Friday, June 17, 2005

Is there an upper bound on complexity?

Jaron Lanier has argued against what he calls "cybernetic totalism" on the grounds that we just don't know how to build reliable software systems larger than 10 million or so lines of code. Similar limits are in play in most engineered systems. This leads to a few questions:

1) Are these limits due to our current ignorance or are they real limits inherent in the nature of complex systems? If they are real limits, can they be expressed in a useful general law?

2) Does biology routinely circumvent these limits? Is this why we see parts within parts within parts -- the emergence of unitization at various scales?

3) Is there a useful quantitative measure for characterizing proximity to the complexity boundary?

1 Comments:

Blogger Radish King said...

Exciting questions. They made me think, immediately, in terms of music composition (which has no upper bound in spite of the octave) and connectivity. It's too early to think, plus it's Saturday. I'll be back.

8:28 AM  

Post a Comment

<< Home