Monday, October 31, 2005

The Legacy of Rosa Parks

Bernadette went to the Capitol last night to see Rosa Parks lying in state. The Capitol Police reported that 30,000 people made it through (Bernadette left when she realized it was a six hour wait after midnight).

I have admired Mrs. Parks for one simple reason: At a time when full-grown men were scared to stand up to injustice, she did. Now I admire her more for something else. Bernadette says that only about one third of the people in line were black. Mrs. Parks managed to transcend race in America while bringing people together and that is a difficult thing.

Saturday, October 29, 2005

Kowari v1.1 in Code Freeze

I am pleased to announce that Kowari v1.1 is now in code freeze. I will set up the CVS repository for v1.2 shortly.

The MP3 and Filesystem Resolvers were removed for this release due to the need to forward-port that tutorial to use the new permanent model names. It was worth it; having permanent model names which are not associated with a URL fixes several bugs and will reduce headaches for people using Kowari on laptops or any host which changes names regularly.

All test are finally passing and we are well on our way to our first real release as a project after Tucana's demise.

Saturday, October 22, 2005

Refuting "Irreducible Complexity"

Bernadette and I attended a fascinating seminar at AAAS on Thursday entitled, "Evolution of Biological Complexity". It was part of AAAS's Dialogue on Science, Ethics and Religion (DoSER) lecture series. The speaker was Chistoph Adami, a professor of Applied Life Sciences at Keck Graduate Institute and a researcher at Cal Tech.

Prof. Adami discussed his most recent research on the Avida software, which simulates Darwinian evolution in silico. The Avida group has managed to show how complex adaptive traits can evolve using only the Darwinian algorithm via the accumulation of mutations. This is a stunning (although, in evolutionary camps, not unexpected) breakthrough in theorectical evolution.

A quick summary of the problem and its recent solution goes like this: Intelligent Designists say that certain biological features are far too complex to have been evolved using Darwin's algorithm. They (especially Lehigh University biochemistry professor Michael Behe) have claimed that since very complex structures (say, for example, the eye) rely on the simultaneous activation of many (tens, hundreds, thousands) of genes, they could not possibly evolve by successive single mutations. This is known as "irreducible complexity", because the complex nature of the structure is not thought to be reducible into a series of single mutations. Adami and Avida have shown otherwise.

Avida has been used to evolve a set of computer programs which compete for CPU time in a race to find solutions to mathematical problems. Adami presented an example of the equality operator, evolved from a series of NAND operations, which required the simultaneous activation of some 20 "genes". He then went on to dissect the history of the evolved operator, showing that it was not active until the last (single gene) mutation, but all the intermediate states were, in fact, functional. Several simpler operators (e.g. AND) were evolved and then lost on the way towards EQU. Surprisingly, the evolutionary train of mutations included several mutations which were downright harmful to the program, severely reducing its fitness. These deleterious mutations survived just long enough to pass on their genes to an offspring which was then mutated in a positive manner. Thus, an examplar structure which met all criteria for "irreducible complexity" was shown to be, in fact, both evolable and emminently reducible.

Importantly, this breakthrough specifically refutes the argument most commonly used by proponents of "Intelligent Design"; that Darwinian evolution cannot account for complex biological features. I hope that this new knowledge can be brought to bear on the recent debates regarding the teaching of evolution in US schools.

Thursday, October 20, 2005

Hurricane Katrina Damage

I've been meaning to post this for a while...

My friend Kirk Benson, who served with me on USS Ouellet, is a Navy oceanographer and was responsible for forecasting weather in the New Orleans area. He lived in Slidell, Louisiana, until Hurricane Katrina blew through. Interestingly, he recommended the evacuation of Navy facilities in New Orleans and was sent to Texas before the storm.

Pictures of the damage to his house are here. It all seems a lot more real when you know someone affected.

Thomas Schelling Wins Nobel Prize for Economics

Thomas C. Schelling won this year's Nobel Prize for economics, and about time, too.

Schelling's article1 on racial segregation, the earliest agent model of which I am aware, was what started my interest in agent modeling, finite autonoma and other ways of generating complexity from simple rules.

The University of Maryland press release is here, which points to some of his other big works. It looks like I have more reading to do...

[1] Schelling, T. 1969. “Models of Segregation.” American Economic Review 59:488-493.

Tour du monde en dansant

For anyone who likes to travel, Tour du monde en dansant is a must-see. Thanks to brownyPoints for the link.

Wednesday, October 19, 2005

Campbell on Memetics?

It is amazing how many scholars have had some internal model regarding the spread of ideas, and many of them seem rather close to (though few are isomorphic with) Blackmore's memes.

I was just reading in Joseph Campbell's Creative Mythology where he says, "A mythological canon is an organization of symbols, ineffable in import, by which the energies of aspiration are evoked and gathered toward a focus. The message leaps from heart to heart by way of the brain, and where the brain is unpersuaded the message cannot pass. The life then is untouched."

That could be Blackmore's memetics, except I think she would tend to say that the last sentence deviates. For, in her theory, an integration of the idea occurs, mixing either more or less with those already resident. Campbell suggests that some ideas may be completely and utterly dismissed if they contradict a functioning mythological system. I think that is worth paying some serious attention to.

Friday, October 14, 2005

Simple Humanism

The history of science may be thought of as the search for simple rules underlying complex phenomena. This search, as the last couple of hundred years clearly show, has been amazingly, even surprisingly, successful. It seems that the world is more simple at its core than we ever expected.

It is not just physics. Sure, Newton ("F=ma"), Laplace (Nature produces "a great number of phenomena, often very complicated, by of a small number of general laws"), Einstein ("E=mc2") and others lead the way, but our contemporaries are busily extending the concept to many other fields. Stephen Wolfram's finite autonoma have sought to place the entire physical world on a simpler footing. Dawkins and Gould extended Darwin's evolution and made it, if anything, simpler. Joseph Campbell found simple underlying rules in worldwide human mythologies. Susan Blackmore, extending Dawkins, is trying to do it for ideas themselves.

I have started to wonder about humanism as a simplification. Humanism would seem to be an expansion of a human world view due to a simplification of the idea of creation. Humanism places humans again in the center of the perceived universe, capable of creating their own world views and solely responsible for their execution. The cause for this shift in perception is directly tied to the successes of simplification in other fields. It worked for physics, biology, economics, so why not God?

Of course, simple (there's that word again) economics tells us that relying on innate human reactions without long term planning (the Tragedy of the Commons) is a quick road to disaster in an economy based on scarcity. Humanism, therefore, is not a silver bullet to solving the world's problems. It just may allow us, though, to look at our problems more clearly and encourage us, however slightly, to plan for our long term survival instead of awaiting Armageddon to clean up our mess.

All that Glitters does not, err, SPARQL

I was asked to review and comment on the specification for the SPARQL Query Language for RDF. The full review is here.

In short, I am impressed that the DAWG was able to produce a specification, but I can't help being disappointed in the results.

The Kowari community plans to implement it anyway, since it will become the standard RDF query language, but will have to extend it significantly. The iTQL language will remain the same, but will be deprecated as of the Kowari release which includes a SPARQL implementation extended to include all iTQL features. There is much discussion whether that is indeed practical.

Tuesday, October 04, 2005

Dates for Kowari 1.1 Release

Kowari version 1.1 dates of interest:

Code slush10 October 2005
Code freeze17 October 2005
Documentation freeze10 December 2005
Release2 January 2006

No code changes will be accepted after the code slush unless they impact currently open bugs of priority 8 or greater. Code changes submitted after the code freeze will be applied to a new CVS branch for the following version.