Cool programs

Here are some cool programs everyone reading this blog should try out:

Aptana Studio

Best free program to write sites with html, css, php, ruby on rails and even iphone applications

Webots

Best robot simulator. The full program costs around $3000 but there is a free version for the Rat’s life robot programming contest

Robotc

This is the best way to program a lego NXT robot. This is unmodified c with a straight forward interface to inputs and outputs. This is the only program that’s not free (1 month trial) and it’s only $30

Dev C++

The best free all platform program to write c and c++ code

Autostitch

Best (free program) for making panoramic pictures from a group of pictures

Into the Cool review

Wow, I haven’t written in a month. After passing through the biology section in my library (I’m always in the engineering, physics, computer science sections) I thought about looking for any theoretical biology/alife books that might be there and I wasn’t dissapointed; I found a book called Into the cool. It was a great read and it made me think about how this could be applied to obtain emergence, to make artificial ecosystems. Doing this could make a new kind of technology; one that doesn’t get ideas from biology but from an artificial biology embeded in technology itself.  For some reason the author doesn’t talk about this but he talks about a really relevant issue; having a unified understanding of biology might be the key to stopping global warming and predicting the effects of other enviromental issues/disasters.

But I also have a few issues with the book; it’s too repetitive, it doesn’t have a lot of diagrams and no equations (it’s for the “lay reader”; come on, if your found out about this book you are looking for more than an essay) and the authors seem to have an implicit dislike of biological simulations and as I explained above they “might” be usefull.

The book’s official site is here: http://www.intothecool.com/  

Karl Sim’s classic video

I just found Karl’s Sims classic video explaining the creatures he came up with using the connection machine.

Horizon episode about Alife

I just found a great posting on you tube. It’s an episode of Horizon, an old BBC series, about alife. It describes most of the things from Steven Levi’s book: Artificial life: the quest for a new creation. I was disapointed Tom Ray was not included.

I know I haven’t posted anything in a while; that’s because I decided to stop trying to post things every 4 days. It was lowering the quality of the content and it was a waste of time considering how busy I am with University. So from now on I’ll keep posting stuff, but stuff that I really find relevant for this blog.  

 

Alife C++ programs

The laboratory of intelligent sytems in EPFL (run by one of the founders of evolutionary robotics Prof. Dario Floreano) has very good multipurpose c++ libraries for artificial life; they are in the Software link under Resources. 

There are also more general science libraries from GNU.

Molecular biology primer

Here is a good unfinished primer on molecular biology (it’s still being written):

http://openwetware.org/images/3/3d/SB_Primer_100707.pdf

Possible life in plasma

Here is an interesting New Scientist article about a computer simulation of dust in interstellar plasma that groups itself into DNA-like structures (coincidence?) and has a lot of features similar to living systems. This is the first example I have seen of people writing a computer program that is not biologically inspired and getting something like an alife program. And I’m sure this might happen again as physics simulations become more realistic, then we can really start generalizing about life-like processes. Here is the paper the article was based on:

http://www.iop.org/EJ/article/1367-2630/9/8/263/njp7_8_263.pdf

Interesting alife paper

I just found a good paper from the Entropy Journal called: Self-Organization of Template-Replicating Polymers and the Spontaneous Rise of Genetic Information. When I read it I was surprised how original it is and how simple open-ended systems might be. It also makes you think that a sytem that displays open-ended evolution might not be possible in a computer but in a mechanical system. Maybe there is something about electromagnetic fields and fluid flow that can’t be recreated in a computer. Probably not and it’s more a question of when but it’s a possibility. It is obiously possible using chemistry because it happened already so I won’t be surprised when Mark Bedau does it.

Artificial life courses

Here are some course websites I have found over the years. They have a lot of information.

Bionics course from Berlin’s Technical University (mostly in German)

“Artificial life as an approach to Artificial Intelligence” course from Indiana University. Taught by Larry Yaeger!

Evolutionary Computation and Artificial Life course from Ben-Gurion University in Israel

“Computational and Biochemical Theories of the Origin of Life” course from the Weizmann Institute of Science in Israel 

Artificial life course from Zurich University (good introductory info)

Digital Evolution and Biocomplexity course from Michigan State University. The wiki has lots of info on avida 

Biologically inspired computing course also from Indiana University 

Complex systems summer course co-organized by the Santa Fe Institute 

Evolutionary and Adaptive Systems course from the University of Sussex in the UK.

Artificial life in MATLAB?

While I was using Simulink (a program that comes with MATLAB) to make a model of a two-wheeled robot (using this equations) I was wondering as I always do: can this be used for making artificial life? After looking for a while on the internet I didn’t anything usefull. Besides EcoLab is supposed to do the same: solve differential equations. And it is made especially for artificial life. But nothing beats “drawing” the differential equations and getting an answer. So I used Simulink to simulate the Predator-Prey equations; otherwise known as the Lotka-Volterra equations. It’s not open-ended alife but it’s the closest I could get to it in an hour in Simulink.

Here is the model:

The initial condition for Integrator is 15 and for Integrator1 it’s 16. Suprisingly you get a similar solution if you use 15 for both initial conditions; in theory both “species” should remain at 15 but because of rounding error the solutions diverge. For the simulation parameters I used a start time of 0, an end time of 10, a fixed step size of 0.001 and the solver “ode5(Dormand-Prince)”.

Here is the solution from the scope:

Here is the output from the XY Graph component that graphs what its name says: