[ExI] mayan forecast

[Anders Sandberg]

On 2012-12-21 17:26, spike wrote:
> Regarding my snarky comment on stone calendars, I can look at it another
> way.  In a few years, all our paper calendars will be gone, but after
> thousands of years, the Mayan calendar is still here.  What are we doing
> or making that has that kind of durability?

I am seeing large scale rock re-sculpting from my window right now, as a 
new train track is being built here in Stockholm and the bedrock is 
being cut. Leaving aside the likely local ice age (which will smooth out 
the cuts a great deal) some of the landscape forming we do is pretty 
permanent.

We have many materials that are extremely resilient, but few of them are 
in wide use: easily shaped materials are usually preferred for mass use. 
Still, from a neolithic hunter-gatherer perspective we are filling the 
world with amazingly durable metals, ceramics, glasses and plastics on a 
vast scale. And if we get atomic precision manufacturing I suspect we 
will have plenty of very tough sapphire, diamond and other designer 
compounds (although we are likely to add voids to keep weight down).

In the end I think the key trick to make something ultra-resilient is to 
have loads of copies of it everywhere. The Library of Congress and the 
Bodleian might burn, but widely pirated files or styles will have a good 
chance of surviving.

[ If the probability that something survives is p, and there are N 
uncorrelated copies of it, the probability of at least one copy 
surviving is 1-(1-p)^N. As N increases this becomes a step-function 
shifting for near zero survival if p < 1/N to near certain for p > 1/N. 
So if you want to save a one-in-a-million survival object make sure to 
make two million copies and spread them around. Spam will outlast 
everything. ]

[[ The real problem is single mode failures, like all electronic media 
failing at the same time. So the real threat is not that CD-ROMs decay 
or computers are volatile, but that we store all our information in too 
similar modes. ]]


-- 
Anders Sandberg
Future of Humanity Institute
Oxford University