[Synapse elist] Artificial Creativity & Machines

Leonel Moura leonel.moura at mail.telepac.pt
Thu Apr 3 22:52:20 CST 2008



Of course that debating "Artificial Creativity" can generate in a 
never-ending discussion about the concepts of artificial and in 
particular of creativity. And here the question for me is not so much 
to know if "artificial creativity" fits our current concept of 
creativity but if we are willing to expand that concept in order to 
accept animal and machine creativity? As we did already with intelligence.

For example, considering Deleuze & Guattari's "unsettling of the 
established order" as inherent to creativity, we can ask if a 
rule-based systems, or algorithmic, is able to break rules? We must 
accept it as a possibility, because emergent processes can produce 
global outcomes which were not previously planned in the code. Some 
algorithms may generate an Aristotelian Eureka!  For Herbert Simon 
(1997) "chaos derives from deterministic dynamic systems that, if 
their initial conditions are disturbed may alter their paths 
radically. Although they are deterministic, their detailed behavior 
over time is unpredictable, for small perturbations cause large 
changes in path."

On the other hand we can say, with Debord, that creativity is not 
just the capacity of arranging objects and forms, but the invention 
of new rules on that arrangement. In this sense we can imagine an art 
process that creates art by itself. "An idea becomes a machine that 
makes the art" (Le Witt, 1967). That is the understanding underlying 
my robot art and in this sense the will to take the "human out of the loop".

Finally I would make a distinction between the practical aspects of 
robotics, applied to art or other things, and the philosophical 
issues that emerge from it. Robots will probably evolve to be hybrid 
bio-machines but for the moment they are still essentially 
non-biological bodies. That does not mean to deny the symbiotic 
relation human-robots and to not recognize that bioprocesses are at 
work in most of these machines.

Leonel



More information about the eList mailing list