you win or you die, or at least get locked up in a small box for a very long period of time with little chance to do interesting work.
I had a quick snippet of conversation on Skype with a friend (the "on Skype" thing is relevant: it probably means it was logged by at the very least Microsoft).
[11:49:51 AM] spiritplumber: hm.
[11:49:53 AM] spiritplumber: game theory.
[11:50:35 AM] spiritplumber: you're playing a soccer game. your team is 10 players (who are allowed to use their feet) and a goalkeeper (who is allowed to use their hands as well). their team is 9 players, a goalkeeper, and a referee (who is allowed to decide what rules to enforce). how win?
[11:51:28 AM] cs_kitty: Will the spectators turn on the ref if he goes too out of line? :P
[11:53:02 AM] spiritplumber i think we've just had every "how to protect your data from a police state" conversation ever.
[11:54:23 AM] cs_kitty: Yup.
This is either a technical problem or a political problem. If it's a political problem, there are two key points here, continuing the metaphor:
1) Will the spectators turn on the ref at all?
2) How "too out of line" is enough for that to happen?
I am extremely unskilled in political solution, and would like to read about them rather than write about them.
My interest is in analyzing the situation from a game theory perspective. Let's say that the problem of "How do I make use of my personal data through my day, with said data remaining personal" -- a problem if you travel internationally and are a journalist, or in my case an engineer with trade secrets or inventions that have not been patented or open-sourced yet -- is a game of capture the flag. One player, called Jeb because Jebediah Kerman is awesome, must take the flag through an obstacle course. The other player, called the State of Arstotzka becaue the indie game "Papers Please" is awesome (and a lot more enlightening about security issues than this diary can ever be, so look it up), must intercept the flag by setting up a difficult obstacle course.
This is fine. The issue is that the State is able to change the rules on Jeb, with some lead time and the requirement to inform Jeb (unless we have secret laws) and to decide the degree of enforcement that any of the rules will receive, with effectively no lead time and no requirement to inform Jeb.
So, does Jeb have a chance of winning? Let's find out.
* If Jeb uses a bike chain to secure the flag around his waist (encrypts his data), the State can beat Jeb up, or lock him in a room, until Jeb agrees to unlock the chain.
* If Jeb hands over a piece of fabric that is A flag, but not HIS flag (deniable encryption), the State can decide that it's not the flag they want and beat Jeb up, or lock him in a room, until Jeb agrees to give them the right flag -- or tell them what they want to hear, in case he didn't have the right flag in the first place.
* If Jeb decides that the only winning move is not to play, he will find it very difficult to do so, in practice.
* If Jebediah Kerman decides to circumvent the problem by building a rocket booster and bypassing the obstacle course (set up your own information infrastructure), he will likely find that after the first few times he does this, the rules will be amended to forbid it it.
So, it would seem that this is the sort of game in which you can't win, you can't break even, and you can't quit. What can be done?
Jeb could lock the flag about his person (encrypt his data), but give the key to a trusted party, for example his lawyer. This raises the question about what would his lawyer do if she knew that Jeb was under duress.
Jeb could simply not carry flags personally (use a side channel for data transmission), although he opens himself up to the risk of being presumed to be carrying a flag and be punished until he provides a flag he doesn't have.
Jeb could do all of the above, and employ a very diverse mix of strategies. While it is true that his opponent has the option of changing the rules as their move in the game, it is a costly and time-consuming move; Jeb would have to be creative and, for each trip, find an obstacle course navigation strategy that lies outside the rules without breaking them. The issue there is that the State not only has the option of changing the rules, it also has the option of enforcing them with an arbitrary degree of precision, which is a much less costly move.
Jeb could try to find an external referee, or arbitrator, or such; the issue there would be that the State has no incentive to abdicate its dual role. This would require Jeb to resort to force to create that incentive (probably in the form of rocket boosters, he's Jeb Kerman after all).
Finally, and possibly as a way to restore the balance of force, Jeb could decide for a mutually assured destruction strategy: build an extremely large rocket booster aimed at himself, and have it navigate to his position in case he feels that he has been put in a no-win position.
On a more fundamental level, I postulate that whichever player resorts to the initiation of force has lost, and will not consider the last two options. However, the "diverse mix of strategies" option has potential if applied as a metastrategy.
If you are familiar with the Prisoner's Dilemma, you will know that the optimal strategy for a single instance of it is very different than the optimal strategy for an iterated game.
On one hand, creative chaos has enormous potential; if there are a thousand Jebs carrying flags with different methods, chances are that one or two will come up with something amazingly clever. Switching metaphors, this would make the State into an environment and Jeb (or rather, the many Jebs) into an evolving organism within the ecosystem. An evolutionary approach is robust and able to keep up the pace with any but the most grueling conditions; however, it is also inefficient and will cause many Jebs to be lost.
On the other hand, if everyone is doing their own thing, it becomes difficult to organize in any but the most elementary ways, and the benefits of herd immunity and the power of numbers are lost.
Is there a way to get at least some of the benefits of both approaches?