To start off 2010, we're invited to join a new “Hot Topics” feature, which will periodically focus the bloggers on a particular topic. The first Hot Topic is this:
The Upcoming Decade in ScienceWhat will the new decade bring in the world of science? What will happen by 2020 in your field?
Making predictions is always a dicey thing, and perhaps most so in the field of computer science. The good thing about knowing that is that one can pretty much “wing it”. If one gets it right, it’s cool. If not, well, no one really expected one to.
So let’s put on some wings, and see whither we fly.
A decade from now, most of us won’t be “using the computer,” in the sense that we do now. That is, there’ll be few desktop machines, few people will think about going to the computer, and there won’t be a sense of logging on, or “getting on the Internet.” Instead, computers will be part of everything we use, and the Internet will just be there for us.
We’ve been making a start on that over the last few years, with smartphones, culminating in the iPhone and its imitators. It snags a network where it can, be it cellular or WiFi, and the user can do what she wants with the Internet. My BlackBerry, in addition to acting as a telephone and keeping my address book, has my calendar, to-do list, note pad (text and voice), and email. It’s also a web browser, a GPS device, a music and video player, and a camera (albeit not a good one). It has applications for all the instant-messaging systems I use, and there are also Twitter and Facebook apps for it. It’s an alarm clock and a calculator. It, as we’d have said in earlier days of computing, does everything but eat.
We’re moving more in that direction with netbooks and, taking the stage at the recent Consumer Electronics Show, smartbooks, lightweight laptop computers that are optimized for Internet operation. Taking advantage of cloud computing, these devices, even more than the smartphones, assume that the Internet is always there. They get their applications and services from the Internet, and they store their data there.
But these are still Internet-specific devices that we pull out and explicitly use. They represent the first half of the decade.
Still in the research stage is the Internet of things, the concept that everything around us is part of the Internet, and it’s all connected. The full-blown Internet of things includes appliances and other “things” that are obviously devices — your television, stereo system, alarm clock, coffee maker, and car will be part of the Internet of things. But it also includes things you wouldn’t expect, things that are not “devices” in any sense we’ve thought of before. Individual items, such as books and magazines, articles of clothing, and cans of soup may eventually be “connected,” in the sense that they’ll be addressable objects that the active parts of the Internet can interact with.
Ten years ago, when IBM Research started moving toward this sort of idea, I posed the following scenario:
You’re invited to an early-morning meeting at your company’s main office next Tuesday. You accept the invitation, and it’s added to your calendar.
When Monday evening comes, you’re watching TV, and just before your favourite program starts (on demand, of course), your TV — which is on the network and is tied into your calendar — reminds you of tomorrow morning’s change from your routine. When you go to bed, your alarm clock (also online) gives you a final reminder, and offers to set an earlier wake-up alarm. You allow it to do so.
You’re awakened an hour early on Tuesday morning, and your coffee maker (it’s online as well, and has a hopper of beans and a water supply) makes your coffee correspondingly early. When you get to your (online) car, the GPS is already set to direct you where you need to go, routing you by the cheap gas station and avoiding a traffic accident that’s blocking one of the main routes.
When you get in the car to drive home, you see a message queued by your refrigerator telling you that you’re almost out of milk. You’re also given electronic coupons for weekly discounts on salmon and chicken, and there’s a message from your husband suggesting pork for dinner (well, not everything can come together perfectly). The heating system in your house, which turned itself down when everyone left for the day, switches back to “occupied” mode to be ready for your arrival — not on a timer, but because it actually has an estimate for when you’ll be home.
Your husband is picking up the kids, and the system has an estimate for when they’ll be home, too. It’s displayed on the “family status” touch screen built into the refrigerator door, which also reminds you that your mother will be stopping by around 8:30. The coffee maker will be ready.
Now, one can debate the usefulness of various pieces of that scenario, substitute delivery for stopping at the store, and so on. But the main point is that all these interconnected things, from the coffee maker, to the refrigerator, to the milk jug, serve to make your day easier, and to make it less likely that you’ll forget something you meant to do. And it’s all still under your control: you don’t have to allow the alarm to re-set itself; you needn’t follow the route the GPS suggests; you might decide not to stop at the store. It’s assisting you, not controlling you.
Events that are of interest to you could show up on the refrigerator screen, and a touch of the screen could order tickets. Another touch of the screen could sent a message to Mom asking her to come at 8 instead of 8:30... or could initiate a voice call on the hands-free system in the kitchen, while you prepare the salmon. When you’re watching television, the program could automatically pause when a call comes in or when you go into the other room for a moment.
This all would have sounded like science fiction not too long ago, but it’s well within our reach in the next ten years.
Of course, it comes with enormous security and privacy issues attached to it. The more connected we are, the more vulnerable we are to electronic break-ins, scams and other trickery, and our own errors. And even within the family, we need privacy and access controls: you don’t want to be alerted that someone’s coming home with a surprise gift, and the kids can’t be buying tickets to every concert in town. Our ability to get the security and privacy right will be what determines how much of this we actually do by 2020.
So there we are: home computing in 2020. We’ll have to wait ten years to see how well I’ve predicted. But what do you think? Is this plausible? Desirable? Where do you think computing will be in ten years’ time?