I read another article, this one by Frank T. Vertosick, Jr., a neurosurgeon. It is a brief history of intelligence, and was quite interesting, if a bit difficult to follow. For ex., knowing nothing about the man previously, I didn't know at first whether he meant an actual computer he knows about somewhere when he referred to "Darwinian computers". Then there were the words like eukaryotes and prokaryotic.

But I am hoping someone here can explain to me more clearly (examples would be extremely helpful) what he meant when he talked about noise. I did infer that he meant noise is something that gets your attention, but that's as far as I could get. I have only the most elementary knowledge of this field.

Okay--he's talking about natural selection, and says: "Darwinian systems learn in the simplest way possible, trial and error, and such systems require just three basic components: 1) a noise source to generate a diverse library of possible solutions to any given problem; 2) a competitive ecosystem in which the noisy library of solutions can compete to solve a problem by "surviving"; and 3) a virtual* storage system to save symbolic representations of correct solutions for future implementation. In conventional Darwinism, spontaneous mutations serve as the noise source..."

Later on he talks about organisms adjusting their noise levels. And, "Noise is a critical parameter in Darwinian systems...Insufficient noise makes their systems inflexible; excessive noise renders them unstable. ... all modern cells have mechanisms for raising their own rate of genetic noise..."
*I think this contributed to my confusion over whether he meant an actual computer.

FTR, he goes on to talk about how the original theory of evolution has itself evolved.

I hope I've put enough to get across his idea of "noise". But, beyond mutations, I truly don't "get" what he was talking about. Thanks for any help, and any errors in copying are mine.

Interesting discussion Jackie.....can you post ref to the article?

I think his "noise" here is just random variations in genetic reproduction. That is, the DNA in the eggs or sperm will not be exact copies of the original. If the DNA is duplicated exactly the organism won't be able to adapt to a changing environment and if there are too many copying errors the incidence of harmful mutations will be too high for the species to continue to sustain itself.

yeah, noise is extraneous information in a sample.

http://en.wikipedia.org/wiki/Statistical_noise

Candy: link
Erk--I just tested the link, and it takes you to the magazine cover, not the article, which begins on p. 20. You can zoom in on it; it'd be pretty hard to read otherwise, I think.

Thanks, all, and I'll welcome any other insights. I just felt like I had begun a class in the middle of the semester.

I wonder if, by 'Darwiniian Computers' he is referring to computer programs that operate heuristically? Such things have been around for a long time - I was programming in the '70s and I knew about them, then, although my own work was in the then-new field of relocatable programs (such as every computer made since c.1990 has used!)
An example of a heuristic program is one that learns to play chess.

I can't tell for certain, but by the end of the article I thought he was referring to biological "computers".

Yep, that would make sense. I have heard predictions that Bio-computers will be the way ahead. I reserve judgement, at present. Maybe I read too much scifi!

Null concept.

agreed.

Thanks Jackie..but the link goes to MENSA page (to log in)...which I cant so!

But I have been reading a bit about from this author and still no clear answer to the 'noise' thing.