Left Home

05 | Hillis
19 | to Friends
19 | Pollack
08 | Taylor
2006: December
03 | Lemmon
12 | Krasnogor
10 | to Friends
03 | Pollack
19 | Wolfram
17 | Travisano
15 | Jewett
13 | Ierymenko
11 | Wallis
07 | Cobb
04 | Buckley
02 | Barbalet
27 | Kampis
24 | Hutton
05 | Klyce

Discussion about the Evolution Prize

If you've become so convinced that a particular theory is right, then you don't do the experiments to prove that it might be wrong. — Denis Noble, "Physiology and the revolution in Evolutionary Biology," 14 Jan 2013

Down Prev | Next Up
5 Aug 2008 | from Dave Hillis | evolutionary computation and cosmic ancestry

10:25 AM: Brig Klyce, I learned about panspermia.org from the INNOCENTIVE challenge. I like that topic and I even went so far as to dust off an old program I'd written. I would like to participate, but the requirements don't make sense to me so far.

I also note that, in order for me to join your Google group, I would have to permit Google to install spyware (web use monitoring software) on my computer. I'm guessing I'm not the only one who will have problems with that. I do read the comments posted there, though.

I have an idea that you might find useful. It doesn't fit into your challenge, so I will just pass it along.

I think that you could get a better reception from the evolutionary computation (EC) community if you approached them differently. Your GP-97 paper is likely to come across as you telling the audience that their algorithms won't work. That's a good way to get them to tune you out immediately. I suggest that an indirect approach would be better received.

You could actually make an interesting case that when we use EC to solve hard problems, we almost invariably find ourselves resorting to algorithms that more closely model Cosmic Ancestry than they do strict Darwinism. That, in turn, might suggest new ways to try to make EC better at solving problems. And solving problems is what evolutionary computation is all about, not modeling biology. Naturally the EC and ALIFE communities overlap, and EC is heavily inspired by biology, but the fields are different and the point of view is different.

After reading much of panspermia.org, I looked at an old project of mine and tried to analyze it from that point of view. I found it quite interesting. For instance, I noticed that my program already has an operator for horizontal gene transfer between species. It plays a fairly minor role; perhaps it deserves another look. And there are some other places, where I see a different way of doing something that might be worthwhile and that I hadn't thought of before.

The canonical GA treats every problem as a black box, uses simple representation and mutation/crossover operators, the smallest possible building blocks, etc. You could say it is a bit like Darwinian evolution. It works very well for some problems, while, for others, it doesn't seem to get any traction at all. But that doesn't stop us because there are a whole host of tricks that have been invented and reinvented to bridge the gap.

These tricks include things like seeding the initial population with "reasonable" individuals, special mutation operators that insert hand-designed genetic material into the population, special repair and local search operators that make it easier to jump from one fitness peak to another, and many others. It tends to make practitioners a little uneasy to modify the algorithm in this way because it seems to be moving away from the natural world that inspired the algorithm in the first place. That's an opening for you to make the argument that these tricks are eerily reminiscent of Cosmic Ancestry, and that the problems requiring them are analogous to the types of problems in nature that require macro-evolution to solve. Perhaps, you might say, they are moving away from Darwinism but not from the natural world after all.

Regards, Dave Hillis

Evolution Prize: You can't get there from here is the subject of additional email from Hillis, posted at panspermia.org, 22 Aug 2008.

13 Jun 2008 | to Eprize friends | another attempt to advance the Evolution Prize

17:35:49 -0500: Dear Eprize Friends -- A challenge titled, "Is Open-Ended Evolutionary Innovation in a Quarantined System Possible?" is posted on Innocentive.com (challenge 6409250 -- https://www.innocentive.com/servlets/project/ProjectInfo.po?f=view&id=6409250. You may have to register first). This challenge grew from an earlier attempt to establish a prize [here] for an answer to this question that was posed to scientists who work in Artificial Life. Now we are offering an award, $20,000.00, to a broader audience for the _design_ of a demonstration of the phenomenon. If that is done, the contemplated main prize can be established.

A blog -- http://groups.google.com/group/OEEI-QS -- has been established as a forum for discussion of the challenge. As the Trustee of the challenge-posting organization (The Astrobiology Research Trust), I will look in regularly and respond as appropriate. If you have thoughtful questions or comments, and especially if you want to solve the challenge, please join in! -- Brig Klyce / Astrobiology Research Trust

4/19/2007 | to Jordan Pollack | question, and video

11:06 AM: Dear Jordan -- ...I doubt that Open-Ended Evolutionary Innovation in a Closed System [OEEI(cs)] is possible. I have tried to come up a way to test the phenomenon. I suggested a challenge for a demonstration. My terms and definitions were not good enough, everyone thought. How, exactly, could the results be measured or confirmed?

You do not doubt that Open-Ended Evolutionary Innovation in a Closed System is possible, right? Why isn't the terms-and-definitions problem also a problem for you? What are the terms and definitions you rely on? What, exactly, is the phenomenon that you do not doubt? What experiment or evidence (besides the big bang) confirms the phenomenon to you? Or is it, in your opinion, still unconfirmed?...Thanks. Best regards. Brig

11:22 AM: Dear Brig, ...You know that my optimism is based on Turing Universality, and the existence of self-organization towards complexity in ecology. But I also think that OEE is "isomorphic" to the Artificial Intelligence problem (which has been unresolved for 50 years so far), so it won't be easy to demonstrate or disprove. The problem is that nobody is really working on it because it is not a priority of defense funding agencies.

So how about setting up a Klyce Institute for Resolving the Question of Unending Evolution (KIRQUE) at Brandeis?

We would have two co-directors working on a "coopetition" framework. One (Me!) working on theory and demonstration (and potential beneficial applications) of open ended evolution in electronic, robotic, and molecular substrates. The other director would be a full professor in Philosophy, who doubts it and works to logically disprove the possibility. The center would endow about 6 professorships and 12 graduate students across multiple departments (biology, computer science, philosophy, biochemistry, physics) to collaboratively focus on your question. There would be an advisory board of which you would be the chair, and the search committees would include nationally renown scholars to hire the best possible people. There would be an annual retreat and annual reporting, monthly seminar series, collaborative grants and growing list of technical reports and papers. Most importantly the development of a new generation of interdisciplinary students who would address your important question on a deepening basis going forward.

locating the Klyce Institute at Brandeis would entail a charitable gift of [lots]. It may seem like a lot, but would cost twice that at Harvard (where they recenty set up a $60M Institute for Innovative Computing and a $30M Program in Evolutionary Dynamics.)

Let me know when you are visiting Harvard again and we can get together.

Cheers, Jordan
Professor Jordan B. Pollack Dynamic & Evolution Machine Org / Computer Science Department / Brandeis University / Waltham MA

11:15 AM 4/20/2007: Dear Jordan -- ...I have asked and you have answered before with Turing Universality, which you also once explained to me on my request. Alas, the explanation left no permanent impression. IIRC, the thing was a bit over my head. Lemme ask you. Is Turing Universality more-or-less the claim that anything from the real world can be modelled in a computer? If so, I get it. Your optimism would be based on the belief that the challenge is simply to get the computer to model the evolution that we observe in nature. Is that right? If it is, I think you are making an unwarranted (big-bang-based) assumption about what we observe in nature.

But if you agree that we may not have observed OEEI(cs) in nature, then Turing Universality must have more power, in your mind at least, than I understand. Can you comment? Is it like the ontological proof of the existence of God? He has to exist because that's part of his definition?

I am not even close to being able to make a gift of [lots], but thank you for thinking so. Best regards. Brig

Thu, 08 Feb 2007 | from Tim Taylor | Another Artificial Life prize (A-Prize)

10:31:30 +0000: Dear all, Maybe most of you folk know about this already, but I just came across details of another prize for artificial life, the A-Prize. Details at:- https://lifeboat.com/ex/a-prize


posted 3 Dec 2006 | from Jay Lemmon | Re: Evolution Prize

11:30 PM 11/21/2006: Let me start by saying I'm probably more sympathetic than other people to the fundamental question of the possibility of open ended innovation. My experience with ALife (Darwinbots) seems to indicate that evolution is far better at breaking patterns than making them. However, I don't think that open ended innovation is impossible. I think it's a matter of need and time span.

In my mind, evolution rarely pushes a creature to be better than it needs to be. Simulations that are placid, eden-esque worlds breed week, uninspiring creatures. A successful simulation that demonstrates open ended innovation will probably require it for survival. Creatures that can't invent some miraculous new solutions to problems will die.

That there is only a single homonid species alive today probably is no accident. The extreme environmental stresses early man had to endure in order to become what we are today were probably too extreme for the majority of other homonid species. As support for this idea, I would use the Toba catastrophe theory. Perhaps a successful simulation would throw down some periodic "mass extinction" events that tax both fitness (killing off the weak) and population sizes (killing off randomly).

Time is also at question. Given infinite time it should be obvious that everything that can be expressed in the genetic language will be observed. I think the real question isn't "Is open ended innovation possible" but "Is open ended innovation achievable a reasonable time frame". More vagueries for you

10:41 AM 12/3/2006: Dear Jay -- Thanks for yours. Sorry to be slow responding.... If OEEI in a code-closed system is possible, it should be demonstratable. How many generations can a computer model run in, say, a year? Isn't that number higher than the number of generations between, say, chimps and humans? Don't some biological innovations appear rather suddenly, according to the fossil record?

To the point at hand, what would you consider a successful simulation actually showing? In my mind, something like demonstrating the Cambrain explosion would fit handedly. If someone could demonstrate the emergence of articulated , discrete multicellularity from unspecialized solitary or colony "algae", I think they would have what you're asking for.

