Re: question on algorithmic complexity

Bill Hamilton (hamilton@predator.cs.gmr.com)
Fri, 19 Jan 1996 12:33:35 -0500

Eddie Olmstead writes

>I agree with Bill that randomness alone is not a measure of complexity.

Hmm, I would say that randomness alone is not a measure of meaningful content.

>However, I think I know where Brian is going. A simple sequence of
>"HTHTHTH..." is not complex because you could write one simple algorithm to
>generate it. The order is very simple--like a NaCl crystal with an endless
>repeation of a simple pattern. There is order, but very little information.

Agreed.

>However, to generate the sentence "DNA is a complex molecule", it takes a
>much larger (but still finite) set of algorithms--rules of spelling,
>grammer, etc. of the English language. The order is much more complex, but
>it contains information precisely because isn't simple repition--like a DNA
>strand. A totally random sequence doesn't obey any rules. Thus, it is
>complex and disordered. Thus, we are looking for "ordered complexity" if
>you will.

I think Brian would say "organized complexity".

>I'm eager to find out if I guessed correctly, but I'm willing to
>let Brian write up his full post that he has been promising us so he can
>define the terms more carefully than I've used.

One aspect of information transmission that seems to throw some people is
that "randomness" can have different meanings depending on whether you are
receiving or sending a message. If you are the sender, then from your
point of view the message is mostly deterministic: you know what you sent.
From the receiver's point of view, however, the message he receives is
mostly random: if it isn't, he already knows it, so why bother to receive
it. Most of what I know about information theory is from a course I took
as a senior in 1963-64 and another course I took as a first year grad
student in 1964-65, so I'm sure most of what I say about information theory
is dated and naive. But I'm looking forward to this discussion. C'mon
Brian :-).

Bill Hamilton | Vehicle Systems Research
GM R&D Center | Warren, MI 48090-9055
810 986 1474 (voice) | 810 986 3003 (FAX)
hamilton@gmr.com (office) | whamilto@mich.com (home)