Want to save a few bucks on insurance? If you let Progressive track your driving, you can get a discount.
If you’re Japanese, want to be part of a healthier society? By tracking your activities, your company can help meet the national goal of improving health.
Sound like Big Brother? Does it creep you out?
These days, maybe. (OK, I’m going to exclude the terminally self-obsessed tweeters who volunteer their every breath to anyone who will listen… they’ll never read this past 140 characters anyway.)
But will that always be the case?
And I’m not talking about a time when we throw up our hands in surrender and simply accept the fact that everything we do is being tracked. I’m talking about a possible time when we actually decide that we’re getting a benefit from it.
At a session sponsored by the IPSO Alliance a few weeks ago, Bo Begole, a Principal Scientist at PARC (yes, that PARC – the one where the Macintosh was invented*; it’s still around, with the express goal of getting things into real products) gave a presentation on this whole notion of gathering information and doing something with it.
As the technology required to make use of this sea of information evolves, one of the foreseeable stages is the point when we figure out how to make context-sensitive intelligent decisions based on all that data.
And my initial response is to scoff. (You should try it sometime; it can feel good. Makes you feel really superior. I recommend it.)
I scoff based on my general experience with things that watch you and try to decide what you’re doing and then be helpful. Dare I bring Clippie into the picture? Yes, he’s quite the shape-shifter, but, for the most part, he’s completely clueless about what I’m trying to do. And annoying.
And I realized, listening to this talk, that this is a realm that has its own “uncanny valley.”
For those of you not familiar with the concept of the uncanny valley, it was coined in 1970 by robotics professor Masahiro Mori. The idea is that, with entities that represent real people or animals or whatever, if you make them look too realistic, you run a risk of creeping everyone out. This can apply to robots or cartoons or any other facsimile of a real living thing.
Take cartoons. Disney representations aren’t realistic; clearly, they’ve been accepted willingly and lovingly, to the tune of many billions of dollars for Walt and his kin. Pixar provided a new level of life-like animation, but it still wasn’t hyper-realistic. And again, it worked well.
And then came the example that everyone trots out to illustrate uncanny creepitude: The Polar Express. It got soooo close to realistic, but missed. So instead of real-looking cartoons, you have living entities that have something wrong with them – they might not even appear living. The eyes are dead, or perhaps that natural facial asymmetry isn’t quite right, or… something you can’t quite put your finger on.
Whatever it is, it creeps you out. All you know is that you want to get away from that conductor.
That’s the uncanny valley. The more realistic you get, the better response you get – to a point, at which you enter a forbidden zone where response plummets until things are perfect.
There’s a similar effect with offers of assistance. I guess part of it is that whole thing your dad probably said when you tried to help him as a kid: “I’ll be faster just doing it myself.”
It’s hard for someone that doesn’t really understand what you’re doing to help you. Adding a helper can actually make things worse until the helper achieves a high enough level of skill. And, not to pick on a frequent pickee, but Microsoft… well, you know, they really do seem to try to help, but man…
Just start typing in Word, and they might go, “Oh, gee, this looks like [some specific document type they heard about somewhere], and the next thing you know, boom! They’ve reformatted everything to help you. Which you then have to go undo. But at least you can undo it.
Even worse is Visio, when you try to place the end of a line at a very specific point, and Visio is absolutely sure that it has a better idea where you want that line to go, and it absolutely will not let you put it where you want, no way no how.
Those are old examples, and people may just write that off to Microsoft being Microsoft.
But imagine now that there are sensors all over the place gathering reams of information about what’s going on and intercommunicating it (the IPSO guy would insist I mention here that the communication is happening over IP… but, between you and me… it might not be… shhhh…) so that some entity can build some kind of a picture of what’s going on.
What’s it going to do with that information? More importantly, how will you feel about what it does? Will you find it helpful or annoying?
Today, most attempts at this are annoying (or at least ineffective). It’s like the early days of internet search: you got lots of hits, but little that was relevant. It took quite some years and changes of technology before internet searches really became useful. In the meantime, internet search was often worse than no internet search.
