications can now ‘learn’ themselves. “If you don’t define the outcome, then you get very different answers, which can lead to endless possibilities,” said Davies.
He described how machine intelligence will take away the ‘mundane’ daily tasks of workers, and automate them, but could equally be used to control cars or tell us the state of our personal health.
Davies described how Google has been working on over 100 projects for the last three years. He used Google Translate as one of the first examples of a self-learning app. “It gets better every day without us having to do anything,” he said.
However, he outlined four examples of machine intelligence apps launched in the last six months by Google that are all automating daily tasks in the enterprise, freeing up workers to be more productive.First, Smart Reply is an app that works out whether it ‘could’ and ‘should’ reply to emails and chooses an appropriate email response from tens of thousands of pre-defined responses. The user can then choose to edit the chosen response, or change it, and will learn as it goes along.
A second app, Voice Typing, is “amazing”, said Davies. While users might have tried it in the past, he said, and thought it “not very good”, it’s self-learning element means that it is now an amazing piece of technology, he told delegates.
Explore, in Google Sheets, meanwhile, offers a button that provides automatic insights and analysis into spreadsheet data. “You might not be a data analyst, but know that some of your data is really important. You can now click a button and the app automatically provides analysis for you, such as trends, insights and so on, based on the type of data you have.
Finally, two weeks ago, Google launched Goals. Recent research showed that when we make a resolution to exercise more, only 38 per cent of us will stick to it for 30 days. Surprisingly, a ‘motivation’ group that was being constantly motivated and told how lack of exercise was bad for them, actually fared even worse, with only 35 per cent still sticking to their regime after a month.
However, of those who scheduled exercise into their calendar every week, the ‘intention group’, an amazing 95 per cent stuck to their exercising regime for more than 30 days. “Goals learns your habits and calendar schedules and automatically schedules exercise time in for you, meaning that you are significantly more likely to stick to your goals,” explained Davies.
Google has open sourced many APIs for self-learning apps such as voice, images, speech, recording, and so on, he said, so that others can integrate them into their own products. “It’s now in your hands, you can do anything you want with it,” he concluded.