Could you provide a short list of some of the things you would probably consider as valid solutions? I think that will help solidify what a "winning" simulation would actually be, and generate some discussion more along specifics instead of generalities. –Jay Lemmon

For a short list from biology, look at category 10 on the webpage "Macroevolutionary Progress ...Without Gene Transfer?" -- http://www.panspermia.org/harvardprep.htm. For computers, one would be this -- if the machine has the hardware for a phone modem, but not the software, and the ALife program composes the missing software and dials out and "escapes." But I am very uneasy about composing a long list. Someone could take one of the examples and make it the goal of a new program. This is how Richard E. Lenski, Charles Ofria, Robert T. Pennock and Christoph Adami found the "EQU" function in 2003, as described at, "Computer model evolves complex functions?" -- http://www.panspermia.org/whatsne29.htm#030511. As in life, the goal cannot be prescribed. Thanks. Brig

02:37 PM 12/3/2006: Brig, Thanks for the interest. I understand the reluctance of specifically giving examples of what does and does not constitute innovation, but the term as given is vague and I think examples will help illustrate the specifics that words cannot. I have a concern that you might be inadvertantly reinventing your definition of what open ended innovation is as after a successful experiment pushes the boundary of current technology a little further. Similar to the AI effect http://www.aaai.org/AITopics/html/aieffect.html

Specifically, I find this quote from the second link you gave me troubling: "Rewarding intermediate steps assures that the fitness landscape has gradual slopes leading to the prescribed summit. Yet in real life, this feature of fitness landscapes — the availability of gradual routes to new genetic programs — is the very thing we question. Furthermore, in real life the summits are not prescribed." While I understand the point trying to be made, the off hand dismissal of this achievement strikes me as self defeating. I have these two points to make:

1. There is an inherant disconnect between the phenotype and the genotype of organisms. This is fundamentally important, and should not be overlooked. Natural selection works with the phenotype. Mutations occur in the genotype. In real life, DNA is transcribed into RNA, and mRNA, translated into proteins, and these proteins effect chemical changes that interact in potentially complex ways to arrive at an end phenotype.

Avida, as many other artificial life platforms do, condenses this gap and has the DNA really be a shorthand notation for the phenotype. Mutations in the DNA of a program in artificial life should not be seen as a mutation in the genotype but rather _a change in the phenotype_. Clearly the phenotype is the agent for natural selection, and the fitness landscape is defined in terms of changes in the phenotype.

Genes should not be viewed as genetic programs. There is only a very indirect connection between DNA and the fitness landscape an organism is evolving on. "The availability of gradual routes to new genetic programs" is a non issue. You could code a protein that does practically the exact same thing and yet has an entirely different base pair sequence. By the same token, you could code a protein that does nothing even similar to the original and yet is different by only a single base pair in a thousand. Add to this the disconnect between what a protein does and its effect on the phenotype.

Clearly then the forces at work in changing, say, my nose into an elephant tusk aren't concerned with _how the genome_ codes it, but rather the phenotypic path. I understand that for Panspermia the source of mutations in the genome is important. However, interstellar genomes incorporating into existing organisms _are_ mutations of the _genome_. The end effect on the phenotype of that incorporation in an organism is entirely dependant on the rest of the genome.

My point is that the evolution of equals in an avida life form _is_ clear evidence that a phenotype _can_ evolve through feedback mechanisms. It doesn't matter if the feedback mechanisms or even the goal were prescribed by the researchers, that wasn't the point of the experiment. Does it prove that open ended innovation is possible? No, but it does provide a very solid link in the path to that conclusion. Phenotypes can very clearly become "fitter" over time if the fitness landscape is continuous. The questions of "is the fitness landscape of coexisting organisms continous?", and "can the phenotypes of coexisting organisms interact in such a way as to cause a runaway growth in complexity?" are still very valid. As are questions of the way that phenotypes and genotypes relate to one another. They should be explored. And small victories (or defeats) should be acknowledged as such.

2. Again, clearly given infinite time the probability of anything evolving is 100%. The question shouldn't be "is open ended innovation possible?" but "is open ended innovation likely given the very real constraints of survive-or-die?". For instance, suppose the planet starting heating at a rate of 1 thousandth of a degree every year. Would life on Earth be able to adapt, and for how long? What if the rate was 1 degree every million years? (Let's ignore the inevitable death of our sun for a moment.) Would life on Earth be able to adapt, and for how long (or the better question, for how hot). Again for 1 degree every decade. Or every year. Or every minute.

As the limit of the rate of change approaches 0, it should be clear that life will have no problem adapting, since there won't be a change. It should be equally clear that if the rate is infinite (that is, it instantly becomes infinitely hot) no life will survive. The question should be "at what range of rates of change can life adapt, and how well?" Beyond a certain temperature proteins begin to denature. Overcoming this would probably be an example of innovation. As the rate approaches 0, this becomes infinitely likely if it is at all possible (again, infinite time gives infinite diversity). What is the rate at which life begins to be unable to cope.

Expand this to uncountably many dimensions. Heat tolerance, sight, digestion, etc. Feed the phenotypical effects of organisms back into the environment. One organism's waste is another's food source. Create a simulation to model this. Can a feed back loop be created that gives the sort of evolutionary rates that we see in nature? Will organisms be able to adapt fast enough to survive? For instance, if a predator evolves to become more efficient, will the prey have time to adapt before its utterly wiped out (think Humans and Mammoths)? What if one organism learns to create food but causes toxic elements to be released? Will the ecosystem adapt fast enough to overcome the new toxicity? (The advent of oxygen releasing photosynthesis). If change can occur faster than life can adapt, even in a closed environment, this would be a fundamental chink in the present idea of natural selection as it stands.

