Anima Ex Machina

I have coauthored, with Jean-Paul Delahaye and Cedric Gaucherel, and made available today on arXiv a new paper entitled Image information content characterization and classification by physical complexity. In the paper we present a method for estimating the complexity of an image based on the concept of Bennett’s logical depth. Unlike the application of the concept of algorithmic complexity by itself, the addition of the concept of logical depth results in a characterization of objects by organizational (physical) complexity. We use this measure to classify images by their information content. The method provides a means for evaluating and classifying objects by way of their visual representations.

The method described in the paper ranks images based on their decompression times and the classification corresponds to the intuitive ranking resulting from a visual inspection, with things like microprocessors, human faces, cities, engines and fractals figuring at the top as the most complex objects; and random-looking images, which ranked high by algorithmic complexity, were ranked low according to the logical depth expectation, classified next to  trivial images such as the uniformly colored, indicating the characteristic feature of the measure of logical depth. A gradation of different complexities were found in the groups between, gradually increasing in complexity from bottom to top.

Along the paper we show that:

The concept of logical depth can be implemented as a feasible and applicable method to approach a real-world problem.

After studying several cases and tested several compression algorithms, the method described in this paper has shown to work and to be of use for identifying and classifying images by their apparent physical complexity.

The procedure described constitutes an unsupervised method for evaluating the information content of an image by physical complexity.

As the theory predicted, logical depth yields a reasonable measure of complexity that is different from the measure obtained by considering algorithmic complexity alone, while being in accordance with one’s intuitive expectations of greater and lesser complexity.

The paper is available here.

Carlos Gershenson, a friend of mine, has developed a suite of games with NetLogo for entertainment at parties. The games have to do with patterns that emerge as a result of the iterative application of  very simple rules by humans or other mobile agents.

Individuals are provided with a single, simple rule at the outset. The outcomes are sometimes independent of the initial conditions and sometimes sensitive to them, but nobody can anticipate them  (except perhaps Carlos and other complexity researchers).Some of the rules are as follows:
- “Approach one”: Each player chooses another player and approaches them one step at a time.  [ Some people ended up in the center of the room while others were  grouped in clusters.]
- “Retreat from one”: Each player chooses another player and then runs away from them. [Everybody ended up on the periphery of the room.]
- “Step between two”: Each player chooses two players, and tries to step  between them. [I had no idea what would happen. As it turned out, everybody ended up in a single tight cluster in the center of the room.]If different rules are issued to different individuals, interesting patterns emerge.

Recently, the New York Times  published an interesting article entitled “From Ants to People, an Instinct to Swarm” with graphs of ants that strikingly resemble  Carlos’ simulations.

As the article points out, people in the U.S. spend 3.7 billion hours a year in congested traffic, but you will never see ants stuck in gridlock. Carlos has himself  worked on improving traffic lights using auto-organization techniques. He recently earned his PhD with a thesis on the subject. Titled 
Design and Control of Self-organizing Systems, it has been published online as as ebook under a CopyLeft licence. It is an enjoyable work.References:
Gershenson, Carlos. Design and Control of Self-organizing Systems. CopIt ArXives, Mexico, 2007. TS0002ENFrom Ants to People, an Instinct to Swarm. New York Times, 2007.

Carlos Gershenson’s suite of games in NetLogo.

The NKS Science Conference 2007 held at the University of Vermont included a special session featuring the contributors to the volume  “Randomness and Complexity: From Leibniz to Chaitin” (see related post),  recently published by World Scientific and edited by Cristian Calude. The session was organized by Calude and myself.

