jack: (Default)
[personal profile] jack
I won't say whether it exists or not, but much in Foucault's Pendulum much attention is paid to the implausibility of the idea there may be secret descendants of the Templars, putting a long-laid plan. Many other books (and conspiracy theories) have similar ideas.

How *would* you go about organising such a thing? First, a few background points.

* Obviously there are *some* secret societies. The masons actually exist :)
* If that helps the members out now, that's obviously a reason to do it.
* Or if they're following some religious purpose.
* Or if you want to shape the future of humanity.
* But would you want to found a society in order to benefit the N00-year-hence members of it? But that's what some fictional societies with a big secret seem supposed to do. (See 1984 for a philosophical discussion of maintaining a caste system.)

If you did want to, how would you do it? First, the parameters:

* Let's say you have a big secret you want put into effect or revealed at the next millennium.
* You want to prevent anyone finding out before hand, including the members.
* But want the secret to survive.

And, so, what:

* One technique is to simply write lots of instructions and bury them. They (hopefully) won't be found too soon and (hopefully) when they're found some people primed by rumour will follow the instructions.

That later point is when many stories are set: with the digging up of a mysterious treasure map and wondering who wrote it.

* Another is to found a small secret society, and trust the members to hand the secret down to the next generation. This is also popular, though fragile.

If too few people know, sooner or later they die at once, and the followers are stuck.

If too many, someone wants to go off, and grab the secret early.

Each man/group choosing a trustworthy successor can work, but they can't *always* be right. And over their life, they may stop caring.

And they can all be hunted down.

* What other tricks are possible? Perhaps estimate Moore's Law and release an encrypted message publicly, and rely on it being decrypted at about the right time.

* Have several groups with part of the secret, to meet at some point in the future.

It's tricky specifying when though. You can say "meet at this place, this time, this day, this year", and rely on statistics to make it robust. If a secret group has a 10% chance of dying out and 10% chance of going rogue, it has 80% chance of doing the right thing. Founding three such groups needing one other to decrypt the secret improves the chances to ~3% of two dying out, and ~3% of two going rogue.

But no "Machiavellian" secret societies anyone's tried to tell *me* about have been so clever.

Alternatively, try to specify a place you can't find in advance, eg. where the next [event] occurs, one month later. But clever groups trying to jump the gun might cheat, eg. by going to the *last* [event], on the anniversary, and hoping another group will think the same way.

* Found a religious order and tell them it's god's will.

* Launch the secret in a space probe designed to intercept the earth in N years.

* Lock it up and trust no-one ever considers trying to cheat mechanically.

* Hide it in a statue, church, or other long-lasting publicly visible place, with a clock designed to go off at the right time. Anyone looking before that won't know where, but at the time it'll be obvious to all comers.

* Similarly, you could hide it in a computer -- or mutation-resistant biological -- virus, designed to spread slowly and go off on that date.

Though none seem perfect. Does anyone have any better ideas?

Date: 2007-07-26 04:46 pm (UTC)
simont: A picture of me in 2016 (Default)
From: [personal profile] simont
Estimating Moore's Law: as I mentioned last year, Arthur C Clarke attempted this with a thirty-year planning horizon in "2010", and was out by a factor of perhaps five. Over thirty times thirty years, I can therefore only assume that someone with the same skill at prediction (at which Clarke has generally been no slouch) would suffer an error factor of 530!

Another problem with leaving your encrypted message is that you have to find some way to be sure anyone will care about decrypting it. I mean, if Jesus had left a strongly encrypted message, we can be reasonably sure that there would have been people trying like crazy to crack it even before they had any real idea of how to, and that the message would probably be decrypted slightly before the moment when the Moore curve first made it reasonably feasible to do so; but if instead of Jesus it had been some completely unknown random in 30AD, or 1030AD, or even 1730AD, we might very well not have given much of a hoot, and might be concentrating most of our attention and CPU power on the latest distributed.net challenge for which there was an actual cash prize to be won.

Date: 2007-07-26 04:54 pm (UTC)
From: [identity profile] http://users.livejournal.com/vitriol_/
Well, yes, the encryption is idea is utterly ludicrous in the long term, it just amused me :)

Date: 2007-07-26 05:50 pm (UTC)
From: [identity profile] cartesiandaemon.livejournal.com
Indeed. It seems a promising *sort* of solution, since it solves (a) keeping it secret and (b) keeping it known. But I know using current cryptography isn't anywhere *near* accurate enough.

And you do need to *also* found some kind of cult that will keep people interested, but that doesn't need to be secret or restricted, so we've reduced it to a simpler, widely considered problem :)