Complexity theory, power laws, bears and honey

Following from my comment on Laszlo's last post, I looked up one of the topics that always burns in the back of my mind: complexity theory.

As a first stab, I read this article from 2001. Old, so perhaps you have heard about it already?

In a nutshell, paleontologists are puzzled by the dramatic discontinuities in the extinction patterns of life on earth over its history. Occasionally there are cataclysmic episodes that causes mass extinctions. Physicists call this phenomenon Self-Organized Criticality (SOC).

"In recent years, some physicists and other adherents of a new field called complexity have argued that the answer involves a simple mathematical theory -- the same one that may explain the collapse of a sand pile or a crash in the stock market. But their suggestions have drawn fire from paleontologists, the group traditionally charged with investigating life's past."

The SOC physicists traced the lengths of time that ancient groups of animals had survived, and found (surprise!) a power-law behaviour: most of the time, things were quiescent, but occasionally there was a mass extinction--the same behaviour typically seen with earthquakes, volcanic eruptions, and solar flares. This didn't have to do with external events (like an asteroid hitting the earth), but evolved as a consequence of how the different species interacted with each other. The complexity theorists say that the power-law "fingerprint" is indicative of a system in a critical state.

The main jam was that the paleontologists re-analyzed the data "with the appropriate statistical tests", and no power-law behaviour emerged. They again tested it with a new database of 36,000 genera showing when the groups appeared and when they vanished ("the best data set that is available at this point in time"), and still no evidence of a power-law pattern was found.

Another complexity theorist put it this way: "It's not enough to say that self-organized critical models give power laws, and we see power laws in the fossil record, and therefore they must be self-organized critical. That's logic on the same level as saying, 'Bears like honey, my wife likes honey, therefore my wife is a bear.'"

As a result, Richard Sole, one of the SOC physicists, has modified his ideas and decided to collaborate with Douglas H. Erwin, a paleontologist at the Smithsonian Institution, to see if complexity theory can help explain the appearance of new species after major biological crises.

I doubt this is the last we hear of complexity theory bashing heads with other fields--perhaps this happens on a power-law basis??--and i'm sure there is a lot of interesting stuff to mine...

How to present a complex idea concisely

Better a narrow description than a vague one.

One reason founders resist describing their projects concisely is that, at this early stage, there are all kinds of possibilities. The most concise descriptions seem misleadingly narrow. So for example a group that has built an easy web-based database might resist calling their applicaton that, because it could be so much more. In fact, it could be anything...

The problem is, as you approach (in the calculus sense) a description of something that could be anything, the content of your description approaches zero. If you describe your web-based database as "a system to allow people to collaboratively leverage the value of information," it will go in one investor ear and out the other. They'll just discard that sentence as meaningless boilerplate, and hope, with increasing impatience, that in the next sentence you'll actually explain what you've made.

Your primary goal is not to describe everything your system might one day become, but simply to convince investors you're worth talking to further. So approach this like an algorithm that gets the right answer by successive approximations. Begin with a description that's gripping but perhaps overly narrow, then flesh it out to the extent you can. It's the same principle as incremental development: start with a simple prototype, then add features, but at every point have working code. In this case, "working code" means a working description in the investor's head.

Source: http://www.paulgraham.com/investors.html

Chimps and grad students

Quick youtube clip:

http://www.youtube.com/watch?v=OVlJv7ZkvGA
Chimps were trained to remember a grid of numbers on a touch screen presented for a brief period. Afterwards they had to touch squares on the grid covering up the numbers, in ascending order. Grad students trained for the same period of time performed much worse in this task! Good thing we don't live in the trees.

Chimps for postdocs!!

Ownership of ideas vs. the value of collaboration in scientific research

This is something that has been troubling us for the past while: What to do with ideas that you are not paid to come up with? Can we work on them in our own time? What if the idea requires validation from data you do not have access to, or critical commentary from someone who pays you to think about something else?

Here is a suprisingly frank debate I came across.

The story of the blue brain project

A very cogent piece of science writing for the lay reader.

What you should know about organic food cerification

MOST of the chicken, fruit and vegetables in Ellen Devlin-Sample’s kitchen are organic. She thinks those foods taste better than their conventional counterparts. And she hopes they are healthier for her children.

Lately, though, she is not so sure.

Read more here:

cudnt resist this

im sure some of us wud remember the "rotten burough" (is it spelt okay?) from black adder where

Mr Black adder is the polling officer, the vote counter, the only voter and personal advisor to boldrick(who ran for those elections)..

amazingly I found this, 

http://timesofindia.indiatimes.com/A-polling-station-for-one-voter-/articleshow/4213160.cms