11-11-2007

Is There Life on the Sun?

I've long had an interest in the artificial (artificial intelligence, artificial life) as one way toward understanding the natural ... typically, I've read books and articles that approach the artificial from the perspective of science in which there is an implicit focus on terrestrial intelligence and terrestrial life, and I've often wondered about ways I could apply some of this thinking (unfortunately, without ever getting around to really doing anything) ... then, on a whim, I picked up a different sort of book: "Metacreation: Art and Artificial Life" ... this survey of artists that apply various techniques of artificial life in their work is even more engrossing than the work of the artificial life scientists from which they draw ... the reason? these artists aren't just exploring life as we know it - they're out there exploring life as it could be ... theirs is an effort to extract the processes of life from the terra-centric (carbon/water) bias.

In that light, here's a short list of the processes of life (mainly from my earlier readings):

Somewhere in this spectrum (roughly arranged from foundational to advanced), "life" emerges. But the perspective of life as it could be illuminates new things for consideration. For example, under the terra-centric bias, thermodynamic disequilibrium is implicitly considered as a constant net inflow of energy (the transfer of heat from the sun to the planets). But what about energy outflows? We can see in terrestrial rocks the evidence that energy outflows can have morphogenetic properties - the outflow of energy in a cooling rock results in order out of disorder ...

... certainly this order is very primitive and the morphogenetic process is extremely short-lived - so I'm not even going to try to say that "life" emerges in a cooling rock in any real sense. However, the sun does not share the transience of a cooling rock ... the sun has been pouring energy out into the solar system for billions of years. Unquestionably, the sun is in thermodynamic disequilibrium with its environment ... and unquestionably, the sun displays order (e.g., convection currents) where an equilibrium system would display dead homogeneity ... does the sun display any other processes of life? does it metabolize? do ordered structures within it reproduce/evolve? do ordered structures within it interact/co-evolve? ... if so, are these structures alive? (Remember, the sun has had billions of years to work at this - longer even than terra-firma has had to work on us!)

At any rate, perhaps artificial-life-as-art is just the muse I need to start doing something with this stuff ... only time will tell ;-)

Images from my apprenticeship ...

Solitary Rose
(evolved image using Thomas Jourdan's Kandid and Scott Draves' Fractal Flames)


FemAlien
(evolved image using Thomas Jourdan's Kandid)


Starlight
(evolved image using Thomas Jourdan's Kandid)


11-22-2007

I decided to try starting from Scott Draves' Fractal Flames software and see where I could go from there. I downloaded the source code and after a bit of fiddling (it's a few years old, requires old versions of some third-party libraries that are hard to find these days, and the latest Microsoft development tools weren't too thrilled with it) I got it to build. But one of the things I noticed while wading through this process was that the software seems to have grown since Draves' published a paper on it - instead of seventeen components in his Iterated Function System, there were fifty five! And neither of the two user interfaces for evolving Fractal Flame images taps in to the more recent IFS components!

First-things-first, I tried replacing the Fractal Flames executable in the Apophysis distribution - but no luck there :-( ... Apophysis seems to only use the executable for exporting evolved images to the file system - during evolution, it must use an integrated Fractal Flames library. I downloaded the source code for Apophysis, but it was developed using Object Pascal (which Microsoft development tools shun) and so there's really no way for me to modify it to take advantage of all of the more recent IFS components. (Darn! - 'cause it was my favorite of the two user interfaces!)

Next, I downloaded the source code for Kandid, recoded the XML for the Fractal Flames plug-in, regenerated the Java skeletons, and then coded in the commands to use the new IFS components. The result was functional, but with all those new IFS components in play the Fractal Flames software was running horridly slow. So I hacked it to skip over the computation of IFS components that were zero-weighted (which a lot of them are in the Kandid scheme of image evolution) and things seem to be running much faster. The resulting images are also more complex now - since each IFS component is designed to lend a particular shape component to the overall image, the more IFS components you have the more variation in shape. Unfortunately, I'm not seeing much in the way of self-organized complexity (but then, I never much did with the Kandid user interface - which is the main reason why I like the Apophysis user interface better). That being said, one can still evolve an aesthetic result (even if it is missing the self-organized signature of artificial life):

Chainsaw
(evolved image using Thomas Jourdan's Kandid and Scott Draves' Fractal Flames)


Conception
(evolved image using Thomas Jourdan's Kandid and Scott Draves' Fractal Flames)


TV Window Shopping
(evolved image using Thomas Jourdan's Kandid and Scott Draves' Fractal Flames)


11-24-2007

A science-fiction plot:

An artificial-life researcher decides to try and discover an artificial physics where the properties found in human designs (explicit purpose, elegance, simplicity, modularity, etc.) are adaptive for artificial beings that live under the laws of (the artificial) physics. The researcher stumbles upon a promising physics and sets evolution in motion. After simulated eons have passed, the researcher peeks into the artificial world to see how things are progressing. What the researcher finds is a society of elegant and simple beings that can easily swap modules (limbs, organs, etc.) of themselves in and out to adapt to any explicit purpose. The interesting thing is that rather than wait for (simulated) eons to pass for new, more adaptive modules to evolve, these beings have tasked one of their own - an (artificial) artificial-life researcher - to create an artificial physics where the properties found in themselves as well as the property of being a module are adaptive for artificial beings that live under the laws of (the artificial artificial) physics ... ... ... ad infinitum.


12-08-2007

"... even if it is missing the self-organized signature of artificial life ..."

