Transcript of "Resisting the all-seeing eye", a talk between Cory Doctorow and Charlie Stross

blog link
Video on
Description of the talk on

I don't have time to transcribe the entire thing, but here's the first 11 minutes. Feel free to work with it (see CC license below), or email me parts at dndculix AT gmail DOT com if you want stuff added.

Tim Phillip: Good evening. Thanks everybody for coming today. I'm Tim Phillip, I'm the executive director of the Open Rights Group. I won't detain you for too long. This event is organized by the Open Rights Group, and we campaign for your digital rights. We currently have activities in two main fields: online privacy, and copyright reform. In privacy, which we'll hear a lot about tonight, the threats are very many, as we'll hear, but in many ways I think the public mood is with us. The threats from both the state, and intrusive new technologies, like those pushed by the company Phorm, are understood and creating creating quite a backlash, not least because of their slightly bizarre tactics. The underlying issues, though, will be with us for a very long time, and that's why our voices need to be heard. So before I hand you over to our distinguished guests - Ian Brown, Cory Doctorow and Charlie Stross - I encourage you to remember after you've gone from this evening to go our website - - and become a paying member and sign up for our Action email list. Because with your help - both time and money - we can win these battles for citizens' rights.


Ian Brown: I'm Ian Brown; I'm going to moderate this evening. We're not going to talk at you, you'll be delighted to hear; we'll try to keep it as interactive and flowing as possible. We're just going to start, very briefly, if I could ask Corey, and then Charlie, for a few minutes to tell us what they see as as the key threats to privacy right now, and what you can do about them.

Cory Doctorow: I think that the problem with privacy that we have right now is twofold. The first is that our intuitions about privacy are very confounded. We've seen this in microcosm before with geeks who were early internet adopters, who found themselves on Usenet in the 80s and 90s - which was, for those of you who don't remember Usenet, it was an early relay messageboard system, that was really what most people thought of when they thought of when they said 'the internet' for a really long time. And people would post all kinds of disclosures on Usenet, thinking that none of it was being archived, and none of it would ever come back to haunt them. And they would post everything from "man, Woodstock 99 was awesome, I spent the whole time in the K-Hole", to stuff like how they cheated on an exam, or how they were in love with someone, or they would say nasty things about each other... just generally do things all the things that that university students do, because the majority of Usenet users were university students.

And then a company called DejaNews went online, and DejaNews had most of Usenet indexed by user. And all of that information that we thought of as a book written on water turned out to be immortal. And not only that but indexable and part of your life. Even the stuff that Deja didn't have turned out to still be around in some form or another. It turns out there was a zoology professor at the University of Toronto whose department wasn't billed for backup tapes, so he never rotated them. So he had literally a giant bookcase full of backup tapes going back to the early 80s


IB: Let me just add: they guy in question, Henry Spencer, is the guy who wrote C News - the first actual Usenet service software to get mass adoption.


Cory: Right, so he put the rest of it online. And all the stuff that we thought of as short-lived turned out to be immortal. We're on the brink of more disclosures along these lines. For example all of your have your cameras out, you're taking photos right now, a lot of those photos will go online with my name under them. Some of them will have pictures of other people in the background whose names won't be associated with them. We're probably not far off from having pretty good face recognition software [e.g. iPhoto's new 'Faces' feature! -ed], and if you can find one picture that's labeled, you can then apply that label to all the other pictures. And so people who today are happily putting snaps up online and thinking those snaps are not really anything we need to worry about, we'll find, soon, that all those snaps will be aggregated, and we'll probably have algorithms to find the most embarrasing ones.


IB: This has been the case for famous people for a long time. There's a photograph you come across in history books periodically: it was taken in one of the central squares in Vienna, on a certain day in summer of 1914, showing a sea of faces looking up as they hear an announcement that war has been declared between the central powers and Russia and France, and later great Brittian. This is World War One. And what makes this photograph notable is that there's usually one face in the sea of faces that's circled - Adolf Hitler. Now we have been able to do this for people like that for a long time. The difference is going to be qualitative when it's something that happens to everybody. It's the flipside of Andy Worhol's dictim that everybody will be famous for 15 minutes; no - everybody is going to be famous forever.


Cory: Well, no, everybody's going to be famous for 15 GB. And right now we create.. Bruce Stirling calls personally identifying information the smog of the 21st century. It's an externality of our network lives that we can't stop. We have no way to carbon sink it - it stays out there forever, and it's building up in the noosphere. Massive amounts of it. And we don't worry about any one piece of it, but as more and more of it gather, it presents lots of threats. And I was thinking about a taxonomy of threats today, and I think that there's at least three different kinds of threats. And they'll all embodied in 'Brazil', the Terry Gilliam movie.

There's the Tuttle/Buttle threat - someone finds you because they're looking for someone else, and they assume the database is infalliable. There's the Cardinal Richelieu threat, which is "Give me six lines in any honest man's hand and I will find a reason to hang him". If you looked at every photo that was ever taken of you would you eventually find one that would disqualify you from holding public office. And then the last one is the one that tends to take center stage when we worry about privacy threats, but is in some ways is the less worrying of them, which is the malicious hacker - the person who wants to spy on you and catch you because you're a pollitical dissident. And it's because we focus on this very lurid threat, which is the Stasi hunting you down because you're an urban geurilla, and most of us go "I'm not an urban geurilla, so I don't need to worry about it", that we miss these other two living threats. So I think our intuition is bad about it.


Charlie Stross: I prefer to take a slightly longer historical view. Architecture tells us a lot about a society's expectations of privacy. And what we learn about it historically be looking at architecture is that our concept of privacy is a relatively recent, relatively frail flower. You go and look around any post-medieval palace in Europe, or any stately home, and look at the architectural layout and you see rooms with muliple doors. The rooms interconnect each other. What you don't see, until relatively recently, is what we have come to know as the corridor or passage - a long room with doors connecting other rooms, each of which only have the one door. Now there's a subtle social reason for this design of houses. Historically speaking the really poor did not have privacy - they lived 20 to a room and slopped out in a bucket, and everybody knew exactly what everybody else was doing all of the time, because there were witnesses everywhere. What we tend to forget is that the very rich had no privacy either. The various quaint titles associated with royalty, such as 'Chamberlain of the Royal Bedchamber', et cetera, this means the guy who mucked out the royal chamberpot in the morning, because he slept at the foot of the royal bed, and when the king woke up somebody had to fetch and carry the chamberpot, and wipe the royal buttocks. This pattern of architecture - the evolution of the corridor - facilitated the growth of the concept of privacy among people who were rich enough to afford buildings with corridors that allowed servants to bypass the rooms they were holding private conversations in. Prior to that point virtually every conversation, every personal interaction, would have had witnesses. It was still common practice in the 17th century for the royal bedchamber to have the Prime Minister and their lord chamberlain sleeping on truck beds at the end of the king's four-poster.


Now fast forward a little bit to the Industrial Revolution: we have aggregation of cities; we have construction of housing which would by the standards of an earlier age would have been considered unbelievably luxurious, for the proletariat and the middle classes - people with individual bedrooms. And then the post WWI labour shortage - a combination of the flu the war killed an awful lot of servants...


CD: the flue kills people?

CS: It did then. Anyway, as Agatha Christie wrote "I never imagined that at any time in my life that I would be so rich that I could afford an automobile, or so poor that I couldn't afford a servant." People with servants have no privacy. We have developed an understanding of privacy during the 20th century as meaning, effectively, "no witnesses". And now we appear to be moving into a period where there will be witnesses all the time, all around us. Silent, mostly unspeaking, electronic witnesses that follow the trail of bread crumbs, that follow the GPS coordinates, or the cellphone triangulation data from our phones; the cameras in the streets outside monitoring where we are - and as Cory put it, this is electronic data smog. How much data smog can there be? How much is there that can be retrieved. Well I did a thought experiment on this a couple of years ago. Let us consider Moore's law and permanence and data storage. Most of our data storage today are volatile, but what is the densest conventional storage medium we can conceive of? Well how about one bit per atom? We can conceive of stable molecular structures such as a diamond lattice which persist over long times and are stable. Imagine an artificial diamond lattice where a carbon-12 nucleus represents a 0 and a carbon-13 represents a '1'. You get about 10^23 bits per 12 grams of carbon, or about 10^23 bytes per ounce. Call this stuff 'memory diamond'. You could store a VGA resolution video stream for an entire century for every human being on the planet in the 21st century in under 100 kilos of this stuff. We're not going to find it easy to bury that data. Although I think memory diamond has a strong future for carbon sequestration.


CD: We solve one kind of smog problem by creating another one.

Creative Commons License

Content on this page is licensed by culix under a Creative Commons Attribution-Share License.

changed November 13, 2009