[ExI] Unfriendly AI is a mistaken idea.

Lee Corbin lcorbin at rawbw.com
Wed Jun 6 02:57:34 UTC 2007


Mike writes

> Another candidate [for intelligence all around us]
> might be the arrangement of animals/insects in an
> ecology.  Surely an individual bee has no 'understanding' of the
> colony's impact on higher forms of complexity to which it may be
> interrelated.

That's similar to most people having no clue concerning
the nature of the economy they're embedded in. Okay,
yes, the economy does have a certain kind of "intelligence",
e.g., in the invisible hand's ability to set prices optimally
(or near optimally, most of the time). 

> The ecology will adapt to change of state from
> rainfall, fires, etc.  If it seems like a stretch, consider how brains
> manage their change in state with blood-sugar levels or nutrient-poor
> diets - aren't we the same kind of reactionary?
> 
> Maybe i'm not thinking of intelligence in the usual definition...

The way I think of intelligence is usually as a characteristic of an
entity---an entity who knows in some near or distant sense what
it means to survive. For example, most animals are aware of
dangers and seek to avoid them. This is one requirement of
all the present day evolved entities, or *sophonts*, as the 
term has just been explained. An entity usually has enough
sense to look out for itself, although an interesting discussion
is taking place whether or not we might be able to create
general purpose intelligences, GAIs, which are very open
ended systems capable of addressing whatever problems
presented to them, yet lack this ability to "look out for 
themselves".  (I vote "yes", or, "probably", by the way.)

The sort of "intelligence" that an ecosystem or an economy exhibits
is something quite different.

Lee




More information about the extropy-chat mailing list