Sunday, November 3, 2013

Is Artificial Life still alive?

Our journal club on Thursday was presented by my new PhD student, Ernest Liu, about a relatively old (1999) paper on the Evolution of Biological Complexity. The authors study what happens when computer programs are forced to compete for access to CPU time and memory. Those programs that work most effectively are able to reproduce and natural selection does its stuff, the fittest programs survive. The framework for these simulations is called Avida has now been around for 20 years.

The paper itself addresses a really fundamental problem in biology: why are there all these complex living forms around us? The answer the authors suggest is that natural selection acts to reduce randomness and make things which are more structured. There is a lot of technical discussion of how to measure randomness and how to define complexity (they define complexity=genome length-genome entropy), but this is the basic result. Natural selection will make genomes less random.

Naturally, the biologists amongst us were not exactly impressed with this revelation. That natural selection eliminates randomness is more or less true by definition. But there are a few additional insights gained in studying evolution of computer programs. For example, there are sudden fitness jumps in the computer simulations, accompanied by decreases in genome randomness. These are reminiscent of 'biological' evolution and are reproduced in the 'artificial' evolution in Avida.

This leads me to my title 'Is Artificial Life still alive?'. It seems to me that research progress since this paper has been pretty slow. Yes, there are Artificial Life conferences and a society with a journal, but the work I have read here is more concerned with engineering challenges and less concerned with the fundamental questions in biology. One nice recent exception to this trend is a paper by Philip Gerlee and Torbj√∂rn Lundh on cross feeding artificial organisms.

One of the reasons for the failure of Aritificial Life to take off in serious biology research is reflected in the reaction of the biologists in our journal club when they read the paper. This research too often looks like Darwin needlessly translated into 'Entroponeese' or some other obscure mathematical language, without providing any new insight. I still think there is potential here, and Philip and Torbjörns' work reflects this potential. I'm interested to hear if anyone else knows of any other signs of life in Artificial Life.


  1. I do agree with you regarding AL in Europe. The situation and the community is a bit different here in the US. For instance, the AL conference in the US (ALife) has a lot more people interested in biological questions than at the AL conference in Europe (ECAL).

    I'd say that at ALife last year, at least half of the talks had a strong biological component (evolution and molecular biology in particular). An example of that (more) biology-oriented mindset was the fascinating main keynote by Jack Szostak who presented an overview of studies on building artificial cells in vitro. Even the "engineering" keynote by Radhika Nagpal packed a fair amount of biology.

    However it's true that at ECAL this year, I felt a bit alone as a biologist.

    Maybe you should come for the next ECAL14 in New York (July 31-August 2) and see for yourself. I'll be happy to introduce you to this small, yet (re)growing and vibrant community ;-)


  2. Hi David, it's great to hear that y'all are covering ALife papers in your journal club, even old ones. Too many of those old ALife papers go mostly ignored nowadays, yet they provide such fundamental insights into evolutionary processes. I've been re-reading the ALife papers from as far back as 1994, and there's some fantastic work discussed there that the ALife field seems to have forgotten. A shame, really.

    To provide a short answer to your question: By skipping from 1999 to 2010, you missed a swath of brilliant ALife papers. I will list several of them that immediately come to mind below. These would be great to go over in your journal club if you would like to see ALife applied to fundamental questions about evolution. To provide a proper full response, I'll have to write up a blog post, so give me a day or so. :-)

    The evolutionary origin of complex features:

    Evolution of digital organisms at high mutation rates leads to survival of the flattest:

    Evolution of biological complexity:

    Genome complexity, robustness and genetic interactions in digital organisms:

    Adaptive Radiation from Resource Competition in Digital Organisms:

    Task-switching costs promote the evolution of division of labor and shifts in individuality:

    Experiments on the role of deleterious mutations as stepping stones in adaptive evolution:

    Note that some of these papers are from the past couple years, so to answer your question: ALife is still very much alive and well. :-)

  3. Regarding your statement: "... needlessly translated into 'Entroponeese' or some other obscure mathematical language"

    I think that's the wrong way to look at it. Translating these phenomena into measurable quantities is a valuable step toward understanding them. No longer are we talking about a vague phenomenon that we all think we know -- in this case, how "complex" an organism is. What does it really mean for an organism to be "complex"? I bet you could ask your journal club that question, have them write it down privately, and everyone would have a different answer.