Compare these results with what the fossil record on Earth says. If the results indicate that the fossil record can be accounted for by natural selection, then Occam's Razor would have us discount sources such as Panspermia as necessary. If in the end the conclusion is that the phenomenon of real life is unreproducable in the current context that we understand it, then we would indeed need to look for additional causes.

Regards, -Jay

begins Sun, 12 Nov 2006 | from Natalio Krasnogor | Updated 20 Nov

16:48:06 +0000: Dear Brig, greetings! how are you doing? and how is your wife? I hope this email finds both of you very well! ...As you know well, I have some reservations (which we discussed in Bloomington) about the way in which the evolution prize is formulated. However I have put two doctoral students of mine to look after the problem posed by the evolution prize albeit in *my* interpretation of it. There are some specific question I wanted to ask you though... When in your paper you say:

"Closed system. The system is closed if no additional instructions in any form, such as keystrokes, commands, viruses, patches, etc., are admitted after the evolutionary process begins. Energy in the form of electricity, sunlight or food, and materials such as air, water, paper or plastic may be supplied as the model requires. System means not just the applications software, but includes the underlying computer operating system and any other software or hardware that might have any role in the process."
am I to understand that for example a Genetic Programming (GP) system will be ruled out because it does introduce new keystrokes, commands, etc in the course pf its evolution? I think that would be too limited. What about reformulating this paragraph on the lines: "... The system is closed if no additional *man made* instructions in any form, such as keystrokes, commands, viruses, patches, etc are admitted after the evolutionary process begins." In this way you could still have entries (like the ones we are planning) that use GP.

The other issue that the above restriction to Closed system has is the following: If we fix the RAM (hardware,+ operating system software to handle it) to a fixed ammount, lets say 1 gygabyte or 10 terabytes, etc, then what happens if our systems does show OEI but runs out of memory? are we allowed to include more RAM (plus operating system to handle it)? or not? I fear that if you fix the ammount of memory then the answer to the prize question is trivially "NO". So, would it be possible to modify the closed system definition in your paper to address the two issues mentioned above?

Also, I think that given that you already allow matter and energy to flow into the system and that you are really interested in programmatic closeness then I would urge you for the sake of making the prize more palatable to the wider scientific community out there (mainly biologist, chemists and physicists and computer scientistis) to change the title from:
"The Evolution Prize: Is Open-Ended Evolutionary Innovation in a Closed System Possible?" to:
"The Evolution Prize: Is Open-Ended Evolutionary Innovation in a Code-Closed System Possible?"

I think that adding this simple prefix to close will remove much (if not all) of the criticism you get. Looking forward to hearing from you. Nat

10:29 PM 11/13/2006: Dear Natalio -- Thanks for continuing to think about this. I agree to adding an additional word to the question. You suggest "code-closed." I have already said "genetically closed" a few times. Either way if it helps. Is "genetically closed" equally good in your mind?

04:40 AM 11/14/2006: Dear Brig, greetings again! and thanks for the quick reply. The reason I think "code-closed" is better than "genetically closed" is because code, understood as "instructions for making something" (even if that something is utterly useless), might end up being introduced into the system by other than genetically means and you want to explicitly forbid man made code *of any sort* getting into the system once the system's evolution kick starts. More specifically (and after speaking with LOTS of chemists & physicists about this after bloomington): when you say that you allow mater and energy to flow into the system it is in principle conceibable to *modulate* that matter and energy flow as to effectively encode instructions, i.e., man made code (of other than genetic nature). In this case, whoemever implements such a system (and I was thinking on doing it as my "entry" into the competition this is why i spoke with chemists and physicists about it) will be effectively "cheating" while simultaneously not answering the question you need answered and wining the prize. Hence, I think the real issue is whether a system which is close to *any* effective man made code (of any implementation, i.e., genetic or otherwise) can or cannot produce OEEI.

As for the memory problem, if that arises, I suggest that the experiment be re-run with the expanded memory installed at the start of the new run. That would remove any questions. On Earth, for example, the memory would be large, but still finite.

This is not quite accurate as there is a flow of material into earth and earth looses material also, e.g. gases to the vacuum, so the "earth memory" is not constant overtime. I think that, given that in your current formulation you already allow matter and energy to flow into the system we should probably explicitely say that memory RAM or hard disk etc and other executing resources (e.g. CPU) are allowed to grow/shrink. This is no contradiction with your current formulation but makes it more clear. As I said before, if the memory amount is fixed in stone i think the answer to the prize is trivially no.

One other clarification to the text of your document that I think might also help is to actually not prescribe that the evolutionary simulation will go in generations that undergo crossover, mutation and selection. This is but *one* model of artificial evolution, namely a sinchronous model, and there are many (probably the large majority of interesting ALife simulators) that actually do not follow a sinchronous model of evolution but rather an asynchronous one.

If it's the system that "introduces new keystrokes, commands, etc" in GP, OK. If a person does it, No. I think you're saying the same thing.

Agreed. The emphasis should be on man-made keystrokes, commands, etc. So GP would be OK if it creates new keystrokes but by itself rather than by the programmers (which generally it does). I am starting get convinced that, were you to use the formulation "man-made code-closed system" then computer scientists, physicists, biologists and chemists will not be able to shy away from the question!. It will not be anymore a matter of disagreeing where the boundaries for the close system are because it will be clear that matter and energy can flow in/out (as per your current formulation) but instructions to make things in *any* form are not. I think this could then raise very profound questions about life and information processing in the universe.

