Chrome is reading my mind. When I type a search query in Google’s Web browser, it offers me the most likely results right in the address bar. It doesn’t just offer suggestions for keywords, as other browsers do, but actual results—when I type Virginia W, there’s a link to the Wikipedia entry on Virginia Woolf right there in the drop-down list. Google Instant, the search engine’s system to update Google’s results as you type, is also baked into Chrome. As I type Virginia, Google results load up with each character I press. I don’t even have to press Enter.
Chrome has had these features for a while, but now it’s started doing something even more amazing. When I type a search query, it “preloads” the first Google result into its memory. Say I’m looking for that old Slate article in which I called Chrome “the best Web browser on the planet.” I type manjoo chrome slate into Google, and the article I want appears at the top of the page. If I clicked on that article in Firefox or Safari, the Slatepage would take four or five seconds, maybe more, to load up. But because Chrome knows that I’m likely to click on that first link, it’s already downloaded all the content from the Slate page. As soon as I click, the page appears instantly.
Taken together, all these mind-reading features make for an incredible experience, one in which I hop from page to page as if I’m on speed. Once you get used to Chrome’s manic pace, you begin to feel lost without it. After using Instant Pages for a few weeks, I get annoyed when I switch to other browsers—all those precious seconds, wasted! I feel particularly pained when I browse the Web on my phone—preloading pages would be especially useful on these devices, which have slower data connections to begin with. Google has said that it plans to bring Instant Pages to Android. I’m hoping Apple, Microsoft, and HP copy the feature for their devices, too.
But I’m not just tickled by Instant Pages as a feature. I also like the philosophy behind it—the idea that my software is analyzing what I do and adjusting its behavior accordingly. If Chrome can read my mind—if it can offer me search results before I’ve finished typing my query, and load up Web pages before I even click on them—why doesn’t every other app?
I spend half my waking life using computers, and most of the time I do lots and lots of repetitive tasks. I’m sure you have a routine, too. Every morning, you might load the same few applications, or open the same few documents, and arrange your windows in the same way. When you open a new spreadsheet document, you might create the same few table headings, and when you start a Word document, you might choose the same font and paragraph settings. Or look at iTunes—you may have tens of thousands of songs, but there’s a good chance that there are just a few dozen you listen to again and again, and perhaps there are some that you’re more likely to listen to on Fridays than on Mondays.
Perhaps you’re not bothered by all these redundancies. Loading up programs and arranging windows is just part of the routine of modern work. It’s also true that most software offers ways for you to customize the settings to avoid repeating yourself—you can decide which programs load up when you start your computer, for example. But often that’s a lot of extra work. Why should I have to tell my computer what to do? After all, it’s watching me all the time. Why can’t it start anticipating my needs rather than simply waiting for my commands?
This is the kind of thing I’m imagining: Once or twice a month, I work on my finances in an Excel document. Along with the spreadsheet, I also load up the websites of my bank, my bill pay service, and Mint.com, and then I arrange the windows across my two monitors in a way that lets me see most of the important stuff in one glance. The whole thing takes about a minute and a half to do. It’s a bit of a pain. An intelligent operating system would notice this pattern, and it would step in to help me out. After loading up the spreadsheet, it might bring up an alert saying, “Hey, you’ve done this before! Click here and I’ll load up everything else you need.” You click, your windows come up and get arranged on the screen, and voila, you’ve escaped Groundhog Day.
There are several other things that my computer could do for me. My online calendar keeps track of when I’ve got to interview someone over the phone. This usually involves calling someone via Skype, and typing up notes in Textpad. Instead of just telling me I’ve got an appointment, then, the computer could load up and arrange the necessary apps on the screen. Or how about helping me out with e-mail? Every couple days, someone asks me for my postal address—why can’t my machine notice these messages and offer to insert my address into my reply? And then there’s travel: Last month I booked a flight online. The computer was right there watching me. It knows the date I’m flying. Shouldn’t it offer to check me in to my flight?
You might worry that the computer’s constant suggestions could get annoying. They could—there’s a fine line between helpfulness and obsequiousness. (Remember Clippy?) But I’d expect the machine to skirt that line by learning my preferences. I’d want it to notice when I decline its suggestions, so that it doesn’t keep offering to do the same thing over and over again.
Is this sort of intelligence plausible? Of all the tech giants, Google seems the most interested in using data mining to make predictions about user behavior. Beyond its search engine and browser, there are mind-reading tricks in Gmail, too. (Type in one e-mail recipient, and it offers suggestions for others whom you typically e-mail at the same time.) Since Google now makes two operating systems—Android and Chrome OS—it’s likely that we’ll see this philosophy extend across entire devices soon. The real coup, though, would be if Microsoft began building intelligence into its operating system. There are more than a billion Windows users worldwide, and we do the same damn things every day. Steve Ballmer, we need your help.