    By establishing a quantifiable measure of complexity, now we have a single definition for it that we can apply to all organisms. If that measure doesn't work out, then we're having a much more informative discussion by trying to figure out how to broadly *measure* complexity, rather than arguing back and forth about something vague like whether having 2 or 4 eyes makes the organism is more complex.

    I really appreciate your efforts to teach your biology students more about ALife. We actually have entire courses dedicated to teaching biology students about computational evolution (ALife among that) here at Michigan State University, largely thanks to the BEACON Center for the Study of Evolution in Action (

    In all, great post, and thank you for asking these questions!

  4. Thanks to both of you for the feedback. This was the reaction I was looking for! Its great to see you are alive and kicking. I will have a look at the papers (I've seen some of them but not all).

    I broadly agree with your point Randy about the definition of complexity and we discussed exactly this at the club and again after. I spent quite a lot of time researching definitions of complexity for a Masters course I teach in 'Modelling Complex Systems'. I could write lots about this too, and probably will in a future post.

  5. I think that although not impressive because it refers to what is by now a mainstream idea, this type of papers should evoke something more that contempt among biologists.
    The idea that organisms are able to become more and more complex is not something obvious from Natural Selection's standpoint : As Gould has pointed out, to every more complex form of life corresponds to a simpler equally well adapted one, for instance in the form of a parasite.

    Complexity is favored by the number of feedbacks present. This, in turn, typically generates multi-stability and hence, variety, flexibility and ability to adapt to complex environments. But at the same time, it carries the risk of favoring the nucleation of undesired instabilities. On the other hand in a "less" complex organism, stability will be easier to secure but it will not be flexible as it carries then the risk to not survive in a changing environment. I think the challenge now is here : to deal with systems displaying multistability. Sure, there should be more surprising results than everything going to the same final state as seems to be the case in Adami et al.

    Finally it will be worth to revisit this type of work in the light of the more physico-chemically oriented theory of molecular evolution developed by Eigen and Schuster (see Eigen, M., and P. Schuster. "The hypercycle: a principle of natural self-organization." (1979) or

  6. Thanks David for mentioning our work and for initiating the discussing.

    I'm currently not that involved in the Alife community, but would say that I've been both inspired and transferred a lot of techniques from Alife into my work on bacterial evolution and cancer.

    What I like the most about Alife is the lack of constrains and hence the ability to let your imagination roam free.

    One of my favourite examples of this is Sticky Feet, an evolutionary framework in which "animals" made out of springs and actuators feed of each other and evolve complex structures.

    These studies might not teach us anything particular about biology, but they help us understand the general principles behind complex and evolving systems. Some of these properties might be reflected in biology as we see it on Earth, while others are particular to the Alife system under consideration (or reflected in some existing unknown system).

    In connection to our work with Urdar it's interesting to note that we have, in order to understand the dynamics of the system, constructed simpler mathematical models that are amenable to analysis. The Alife system in this case represents an intermediary between the real system of evolving bacteria and a more comprehensible mathematical representation.

  7. That is really cool, Philip. I think it gets a good balance between what assumptions on the local level are put in and the complexity of the structure that comes out. I wonder if there is any way of getting biodiversity in the 'sticky feet'?

  8. There was a interesting video about Karl Sims' virtual creatures although it was in 1994.

    I think they're more interesting questions that why "complex" creatures (e.g. human) appears although there are still much more simple creatures (e.g. bacteria) on earth, and whether it's an inevitable trend or not, than that whether complexity is increasing or not as time goes on.

    I also want to mention dynamical hierarchies, which should be also associated with these questions. Here is an interesting work by Rasmussen, but I cannot get the full pdf.