Netflix is a master at it, Google is no slouch either, and Amazon announced earlier this year that it was getting into the business: predicting what you want even before you ask. The data-driven science of anticipatory computing seems to be here to stay, and infiltrating business for better or for worse. Is this just the first step toward artificially intelligent machines that make our life choices for us? Are we selling our souls?
Weeding through the data you leave behind
Earlier this year, Amazon filed a patent for anticipatory shipping. Basically, the company plans to figure out what you want and ship it to you long before you ever order it. While this technology is not entirely new, the fact that Amazon wants to forecast a unique customer’s desires would indeed be a novelty.
How will Amazon know what you will order next? The answer is obvious if you have followed recent trends: data. It’s been reported that Amazon will focus on what you have ordered in the past, what products you search for, which ones are in your shopping cart or on your wish list, and even how long your cursor hovers over a product.
Netflix is seemingly doing something far less invasive than Amazon to predict what you want to watch next—monitoring what you have watched in the past and how much you liked it. But it turns out they are also pushing the limits of what a company is willing to do to get into your head.
To most effectively pitch new TV shows and movies, Netflix has undertaken a tedious process recently revealed in an article in The Atlantic. Apparently, they took every movie and TV show in their library apart by analyzing and tagging them down to the characters’ jobs and moral status. With that data in hand, they are not only in a position to suggest tailored content, but also to produce original content that is exactly what users want. The results are shows like award-winning House of Cards.
Only the beginning of reading your mind
Going beyond the niche services of Netflix and Amazon, Google has taken on the challenge of providing a predictive technology that is with you 24/7 with Google Now, a sort of digital assistant, or data butler, as some call it.
Google Now is fed with information from multiple sources like your calendar, email, current location, and more. In addition to gathering relevant data, it has the ability to learn your habits, making it even more powerful. Anyone else reminded of the movie Her?
Google Now can learn your commute and, based on traffic data, propose that you leave home early to make sure you aren’t late for work.
Google Now fits perfectly into the future of the company, which revolves around artificial intelligence: A computer-centric world with digital assistants, self-driving cars, and various gadgets that make use of Google’s tremendous stockpiles of data destined to make our lives easier. Google’s recent acquisition of the Nest smart sensor company is another testament to this.
But back to prediction. Startups like San Francisco-based MindMeld and Cover are also based on anticipatory computing. MindMeld is an app that listens to your conversations and provides relevant sources, such as websites or maps, before you ever run a search. Cover changes your smartphone’s home screen whenever you change location, displaying the Netflix app when you are at home and the Salesforce app when you are at work, for example.
Meanwhile, Silicon Valley heavyweight Facebook recently revealed that it can predict when two people are about to enter a relationship. They also began experimenting with so-called context cards that pop up in your mobile feed with relevant comments from friends based on your status. “Jane likes the club sandwich” would appear at that restaurant you just checked in at, for example.
Dennis Crowley, the CEO of Foursquare, said that his company is focusing strongly on anticipatory computing as well, mapping interests by collecting data on which restaurants users like, when they eat, and what they order to help provide suggestions.
According to Fast Company, the four megatrends will accelerate predictive computing even more are increased use of mobile devices, sensors everywhere, cloud computing that makes data easily accessible anytime, and of course Big Data.
Is prediction a blessing or a curse?
It looks like anticipatory computing is here to stay, but there is an elephant in the room. Without a doubt, the technologies outlined above are considered by some to be a serious invasion of privacy. The question is, will these privacy concerns slow down the trend or will the convenience factor win out?
Others warn that by allowing companies to shape our intents instead of guessing them, we have never been closer to selling our souls. But venture capitalist Om Malik argues that this new wave of computing will simply declutter our lives by learning our behavior and anticipating our needs, and that most of us will gladly trade convenience for security squeamishness.
The concerns are valid, but if anticipatory computing is really the next big thing, the players involved could very well decide to put their efforts into making sure that the new technologies will please people and not alienate them. Google has reportedly even set up an ethics committee to ensure that intelligence technology isn’t abused.
The creation of this committee may well be related to Google’s recent acquisition of the artificial intelligence company DeepMind. Is that where anticipatory computing is going, towards machines that can think on their own—and think for us? Perhaps.
But before computers can replace humans, a lot still has to happen. Siri’s isn’t nearly as human as the Samantha operating system in the movie Her. Google Now seems to offer public transportation for places that are just around the corner. These technologies clearly need to step it up a notch if they are to become more trusted and widely used—dare we say loved.
Even if these technologies get significant improvements, such as better interfaces, we might never fully get there. At Stanford University’s Future of Media Conference in March 2014, the role of anticipatory computing was discussed in the scope of content curation. While content discovery will probably be based more and more on anticipatory computing, no algorithm will ever replace humans.
A computer can’t detect humor, nor can it detect early on if an article is bound to go viral. At best, it can develop a very close approximation, by closely monitoring the data we leave behind.
While anticipatory computing seems here to stay, the human factor is something that will probably never be replaced. So go ahead, Amazon, order my next shipment. Netflix, tell me what I’m going to watch. But you never know, I just might change my mind and do something unexpected that you’ll never be able to guess!