Re: >H Optimal Intelligent State?

Anders Sandberg (
07 Sep 1998 18:55:35 +0200

Doug Bailey <> writes:

> As intelligence and the societies/civilizations/structure-
> complexes it creates become governed more by directed
> intelligent evolution than natural evolution, it seems to
> me that it will only make changes that increase its
> overall "state". Assuming we evaluate the state of a
> civilization by comparing it to the limits prescribed by
> our current understanding of physical laws, it appears that
> there should be an "optimal state" that intelligent systems
> should tend towards. Regardless of the initial environment,
> conditions, substrates, characteristics, etc. of
> civilizations, it seems that once they has harnessed the
> power of directed evolution that they should tend to
> strive for this "optimal state".

(Formally, I would distinguish between the state of the civilization and the "value" it is optimizing)

I think there is a large amount of contingency here. If we assume that all civilizations seek to become better at doing stuff, until they reach the limits set by physics, then their structure would still be rather contingent.

Imagine (somewhat arbitrarily) that the ultimate limits were medieval-level technology. What kinds of societies and cultures could exist in such a strongly constrained system? Looking back at history, we see that we could have very many styles (ranging from icelandic anarchy, european feudalism over the arabs and indians to the chinese empire), and that a lot of possibilities that could have happened were unexplored (like a more technically advanced Atztec empire, for example).

The space of possible societies and cultures appears to be so large that there isn't time to search through it for an optimum of some kind, very much like the situation is in evolution of species today. Most likely societies will end up in local maxima, and if there are several their mutual competition might lead to an overall co-evolutionary random walk through the space of possibilities without end.

> 2 - What insights can we draw into the nature of this
> "optimal state" from our current knowledge of the natural
> limits (e.g., the Planck Length, the speed of light, etc.)?

We can determine some of the limits of doing stuff, which gives us some constraints on their possible actions. But what they actually do, that is hard to tell. Unless it is possible to develop some kind of "psychohistory" that describes the evolution of cultural goals - perhaps possible, but tricky.

> 3 - What's the merit of evaluating optimality based on physical
> limitations? What other possible criteria might there be?

Mihály Csíkszentmihályi suggested the amount of flow experienced - how many of the citizens/subsystems achieve their goals in a constructive (for them) way?

> 4 - What would a civilization do once it reached the "optimal
> state"? In concert with #1, would such a civilization "die"
> of boredom having exhausted all the wonders of existence?

Not necessarily. It might remain static, or in some kind of local attractor. It might even turn out that once one factor was optimized, new factors become important to fix - it would wander in a subspace of the total space of possibilities.

> 5 - Maybe its errant to think of the "optimal state" as an
> evolutionary pinnacle and instead as an evolutionary
> singularity. What could possibly lie on the other side of
> an evolutionary singularity?

I can think of one thing: the end of evolution, or the start of a new form of evolution.

Anders Sandberg                                      Towards Ascension!                  
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y