Evolving Evolving … Evolving Evolution

One of my favorite ideas is that the “theory of evolution” is a self-referentially robust. By this I mean that the “theory of evolution” itself can undergo “evolution” and this strengths the theory of evolution (in the sense that theory is more consistent with experiment and history.)
Consider now genetic algorithms where you want to evolve a program to carry out a task. Then we can also talk about evolving the way in which we implement the genetic algorithm which will evolve the program to carry out the task. Well, once you’ve gone to one level of recursion, why not continue on down this path? We can evolve the program which is evolving the genetic algorithm for the program which solves the problem. It seems intuitively clearly (and there are some papers on this, I believe) that (Evo)^(Evo) > Evo, i.e. evolving a program to evolve programs is better than simply evolving a program. So it is interesting to consider (…((Evo)^(Evo))^….^(Evo)). What is the power of evolution recursed?

5 Replies to “Evolving Evolving … Evolving Evolution”

  1. I think you’re falling into H. Spensers black hole of over extending the Theory of Evolution beyond it’s intended or demonstrable bounds. The theory is very strictly applicable to certain types of systems, and is cheapened when kidnapped by meta-physics for imaginary (or ‘complex’, if you like) theological purposes.
    on a meta- theory level, there’s no reason to rehash the Russell-Whitehead PM assumptions that were skewered by Godel in 1931…. and evolution isn’t even a Law or Logical principle, it’s just a physical theory! If we’re discussing “theories of Everything” why don’t we go to the nearest chruch, because Creationism is a “Theory” (although they might claim it’s a law), and it certainly is a “Theory of Everything’. This kind of holy-roller power tripping should be left to preachers, and avoided like the plague by scientists.

  2. Dear sirs,
    I have some ideas, and I think if these ideas can be used with the facilities you have, there will be a great benefit for the mankind.
    This means that a great part of today’s technology and a variety of scientific fields will be affected.
    With using this idea the whole world will encounter a huge mutation in a few years, and there will be a new season in science.
    It should be mentioned that your the first organization I have contacted.Please inform me about your comment on this matter.Thank you.
    Yours Faithfully,
    Adel Mohabatian

  3. 1. I disagree that extending a principle beyond its intended context automatically renders the principle useless and that simply meta-sizing for the sake of meta-sizing is fundamentally wrong. Applying the theory of algorithms to quantum systems doesn’t mean the theory of algorithms won’t work. It means we get quantum computers. Evolving programs to evolve programs is a well defined and scientific task, no? What is theological about a set of experiments that I could sit down and run on my computer? There is no God in my computer.
    2. I do agree that my intro paragraph, where I say that evolving evolution makes it stronger is a statement which is not on the level of a perfect scientific statement. It is the sort of subject a philosopher of scientists, however, would find interesting. That is to say it is not a statement of religion, but a statement about the content of theories, how they change, and in particular the role of selection processes in a theory of selection. This is a process, well not on par with the rigorous (or so we hope!) scientific method, but an idea which tells us something about ourselves. About or frailties and our strengths.
    3. Evolution is NOT a physical theory. There is no fundamental law of evolution which we think governs the evolution of the universe. Evolution is a collection of observations on the complex system of biology which has been turned into a model to produce statistical predictions about the evolutions of these complex systems under different environmental factors. This is far from physics.
    4. I do not understand your claim about a theory of everything. I make no such claim about any such theory in my entry. Godel’s proof does nothing to talk about the power of meta-theorizing except in the completely unrealistic limit of meta-theorizing to the infinite degree. Indeed much of the power of our brain probably comes from the different levels of meta-structure and the power gained from such structures.
    5. I agree that discussing theories of everything as if they are the end all of everything is not scientific. However, conjecturing such theories and testing them is totally scientific. We do not know whether such a theory can exist. But we can test any such theory. Test it over and over again. We can know nothing but certainty, but to limit yourself from talking about a theory of everything is the kind of preaching which should be restricted to churches and U.S. political discussions.

  4. JJ, Alexandar’s argument was that meta-theorizing is a theological exercise and cheapened the theory. I was saying that this was not so. As to whether there is a “programmer” or not for nature, I have found no evidence for or against this proposition.

  5. “What is theological about a set of experiments that I could sit down and run on my computer? There is no God in my computer.”
    Since in the computer, evolution is a program that must be written by you, that proves that if evolution does actually exist in this computer we call “nature” it also required a programmer, “God.” That’s what is theological about it.

Leave a Reply

Your email address will not be published. Required fields are marked *