Features

net.wars: Thinking time

If it's Christmas it must be time to come up with program proposals for Computers, Freedom, and Privacy, the 17th edition of which will take place in Montreal, May 1-4. Submissions (via email or the Web system) are due on January 20. It's something to think about after you've finished listening to the Queen's Speech (UK)) or watching It's a Wonderful Life (US). Who needs trivia quizzes when you can think up and propose speakers for CFP?

CFP is a cross-disciplinary oddity of a conference. Others focus more tightly on privacy, data protection, cryptography, software, hardware, games, music, new technology...but CFP is the only one where, as I'm so fond of saying, for four days you never want to finish anyone else's sentence.

The 2007 theme is autonomy, which should include one of the subjects CFP has long neglected, disability. Generally speaking, redesigning anything to make disabled access easier has benefited many other people (curb ramps, for example, help not only people in wheelchairs but those dragging luggage or pushing babies in strollers). But it's been contentious in designing electronic voting systems, especially when you're trying to design a paper trail voters can verify – how does a blind person verify the paper?

It's worth noting, by the way, that the UK, apparently refusing to believe the stories it reads about snafus (PDF) everywhere else, has decided to run trials in 2007. Look for a talk by Rebecca Mercuri on the problems encountered in the last US elections on February 8 in London as part of a conference on e-voting that will attract speakers from Ireland and Italy; it's being organized by Jason Kitcat for the Open Rights Group

The disabled, the elderly, and even the seriously ill, figure in another trend: one of the promises people talk about with respect to ubiquitous computing is the ability to monitor people at risk and make sure they're all right. It sounds so warm and fuzzy: install a bunch of sensors in Grandma's house so you can check in every day via a Web interface and make sure she's still alive. Or give her a robot to make sure she eats every day and doesn't spend all day sitting around in that one stained bathrobe. Is that the life you want when you're 87? (I can hear my mother saying even now, of her old robot that's been replaced, "It never calls, never writes…")

Autonomy is also an umbrella for the many trends that, compared to the glamour days of the file-sharing wars, sound too dry or remote to be dangerous. What could be harmful about putting medical records on a centralised database? Wouldn't it be better if emergency personnel can quickly find out the medical history of an unconscious person – what medications they take, what allergies they have, what their health problems are? Speaking as someone with almost no medical records at all (most of my doctors are dead; those that aren't shredded their records rather than show them to me), that sounds appealing. But who will have access? How will that information be used and protected? Where will it go once it's collected? The UK's proposals in this area are so weak that Ross Anderson is heading a movement to help people opt out of having their patient records uploaded.

As much data as is now collected about us all – credit card trails, online shopping, medical data, government dealings, phone bills, Web logs – it's nothing compared to what's coming our way. Location-tracking (primarily but not solely via mobile phones), national identity databases, border controls that require fingerprinting and other biometrics will all generate far more data than anything we have now. And that's without RFID. A friend points out that in the US foreigners are required (although it's rarely enforced) to carry their I-94 entry forms with them at all times; trials are underway to include RFID chips in these, and the privacy flaws are already being reported

Which leads to another strand: technologies that don't work. Despite the fact that everyone who's ever installed new software has had the experience of having it utterly fail to work, hope seems to spring eternal that any IT project will do what its vendors promise if only it's sufficiently large and commissioned by a government. In a way, we have to be grateful when those hopes are crashed; an identity database that fails obviously, frequently, and undeniably is much less damaging to the person who is the object of that failure (the failee?) than one that fails subtly and rarely. The real problem is not that technology fails – all technologies fail sometimes – but our faith that it can be trusted.

The Hitchhiker's Guide to the Galaxy summed up the three stages of civilization thus: How can we eat? What shall we eat? Where shall we have lunch? There is a similar thread running from the natural desire for greater safety for ourselves and our children (warning labels on bags of marbles) to surrendering control over our own lives (a database to make sure that marbles aren't sold to anyone who isn't bright enough to know not to eat them).

Enjoy the holidays.