For many years Artificial Intelligence (A.I.) has been hyped as the world’s next big groundbreaking technology. But till recently, the closest we’ve got to seeing A.I. in action was on the big screen. From evil computer HAL 9000 in 2001: A Space Odyssey, to Scarlett Johansson’s breathless A.I. in Her, there’s been a rich tradition of artificial intelligence in the movies. Now, thanks to several recent innovations, A.I. has become a reality in our daily lives. You may not even be aware of it, but whenever you use Google, Facebook, or Amazon Echo, you’re using A.I.
All of the major Internet companies of this era have made heavy investments in A.I. The “big five” of the Western world – Google, Apple, Microsoft, Facebook and Amazon – have been snapping up leading A.I. scientists and pouring money into A.I. initiatives. Of the five, Google has gone deepest. This has big implications for our future, because Google is actively developing what it calls an “A.I. first world.” Our masters from Mountain View believe that the popular Internet devices of today – most notably, the mobile phone – will slowly be usurped by ubiquitous A.I. services. In effect, A.I. will become a utility, much like electricity or water. Which means it’ll be available anywhere and at any time, like switching on a light or turning on a tap.
So why has A.I. become so prevalent now? It’s due to several intertwining trends in recent years: neural networks, machine learning, and big data. In a nutshell, a neural network is a computer architecture that mimics the way the brain works, machine learning is a computer teaching itself rather than humans telling it what to do, and big data is the zettabytes of digital data now available to learn from.
Google’s version of an A.I. first service is called Google Assistant. In its current iteration, Google Assistant uses the chat interface as its primary design paradigm. Using the service is just like chatting to someone on Facebook Messenger or iMessage, except you’re chatting with an automated bot. But don’t be put off by that, because Google Assistant’s design will evolve over time. In future you may simply talk naturally to a device, like your TV or your coffee maker, and Google Assistant will do what you ask without needing to reply. Or even better, do it before you even ask. The A.I. will be the electricity in the background that makes all your devices work.
Another way to look at Google’s device-less vision is that in the near future, the best technology will largely stay in the background. That’s different from the current era of the Internet, in which fancy devices have been the main attraction. From the iPhone in 2007, to the iPad in 2010, to Amazon Echo in 2015, to Facebook’s Oculus Rift in 2016, this era has been dominated by shiny new devices that have captured our collective imagination. While Google has more than kept pace with these developments – its various Android devices have been enormously popular – its real strength is the power of its back-end. And increasingly, A.I. is the foundation.
We got a glimpse of the future of A.I. late last year with the vast improvement, in a matter of months, of Google Translate. After its back-end was overhauled with A.I., Google Translate “suddenly and almost immeasurably improved.” Yet the front-end changed little. That’s because it’s relatively unimportant in this emerging consumer A.I. era. Indeed, the ultimate goal of Google Translate is for its user interface to disappear entirely. That’s already started to happen. If you go to Gmail, it takes just one click to translate something; there’s no typing required. A little further into the future, you won’t even need to click anything. Or look at a screen, for that matter. Imagine talking to someone who doesn’t speak your language, and Google Translate doing an on-the-fly translation into your ear. That’s what Google means by an A.I. first world.
By the time we get to seamless translation, Google Translate will probably no longer exist as a standalone product. It’ll just be folded into Google Assistant. Or, even more likely, the whole shebang of Google A.I. services will simply be known as…Google. Which perhaps its founders foresaw all along. As far back as 2002, Google co-founder and current Alphabet CEO Larry Page is reported to have said, “we’re really making an AI.” With the benefit of hindsight, the rest of us can see what Page was on about. Google search, the company’s core product to this day, has basically evolved into an artificial intelligence machine. It’s that smart, and soon it’ll be ever-present.
Of course there are dangers to having an A.I. first world, particularly if one company – like Google – becomes the dominant supplier of artificial intelligence. To get a glimpse of how this may play out, we just need to look at an equivalent scenario today: the dominance of Facebook in how news is disseminated and consumed. Since the election of Donald Trump, it’s been revealed how his campaign used “dark posts” on Facebook to convince people to vote for him. Dark posts are narrowly targeted Facebook posts, designed only to be seen by a specific audience. The Trump campaign delivered such posts to key swing demographics, in order to both promote its candidate and spread fake news about the opposing candidate. It was very effective, in large part because Facebook has become the primary way that many people get their “news.” While it’s hard to see how Google A.I. might become politicized to that extent, it does show how popular Internet services can (and almost always are) used for nefarious purposes as well as good.
The biggest danger though is that, over time, we will be conditioned to rely on A.I. too much. We already have an over-reliance on Facebook to deliver us news of the day – and look where that’s landed us. If we let A.I. services tell us what to do all day and every day, there is a similar danger that our human judgement will wither and erode. Just as our ability to pick out fake news from real news has eroded over the past few years.
Remember Google’s original motto? “Do no evil” will be severely tested in the years to come, as its “A.I. first” vision becomes a reality. Yet despite all the Hollywood fables about A.I., it won’t be computers (or Google itself) that cause evil things to happen. It’ll be what we humans choose to do with it. It’s up to each one of us to take responsibility for how we use technology. Today, it’s about not letting fake news fool you. Tomorrow, it’ll be not letting Google’s A.I. run your life.