Please let me know what your students uncover! Thanks. Best regards. Brig

I will. Nat

Wed, 15 Nov 2006, 11:13 AM: Dear Nat -- ...I think your rephrasing of the title question is fine.

I am still uneasy about allowing memory to be added after a run begins. I disagree with your view, "if the memory amount is fixed in stone i think the answer to the prize is trivially no." A point often made by origin-of-life researchers is that keeping the process confined and uncontaminated by inappropriate reactions is critical. These researchers prefer a "warm little pond" to a primordial ocean. I still think you should be able to ascertain how much memory you need and make it available at the start. Otherwise we would be exposed to the same danger you mentioned with respect to the modulation of matter and energy flow. But mainly, it's just not as clean with RAM flowing in or out.

Thanks again for your interest. Keep me posted! Best regards. Brig

17:38:32 +0000: ...Brig, "if the memory amount is fixed in stone i think the answer to the prize is trivially no." is not a matter of opinion but of sound mathematics (see the many papers by Crutchfield, wolfram and others) , there will be a finite phase space for the system hence by definition oeei will not be possible. Also, increasing the ram and the operating system only means that you potentially increase the size of the "universe" where your simulated evolution runs, you are not changing the simulated evolution in any sense so it has no further implications to the title question. I think this is crucial otherwise there is simple no potential way of wining the prize (hence nobody will be interested) because a finite phase space will stop evolving sooner rather than later. (note that "sooner" still might mean a very long time but that is not the point).

02:50 PM: Nat, I cannot follow this at all. Let's say you do add memory during the runs, and you think you've won the prize. Total up all the memory that was required, make that much available from the start and rerun it. (You didn't add infinite memory.)

Isn't darwinism built on the assumption that DNA on Earth occupies a finite phase space? Darwinian logic does not require that life on Earth will stop evolving "sooner or later."

Mon, 20 Nov 2006 09:45:31 +0000: In all honesty (and please accept this as constructive criticism) I think the Evolution Prize (EP) formulation as it is being shown in your website is too woolly if I may say, and unless you seriously reformulate it in a way that makes sense to the scientific community it wont work. I think that its quite disapointing to spend time thinking, giving feedback and eventhough that feedback is accepted ...it does not result on a reformulation of the EP. How am I to interpret that on the one hand you say, well that's fine but then the formulation does not change? it is not reasurring to say the least. From reading the posted messages i see that other people as well have given you very reasonable feedback and not much of it has made it to the EP formulation which is a pitty.

[Brig wrote] Total up all the memory that was required, make that much available from the start and rerun it. (You didn't add infinite memory.)
No, i dont think it will work like that, by definition your prize requires that a system that is a candidate for the prize must be observed for very long times so you cannot just stop it re-run it with (again) a fixed memory. By your very definition if the system stops then it is not oeei so the system will need to keep consuming as much memory and CPU resources as it needs to keep working. The whole problem with the prize is that essentially you are asking to recognise and undecidable language (i.e. the language of all the words that encode and open ended evolutionary history). The closer i suspect you will get to formally recognising oeei is through some kind of turing test for oeei. But leaving that "detail" to the side i really think its time to rewrite your draft into a more rigorous language that could really inspire legitimate research. I think there are serious issues with the closeness of the system (as i mentioned before), the type of evolutionary algorith you want to run (as i mentioned before), your conception of optimisation VS innovation, etc, etc, and these need to be addressed.

[Brig wrote] Isn't darwinism built on the assumption that DNA on Earth occupies a finite phase space?
not at all. where that requirement comes from?

[Brig wrote] Darwinian logic does not require that life on Earth will stop evolving "sooner or later."
I did not say that *but* just to clarify: put a shell around the earth and stop energy/matter flowing in/out and then you will see how darwinian evolution does stop sooner rather than later. ...Nat

01:45 PM 11/20/2006: Dear Nat, Thanks for yours of today. I am most grateful to you for your continuing interest. Please take my responses as goodnatured. I will attempt to take up your points in order.

reformulation of the EP--
Your feedback has already become part of the thinking behind the Evolution Prize, simply by being posted on the website. As you know, the prize proper has been moved to the back burner for now, in favor of prizes for the best papers that advance the idea. Also, I would like for the final formulation of the prize wording to be done by a recognized committee or group meeting in one place. We are far from that. Meanwhile I have not made any changes to the prize wording. I want people who read the original abstract to recognize that we are talking about the same prize. Anyway I certainly appreciate your comments and they will not go unheeded when the prize wording becomes official.

must be observed for very long times, etc.--
This was the issue for biological experiments, but computer models can run generations in seconds. And stopping the system does not preclude oeei. (To clarify this, perhaps we should say "Virtually Open-Ended....") In my 2-page abstract I write only, "the process does not halt after one debatable innovation, but produces more."

recognise [an] undecidable language--
The prize criteria, at their present level of precision, could already recognize countless winners, if they emerge. Examples in analogous (open-system) fields, especially biology and commercial computer applications, are abundant. Is it unreasonable to expect such things in closed-system demonstrations?

Yes, working on the language would help when the innovations are not so clear. I had originally hoped that the prize could be launched, and entries or proposals would be submitted, and clarification would come from considering those specific cases. But meanwhile a blockbuster could simply claim the prize and settle the issue.

where that requirement [DNA on Earth occupies a finite phase space] comes from?--
In the Darwinian account: the planet has only so much DNA and at some early stage it had no DNA. OEEI began with limtited DNA, in the Darwinan account.

I did not say that--
Right, sorry. You did not say that life on Earth will stop evolving "sooner or later." You said,
"a finite phase space will stop evolving sooner rather than later," and DNA on Earth is finite. I thought I was just following your logic. I see now that you include energy and mattter in the phase space. In life these are analog variables and not part of the digital sequence space of DNA. In a computer model, must they, too, be expressed digitally? Even if so, they don't have to vary. Can't they be constantly supplied by simple routines? Also, I do not think they must be unlimited in order to observe OEEI. Sure, if the sun goes dark, life on Earth would die out or go dormant. But we have already seen OEEI here. It doesn't have to continue indefinitely.

I don't know why we are having this difficulty. To possibly change the subject, would it be possible to define Innovation by listing some sure examples and extracting their common features? ...Brig

10 Nov 2006 | to: Evolution Prize Friends

Dear Evolution Prize Friends -- Beginning in June I have attempted to launch The Evolution Prize. It is described in the abstract, "The Evolution Prize: Is Open-Ended Evolutionary Innovation in a Closed System Possible?" posted at http://www.panspermia.org/evolutionprize/oeeipossible3.pdf and http://www.panspermia.org/evolutionprize/oeeipossible3.doc

The launch has hardly proceeded at all. The interim idea of awarding a cash prize for a paper advancing the understanding of issues raised by the proposed prize is still the one favored by most of you. The Astrobiology Research Trust hereby announces that it will make such a cash award for papers published ("published" may need more specification) in calendar 2007, with an initial budget of $3,000. This sum could go all to one paper or be split among two or three papers if they merit it. Prizes will be announced by June 1st, 2008, or at an ALife conference around that time.

No committee or board of directors or jury for The Evolution Prize has yet emerged. (However the Cambridge artificial life club, "Greythumb" was nominated by Jordan Pollack when I spoke to them about the prize, 6 Nov.) Until a formal committee is established, the Trustees of ART (Stan Franklin, Lisa Jennings and I) will form the jury. Of course, comments from you friends of the Evolution Prize will certainly count. And you will definitely constitute the Nominating Committee. Nominations should come with your own supporting comments.

To qualify, the papers must advance our grasp of the issues listed in the abstract, or other equally relevant issues. Please prepare to send nominations!

Thank you. Best regards. Brig | Astrobiology Research Trust

posted 3 Oct 2006 | Jordan Pollack

March 2006: Dear Jordan -- Until this process [invention in biological evolution] is demonstrated [in some medium], we can't be sure it works. Maybe what we are witnessing [on Earth] is not automatic design, but the installation, sorting out and optimization of existing programs. I know you think this is unlikely, but if it is even possible, your quoted words should be qualified. "Seems to have automatically designed...?" I am trying to establish a question -- Is Open-Ended Evolutionary Innovation in a Closed System possible?. I thought you agreed that it is a valid question. Yet your recent article treats it as invalid. ...Brig

Mindless Intelligence, by Jordan B. Pollack, p 50-56 v 21 n 3, IEEE Intelligent Systems, May/June, 2006.

Jordan Pollack replies: Evolution is accepted, unqualified by "seems". So is the equivalence of all computer programs to the set which runs on a universal turing machine - a finite state machine with unbounded memory. What I agree is lacking is a demonstration of a formally implemented artificial evolution on a utm which keeps developing....

Later: We agree that there has been no convincing formal/computational demonstration of OEE, despite a decade of my working on it. But I also know that it is "isomorphic" to AI, which has withstood 50 years of many brighter people than I, as well as million dollar chess prizes and turing test prizes.

I plan to keep working on my question, of finding the missing pieces in our explanations of self-organization in Nature (not only regarding Darwinism, but also symbiogenesis, non-linear dynamics, non-equilibrium chemistry, pattern generation, emergent layers and abstraction, etc).

I acknowledge your oft-stated question is not answered satisfactorily from a computational viewpoint. I but can't speak for what biologists or physicists think about the question, as they do not place much value on working computer models.

The question may still be asked in an incorrect way which doesn't enable a scientific process. Much of science progress seems to come from reframing questions in a way which allows testable questions whose answers inform that progress. I think that my framing intelligence as a result of a runaway complexity arms-race - equivalent to a 10 Billion line computer program which no human engineering team could engineer -- allowed me to work on OEE even as a member of the AI field. Working on coevolution for many years, subsequently let to my realizing that a previous answer - competition - as motivator in natural self-organization - was really a question. In particular, "closed system" might be expanded to mean "finite set of primitive elements combining" since the system is open both to incoming usable energy and to random fluctuations.

22 Sep: As I mentioned earlier, like Artificial Intelligence, ALIFE's possibility is "proven enough" by the biological example and the Church-Turing result on universal computation. ...Jordan

23 Sep 2006: Dear Jordan -- I admire your diligence on behalf of AI and related fields of interest. And I am grateful for your patience for my somewhat alien project. I will try to be brief.

