what is artificial life (alife)? more to the point: what is life?

some people claim that `alife exists'---that we have already created software systems `complex enough' to warrent the label `life'. the real achievement here is defining the axis from life to non-life, rather than the systems so far.

because really, the `liveliest' software systems to date are still only simple shadows compared to even bacteria. but animism is popular: a virus is alive, the sun is alive, the telephone system is alive, a hive has `a life' beyond the sum of the lives of its members. why not the latest GA? because they're not complex enough? where do you draw the line?

[expansion: compared to nature, the programs that people write are tiny, non-complex, and have little content. when comparing bacteria or cells to computer programs, sometimes people add up the RNA information content and compare this to some source code, and conclude that software has equaled nature. but this leaves out the molecular state of the organism, a huge resource (though it is quite random, thus has a large discount). [link difference between human layers (eg computer languages) and naturally evolved layers (eg cells, human language)]]

in An Evolutionary Approach to Synthetic Biology Tom Ray defines life with a checklist of two: self-replication and `open ended evolution'. he considers evolution to be the fundamental process of life. but most of us would say a turing-successful AI is alive because it can think and learn, even though it lacks replication . afterall, darwinian evolution is only one of many learning algorithms.

furthermore, you and i are certainly alive, but our brain's resources and capacity for learning are both finite (though very large). it's not even clear that life on this earth or in this universe is open ended (though it's certainly very very large).

my definition of life is based on complexity and emergence.

Schrödinger answered What is Life? with Living matter evades the decay to equilibrium, see here.

Gell-Mann's definition is based on transmission of information, see here

Nell Tenhaaf's Simorg Culture


isbn 0-942299-29-9; Zone 6; `incorporations' ed Crary, Kwinter Manuel DeLanda `non-organic life'; 909.82 I37 (hunt 2nd floor)

p138 ``Roughly, we could say that phenomena of self-orgaanization occur whenever a bifurcation takes place: when a new attractor appears on the phase portrait of a system, or when the system's attractors mutate in kind.''

cf MST