My Photo

Your email address:

Powered by FeedBlitz

April 2018

Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30          
Blog powered by Typepad

Become a Fan

« Asserting Context: A Prerequisite for Smart, Sensemaking Systems | Main | The Christmas Day Intelligence Failure – Part I: Enterprise Amnesia vs. Enterprise Intelligence »

August 16, 2009


Feed You can follow this conversation by subscribing to the comment feed for this post.

Laura McClure

Orwell... wow

Brand Hunt

I think this is great. Someday AT&T service reps will be calling customers as soon as they walk up to a T-Mobile counter near the end of their contract. For consumers it'll yield better choices from vendors (and less waste in marketing.) Ignore the downside!

Allan Engelhardt

I have been using everybody's favorite search engine and I can't find any information to substantiate your claim that the operators are sharing their detailed network data in any systematic way (other than for pure research -- and we did some of that so that is happening). Google's service is different, of course, I am talking about your "cell providers" here. Asking internally (here in Europe) has so far drawn a blank.

Can you provide any more information / references / links?

Allan Engelhardt

Jeff doesn't go far enough in one respect. When he points out that the Blackberry generates a transaction every minute or so it is true. The significance it that these show up in the Call Detail Records (CDRs) which is where most analytics have traditionally been done.

But of course every mobile device checks in to a nearby cell tower ever few seconds: the network and the device is in constant communication and they need to be in order to deliver your calls and messages. (The geographical accuracy is lower on this data.)

This raw network data is big and processing it requires non-trivial hardware (which I suspect Jeff's company will be very happy to sell you).

But it is also an absolute goldmine. I know how long you spend in that shopping mall, in *every* shopping mall, even if you did not make or receive any calls and even if you do not have a Blackberry.

Ian Story

It seems to me that this information shouldn't be made available - any more than it'd be "okay" for the providers to follow you around all day recording your location ;) Or that it'd be "okay" for them to mine your voice calls or data transmissions or addressbook on the phone. Just because it is possible to get this, have I explicitly granted them this access? Of course, that is all nice from a utopian view, but in reality, I don't get any say on this. People would (well, ACLU types at least) get all up in arms if the cell company was following you around or indexing your voice calls (although they don't seem to mind if Google does it to your Gmail), I suspect, that if they realized it, they'd be up in arms about this as well. There are of course other examples of this aside from mobile devices - for instance GM's OnStar system in cars (presuming people still buy GM cars).

What's a person to do though? Short of donning a tinfoil hat (or lead lined cellphone case, which kind of defeats the purpose!), it would take a massive campaign (probably via privacy legislation of some sort) to even attempt to prevent this (and attempt is the right word, as nefarious people will still do what they will do).

Perhaps the best way to combat this is to exploit it for your own gain - IE once you start to see Starbucks handing you your grande vanilla soy latte before you show up, refuse to take it unless they make it half price :) Or once AT&T starts calling you when you approach a T-Mobile counter, use that to negotiate better rates, with no true intention of switching.

I'm sure that for every person that would exploit this information on the backend, there will be at least one like myself that will counter-exploit it on their side - not that it makes it "okay" but at least a little easier pill to swallow.

Jack Repenning

It's not just mobile devices:

For another view of the "Vanilla Soy Latte" story, see "Minority Report."

Stefan Dreverman

And some vital points are missing:

1. Cell-phone time-space data is not the truth! If you forget your phone, give it to someone else or just leave it somewhere, that doesn't mean you were there.
Imagine that courts start accepting this data as an alibi. Disasterous! The stalker in your example simply leaves his phone somewhere else to create his alibi.

2. Too much structure kills/numbs. People could start (or be forced to) living/being their data, eradicating all spontaneous and creative action from their lives.
Imagine police tailing you because you've changed something in your time-space pattern (and that's suspicious, because the data says people generally don't change their pattern in the way you did). You could become afraid to change it again.

Sure, this time-space data can be helpful in some ways, but it surely raises some moral and psychological questions. The debate on how far we want to stretch it should be held with caution. Decisions on where and when to use it should be made very carefully.

Jay Levitt

Great insight, Jeff... CDR becoming space-time-travel analytics is yet another case where, when price/performance curves cross, a change in degree becomes a change in kind.

And it goes oh-so-well with Paul Ohm's point that we have no privacy protections under most current laws and privacy policies:

Carl Youngblood

I think it's also important to recognize that along with increased opportunities for surveillance come increased means of collaboration and dissemination of information. It seems to me that individuals are winning out, on the whole. Political establishments and corporations, even dictatorial ones, are being increasingly circumscribed as they attempt to foist undesired policies on their constituents and consumers.

Paul Rosenzweig

Here's an interesting link to someone who mostly thinks this analysis is correct.

The comments to this entry are closed.