The program was as follows:
9:45am-12 noon
A. Presentations from “Randomness & Complexity: From Leibniz to Chaitin”, Angell Lecture Center B106:

* Cristian Calude, “Proving and Programming”
* John Casti, “Greg Chaitin: Twenty Years of Personal and Intellectual Friendship”
* Karl Svozil, “The Randomness Information Paradox: Recovering Information in Complex Systems”
* Paul Davies, “The Implications of a Cosmological Information Bound for Complexity, Quantum Information and the Nature of Physical Law”
* Gordana Dodig-Crnkovic, “Where Do New Ideas Come From? How Do They Emerge? Epistemology as Computation (Information Processing)”
* Ugo Pagallo, “Chaitin’s Thin Line in the Sand. Information, Algorithms, and the Role of Ignorance in Social Complex Networks”
* Hector Zenil, “On the Algorithmic Complexity for Short Sequences”
* Gregory Chaitin, “On the Principle of Sufficient Reason”

Calude began by talking about  “Randomness and Complexity: From Leibniz to Chaitin”, published to mark Gregory Chaitin’s  60th birthday.

The blog entry of my presentation is posted here:

http://blog.wolframscience.com/

while an extended version of the published paper (co-authored with Jean-Paul Delahaye)  from which that presentation was culled is available here:

http://arxiv.org/abs/0704.1043

Following the  presentations, there was a panel discussion on the subject “What is Randomness?” organized by myself  in collaboration with Cristian Calude (who edited the book), and Wolfram Research’s Catherine Boucher and Todd Rowland. It was held at the Angell Lecture Center and  featured Cristian Calude himself, John Casti, Gregory Chaitin, Paul Davies, Karl Svozil and Stephen Wolfram.

Gregory Chaitin cutting his Omega cake surrounded by Leibniz cookies

We  had a good time discussing various topics of interest  at a  luncheon on the university campus and again at dinner the following night in downtown Burlington. At the luncheon, Stephen Wolfram provided an overview of Chaitin’s prominent career as a pioneer of  algorithmic information theory and then invited Chaitin to cut an Omega cake surrounded by Leibniz cookies.

My paper On the possible Computational power of the Human Mind (co-authored with my BS thesis advisor Francisco Hernández-Quiroz of the Math Department of the National University of Mexico [UNAM], which I delivered as a lecture 2 years ago at the Complexity, Science & Society 2005 Conference at the University of Liverpool, U.K.) has been recently published by World Scientific as a book chapter. It is available from World Scientific at http://www.worldscibooks.com/chaos/6372.html ; also as a paper or from Amazon.
The book is edited by Carlos Gershenson, Diederik Aerts (Brussels Free University, Belgium) & Bruce Edmonds (Manchester Metropolitan University Business School, UK).

Introduction: Scientific, technological, and cultural changes have always had an impact upon philosophy. They can force a change in the way we perceive the world, reveal new kinds of phenomena to be understood, and provide new ways of understanding phenomena. Complexity science, immersed in a culture of information, is having a diverse but particularly significant impact upon philosophy. Previous ideas do not necessarily sit comfortably with the new paradigm, resulting in new ideas or new interpretations of old ideas.In this unprecedented interdisciplinary volume, researchers from different backgrounds join efforts to update thinking upon philosophical questions with developments in the scientific study of complex systems. The paper contributions cover a wide range of topics, but share the common goal of increasing our understanding and improving our descriptions of our complex world. This revolutionary debate includes contributions from leading experts, as well as young researchers proposing fresh ideas.Contents:* Restricted Complexity, General Complexity (E Morin)* Complexity Science as an Aspect of the Complexity of Science (D Mikulecky)* On the Importance of a Certain Slowness (P Cilliers)* Simplicity Is Not Truth-Indicative (B Edmonds)* Why Diachronically Emergent Properties Must Also Be Salient (C Imbert)* Some Problems for an Ontology of Complexity (M McGuire)* Physical Complexity and Cognitive Evolution (P Jedlicka)* Informational Dynamic Systems: Autonomy, Information, Function (W Riofrio)* The Complexity of Information-Processing Tasks in Vision (J Symons)* On the possible Computational Power of the Human Mind (H Zenil & F Hernandez-Quiroz)and other papers