... and here it is, complements of an old cellular automata (see Conway's Game of Life) software program I found on the web ... I was skimming through the large number of CA rule sets when I stumbled across one that had been given the name Digital Inkblots ... at first, it was pretty much a psychedelic light show - but the longer I watched it unfold, the more interesting things became:

In the beginning: random noise.

Time in this universe begins with the appearance of a psychedelic light show.

After a while, the screen becomes nearly monochromatic at any single point in time, but the CA continues to rapidly cycle through all the colors in the rainbow, over and over again (a serious strain on the eyes!) ... all of a sudden something odd happens at the borders of different color shades - off-color "walls" momentarily erupt to separate the different color shades, but some of the walls contain irregularities ...

The walls fade away, but the irregularities persist - and grow! To give you a sense of their stability within the chaotic flashing of this artificial world, here are three snapshots taken only a handful of time-steps apart:



A close up of one of the irregularities shows a very terra-morphic structure - a "cell wall" that keeps the chaotic flashing of the external world from destroying the complexly balanced inner workings of the growing "cell" ...

What happens when "cells" collide? - let's watch!

The cell walls begin to merge at the point where the two cells touch ...

The border between cells continues to disappear ...

We've just watched cell division in reverse!

... and what's most incredible about all of this complexity? - it's all occurring based upon a few simple rules that describe how the state of a given pixel in the CA changes based upon the states of the pixels in its immediate neighborhood - that's all!

Signs of Life (borrowed from the title of a book):

Certainly, this is still a looooooooooong way off from anything that could be called "life" - but it just goes to show the Deep Simplicity that may underpin all life on Earth.


5-21-2008

Rules for autopoiesis from a book I'm reading entitled "Mind in Life":

  1. Semipermeable Boundary: Check whether the system is defined by a semipermeable boundary made up of molecular components. Does the boundary enable you to discriminate between the inside and the outside of the system in relation to its relevant components? If yes, proceed to 2.
  2. Reaction Network: Check whether the components are being produced by a network of reactions that take place within the boundary. If yes, proceed to 3.
  3. Interdependency: Check whether 1 and 2 are interdependent: are the boundary components being produced by the internal network of reactions, and is that network regenerated by conditions due to the boundary itself? If yes, the system is autopoietic.
So the $64K question is this: is the cellular automata above autopoietic? And if so, does it then constitute life?

Referring back to "Mind in Life":

The autopoietic characterization of minimal life, however, is not dependent upon this particular structural arrangement of nucleic acids and proteins (RNA and DNA). The autopoietic characterization is more general and so could hold for an "uncoded life" - for instance, an early terrestrial protocell ...

Hmmmmmmmmm - does an early digital protocell count? ;-)


5-27-2008

According to the author of "Mind in Life", the answer to the above question is "no" ... through various philosophical travels, the author arrives at the conclusion that "life" requires both autopoiesis and cognition (minimally, the ability to sense and respond to the environment) ... but what's interesting is that the author goes on to propose an "it takes one to know one" argument ... that is, if we were an utterly alien type of intelligence that had no concept of "self", "environment", "other", "food", "cognition", and so on, and we were studying a Terrestrial soup, would we be able to get past a physical/chemical description of that soup? ... stated another way, without the phenomenological experience of being human and attributing the quality of "life" to ourselves, would we be able to recognize a bacterium as "life"? (or would we merely see a statistical distribution of chemical reactions?)

Now, I've heard this question posed before (if we were to find extra-Terrestrial life would we even recognize it?) ... but usually it's phrased in a very physical/chemical way - for example, maybe silicon-based (as opposed to carbon-based) life is possible (from an organic-chemistry standpoint; not a computer-chip one!) ... but this is the first time I've heard the question phrased in terms of emergence!

And does this have something to say about controversial thinking such as the Gaia hypothesis? ... and, closer to home, life on the Sun? ;-)


1-1-2012

A book entitled "What Is Life?" seems relevant to this meandering stream of thought. There, Margulis and Sagan essentially conclude that autopoiesis is cognition (i.e., "being is thinking"). For example, certain strains of bacteria often demonstrate a preference for certain kinds of food - when a preferred food is available, a bacteria will select it over another food that can provide sustenance but is less preferred. From this perspective, "consciousness" enters into the universe with the first appearance of an autopoietic system. And while this would appear to infringe upon the exceptionalism that many attribute to humans, it no more does so than the recognition that bacteria and humans also share locomotion in common. It may simply be that the consciousness of a bacterium is as vastly different from human consciousness as a nano-screw proton drive is from musculo-skeletal bipedalism. At bottom, it all comes down to definitions, but a definition of consciousness as an attribute of autopoietic systems no matter what the scale seems to be more clearly defined than one that arbitrarily claims that, say, humans are conscious but dogs are not.

Of tangential interest are Margulis and Sagan's descriptions of the processes of bacterial evolution. Essentially, bacteria can evolve either by leaking/absorbing DNA to/from other bacteria and/or by appropriating/infecting other bacteria. The latter is an instance of what Margulis and Sagan refer to as "symbiogenesis" - eukaryotic organisms such as ourselves can be the result of such bacterial couplings. This is very interesting, given that the field of "evolutionary algorithms" focuses primarily on the evolutionary process of eukaryotic organisms - i.e., mitosis (recombination of chromosomes). That is, by giving short shrift to bacterial evolution, computer scientists may be missing out on two significant evolutionary algorithms routinely practiced in the biosphere.