Even old tech hasn’t gotten it right yet. All of the marketing databases and data mining capabilities are billed as good for us because it lets merchants target us more specifically – look at how much junk mail we’ll stop getting! So… why has junk mail increased? (Because it’s now cost-effective to do junk mail when it wasn’t before.)
Do you find that you’re now getting more relevant spam in your inbox? How often do you click on a web ad? Are you seeing what’s relevant to you? How many Failblog posts have there been where a news article on, say, an airplane crash is paired with an ad for an airfare special? When you do a Google search on, say, “pestilence and misery,” do you get some returns promising “the best prices on pestilence and misery”?
The reality is that the amount of information and the algorithms really needed to figure out what someone wants in some situation (“Please, I want no ads” doesn’t seem to be an acceptable desire) is enormous, and we really haven’t gotten it down yet.
So as we get more and more information, our feeble attempts to look like we’re doing something useful with it take us further and further into this uncanny valley. And “uncanny” is really the wrong word. “Annoying” is really a better one (if occasionally understated). People are trying their darnedest to “make offers” that we want, and 99% of it is just, well, annoying.
Not all of that is a result of early technology; some of it is under the control of marketers. I was talking today with Stuart Taylor of ViVOtech, a company focused on near-field communication (NFC) technology. He gave an example of one NFC use model where you might walk into a store and take an interest in a poster that has some useful information. By tapping the NFC-enabled poster, you could register your interest.
But what happens with that information? The marketing program could selectively target one or two things – say, a coupon or two, or perhaps a helpful hint – and send them to your phone. Or, a more aggressive marketer could say, “AHA! Got an opt-in! Sign him up for every newsletter, and forward him 50 coupons for things we think he should want.”
The first case would probably be appreciated; the second… not so much. So part of a marketer’s positioning with respect to the annoying valley is under his or her control.
But when you expand your view to a vision that has data gathered everywhere for use in much more specific ways, the problem gets harder. My examples have all been focused on advertising, but, let’s face it: at this point, value propositions for these types of technologies talk primarily about better ways of target marketing messages to you. Perhaps a sign of the technology maturing will be a migration into areas other than sales pitches.
If, for example, your car is going to interpret what it is you want based on some sensing of something, you really want the car to do a better job than Word does. If you’re closing on the car in front of you and the car doesn’t think you’re paying attention, then having the car do some braking on your behalf probably makes sense. If, on the other hand, you’ve put a destination in your GPS but you’re trying to take a secret shortcut, having the car say, “No, you’re going the wrong way; I’m going to take us this way instead” would probably be moderately infuriating.
I believe that this is going to be the challenge that determines the point at which people become more comfortable being tracked and having decisions or suggestions made for them (setting aside privacy and data control issues, which are separate – and also important). The benefit of the technology has to be clear, and the technology has to deliver on that benefit reliably.
In the case of Progressive Insurance, it’s not so complicated. You have a clear choice and potential benefit. If you want to spend less, you can let them track you. If you don’t want to be tracked, then you can decline the offer.
On the other hand, the Japanese health initiative is a bit fuzzier. Obviously, cultural differences apply here, but this tracking may not result in a specific benefit to you – no surprise, since it’s done for a societal, not an individual, benefit. Do the results mean that perhaps you get better information on exercise options? That’s not bad… Does it mean that your boss berates you for a sloth-like demeanor? That’s not so good. Does it mean that, when you sit down to order a sizzling, juicy tonkatsu, your order gets replaced with a salad? That’s truly terrifying.
If the technology gives us a benefit and the users of the technology don’t screw up the benefit, then we all might be a bit more OK with it. But for now, there’s something of a climb ahead of us as we try to escape this annoying valley.
*KIDDING! Only the mouse and graphic user interface, for which the Mac gets credit, were invented there. Well, that and lots of other cool things from which Xerox never got much tangible benefit…