[ExI] Universal timeless principles

Anders Sandberg anders at aleph.se
Tue Oct 6 08:22:16 UTC 2015

I attended a lecture on big History by David Christian yesterday. Great 
guy, and a very good lecture too. Big History is a fun approach to 
learning about the world, https://www.bighistoryproject.com/home

He made a really interesting comment related to the emergence of life 
that also relates to this thread. He pointed out that there is universal 
information: gravity, electromagnetism and iron behave the same way 
everywhere, so an organism does not have to store this information 
internally. On the other hand, how the organism is organised is local 
information: it is contingent, and if the descrption is lost the 
organism cannot survive or reproduce.

(Note that edge cases like viruses to a great extent rely on cells 
containing the neccesary information to reproduce: they just contain 
information on how to make *themselves*, not how to make proteins or how 
to assemble them.)

In regards to universal timeless principles, they might be like 
universal information: something that is always true and acting on 
relevant systems whether they know it or not. Game theoretic equilibria 
and evolutionary stability comes to mind. In many situations that means 
you do not even need to know them, they just happen naturally. However, 
it might be *efficient* to have explicit representations of such 
information. Knowing the laws of electromagnetism allows us to make good 
predictions and inventions rather than hope some random configuration 
will do the job. Knowing that cooperation is fragile but reinforced by 
reputations and altruistic punishment, we can set up our societies to be 
better at reputation management and sponsor discouragement of misbehavior.

Conversely, if a principle can be forgotten forever and never 
rediscovered, then it is not universal.

So maybe we should consider what kind of "ethics" would be independently 
discovered by (1) humans in a parallel world or isolated continent, (2) 
aliens, or (3) intelligent agents. Something truly universal would show 
up in all categories, something universal-to-evolved-beings would show 
up among 1 and 2, something human-universal would be in category 1.

Anders Sandberg
Future of Humanity Institute
Oxford Martin School
Oxford University

More information about the extropy-chat mailing list