Yes, a formal/computational demonstration of Open-Ended Evolutionary Innovation in a genetically Closed System (OEEI/gCS) would be a huge breakthrough. However, the demonstration is lacking not only computer models, but also in biology. The apparent Open-Ended Evolutionary Innovation seen on Earth may actually be accomplished by the installation, sorting out and optimization of existing genetic programs in a genetically open system. You seem unwilling to admit that this alternative exists.

1. Are you unwilling? 2. Why? 3. Or if you are willing to admit the alternative, then do you see that the lack of a demonstration in any medium may be more than just a technical challenge? 4. Because, without a demonstration, how do we know if OEEI/gCS is possible?

Is it possible? The importance of the question seems obvious to me. However, I do not know how to go forward with it. That is why I am seeking the involvement of people like you, who might have suggestions, like your proposed expansion of my "closed system."

But I haven't really gotten to first base. First base would mean you agree that the question is unanswered, in biology or computer models. Second base would mean you agree that the question is important. Thanks again for your patience. Still at bat, Brig Klyce | Astrobiology Research Trust

30 Sep 2006: As you know, I don't think the biological example proves Open-Ended Evolutionary Innovation in genetically closed systems yet, because the biosphere is open to input from space, and the big bang is too remote to be conclusive. If you disagree, I am curious, why?

Mon, 2 Oct 2006 16:00:55 -0400: The logical problems in your thesis are too great. Life is arising out of macromolecular soup on earth, but random dust from outer space is what makes OEE possible? It is like the Clarke 2001 hypothesis that human evolution was triggered by an obelisk. Your idea of "open from space" affecting OEE theory in this context is like "a random number generator", as some science fiction presupposed AI would be impossible without "true randomness". Programs can be written and composed into bigger and bigger programs, and many alife systems inject pseudo-random objects into their simulations, but to presume the simulations wouldn't work without a more sophisticated input floating in from another simulation renders the questions unscientific. If the dust from space was of a higher complexity than the hierarchal self-assembling molecules on earth, it was point to a different planet where higher life evolved on the same chemistry. If not, they would not be able to invade and "colonize" Earth."

[WRT the Church-Turing result on universal computation] This is a rough idea: In early computer science, which was pure philosophy, at least 4 different abstract models arose: Turing's Machine, Church's Lambda Calculus (Lisp), Post's Production Systems (grammar) and recursive function theory, and register transfer systems (so-called Von Neumann machines). The questions which were hardest to answer was as to the set of effective calculations each could perform, and whether one model was more powerful than another, and whether a Turing Machine with an Infinite-state (e.g. ANALOG) automata was more powerful than the machine with a finite state machine, etc.

Eventually, it was proven by several people that all machines could run the exact same set of effective procedures, and that a Finite turing machine was the limit of power, since infinite memory would require infinite access time. In computer science, this is an accepted limit, like the speed of light, and claims of super-turing computing are considered fiction, like warp drive.

This model-equivalence gave everyone in cognitive science the confidence that whatever the brain is calculating, it is within the scope of what other universal machines could compute, brains are not magic or spirit based. My earlier point is that this theoretical certitude in AI expands to nearly everyone in ALIFE, that whatever biological systems are computing, they are not using magic....

12:37 PM 10/3/2006: Dear Jordan -- You have misunderstood me. You wrote: "Life is arising out of macromolecular soup on earth, but random dust from outer space is what makes OEE possible?" No. In the alternative I promote, life *arrives* from space and *nonrandom* genes from space make possible the apparent OEEI on Earth. At least get it right!

You also wrote: "...But to presume the simulations wouldn't work without a more sophisticated input floating in from another simulation renders the questions unscientific." No. First, I simply observe that the simulations haven't worked. Then I describe a scenario in which they don't have to work. And in this scenario, genetic input does not come from "another simulation." I never said that. It may already exist in the environment, implausible as it may seem to you. If simulations ever work, I'll change my mind. You seem already sure that they will. Who's unscientific?

But my main frustration is that you sidestep my questions. I asked you to say whether you agreed that genetic input from elsewhere is even a possible source for the genetic programs that produce apparently new innovations on Earth. You have answered instead that you think the account is implausible. Meanwhile, the mainstream account of evolutionary innovation has no direct supporting evidence, in biology or computer models. It seems that the standard tactic (you're not the only one) is to ignore the difficulties and ridicule the alternatives.

Since this is so, perhaps I should spend more time explaining why [the strong version of] panspermia is not implausible. This would include evidence supporting biology in space, life on Mars, fossilzed germs in meteorites, etc. But the main thrust would be about evolution, and the building evidence for gene transfer as the source for new genetic programs, and why this fact already weakens the darwinian account. And I would show that the darwinian account for new genetic programs is hollow. Would your [regular discussion] group be interested in a presentation like this? (I'll have to learn PowerPoint.) What do you say?...Thanks Jordan. Thanks for all the info on Church-Turing....Best regards. Brig

Tue, 03 Oct 2006 15:06:05 -0400: I certainly mean no ridicule and apologize if it seemed that way. Maybe evolutionary biology suffers from a kind of religious persecution anxiety because of its origins as overturning long held religious conviction. Even symbiogenesis was apostosy and besides HGT violating the "germ cell exclusivity" there is also been discovered lamarkian inheritance of brain chemistry through the placenta. ...But you have to be pressed to say at what point in the fossil record does life arrive from elsewhere?

02:46 PM 10/3/2006: ...In panspermia, microbes are abundant in space and would therefore be arriving all the time, although perhaps more arrive when the comet traffic is heavier. Comet traffic may be cyclical, as when Earth passes through the plane of the galaxy or through a galactic shockwave arm. (But why do I have to be pressed? Darwinists are not pressed to say when, where or how life originated. They can simply say it must have.)

...We should be able to detect alien DNA in the environment which has yet to transfer as genetic input.

Right. please have a look at some of these stories I posted in the past year. Some of them pertain to genes already input, but also still found in lower forms that apparently do not need them.
-- Common bacteria share an infinite gene pool?!
http://www.panspermia.org/whatsne39.htm#051125 -- A small marine worm has complex genes like humans'.
http://www.panspermia.org/whatsne39.htm#051208 -- ...genes ...exclusive to higher animals and some plants
http://www.panspermia.org/whatsne40.htm#060228 -- Can Viruses Make Us Human?
http://www.panspermia.org/whatsne41.htm#060324 -- Viruses And The Evolution Of Life
http://www.panspermia.org/whatsne42.htm#060511 -- ...many more genes in the viral world....
http://www.panspermia.org/whatsne42.htm#060623 -- Viruses' staggering diversity and genetic promiscuity
http://www.panspermia.org/whatsne43.htm#060823 -- ...they hold a reserve of genetic information....

...There is another line of reading you need to do on open-endedness, besides computational theory, which is the history of the "generative enterprise" (Chomsky) in linguistics. There the problem of open-ended invention of new language has a similar epistemological status as "genetic input". Can we truly say new sentences which have not been said before? Certainly new language is formed by creative combination of existing elements (sounds, meanings, cultural references) without extra worldly input....Professor Jordan B. Pollack | Dynamic & Evolution Machine Org | Computer Science Department | Brandeis University

Spiders make unique new webs every day. Our use of language, like their web-spinning, may be the simple expression of our genetic makeup. Chomsky and Pinker talk about a language instinct, which would be genetic, I think. ...Brig

posted 19 Sep 2006 | Wolfram Research

Wed, 28 Jun 2006 06:56:04 -0400: Dear Mr. Klyce: Thank you for the card you sent to Stephen Wolfram, and thank you for coming to our conference. Dr. Wolfram enjoyed his conversation with you, wishes you and your wife the best in your endeavors. He remains unclear, though, what you really mean by "open ended evolutionary innovation". Although it is not yet as widely understood as it no doubt will sometime be, Dr. Wolfram's work changes the intuition about what can and cannot happen in systems, and with many interpretations of your criterion, shows that even very simple programs can satisfy it. Of particular interest to you might be:
http://www.wolframscience.com/nksonline/section-12.10 and

We hope you have a chance to interact with the NKS community and Dr. Wolfram further in the future. Sincerely, Catherine Boucher / Wolfram Science Group / http://www.wolframscience.com

10:51 AM 7/3/2006: Dear Catherine -- I am delighted if Dr. Wolfram is curious about open ended evolutionary innovation. In the hope of clarifying this concept as I mean it, I attach a 2-page pdf, "The Evolution Prize: Is Open-Ended Evolutionary Innovation in a Closed System Possible?" I would be interested to know of any specific NKS example that Dr. Wolfram believes would satisfy the criteria given there. Thank you. Best regards. Brig

11:34 AM 7/17/2006: Dear Mr. Klyce: Thanks for your mail. I did get Dr. Wolfram to look briefly at your piece. I'm afraid his responses weren't altogether positive; he found it rather "woolly". So I don't it's realistic to expect Dr. Wolfram to make a detailed analysis of your piece; if you want to proceed, I suggest that you write a piece dissecting Dr. Wolfram's writings on the subject. Best wishes, Catherine

09:26 PM 8/29/2006: Dear Catherine -- I have now read the readings you (or Dr. Wolfram?) suggested. I was not able to see a clear link between them and the question I am asking. The question is -- In life or any system analogous to it (by having an encoded instruction set (genotype) and a resulting output (phenotype)), is Open Ended Evolutionary Innovation in a Closed System Possible (OEEI/CS)?.

I see that simple programs can produce complex patterns, but this is not analogous to evolutionary innovation. The latter requires a feedback loop -- between genotype and phenotype -- that can write new programs. The phenomena producing the complex patterns that Wolfram admires do not appear to write new programs, but merely to execute existing programs. That this happens "without the need for any additional input from outside..." (p 752 [A New Kind of Science]) is interesting. But where are the new programs?

Elsewhere Wolfram writes, "Algorithmic complexity theory ...asks instead about how large the programs themselves need to be. The results of this book indicate however that even programs that are very small--and thus have low algorithmic complexity--can nevertheless perform all sorts of complex computations" (p 1143). But a bacterial genome cannot generate features possessed by higher animals and plants. The latter require larger genomes with additional programs. The features encoded by these programs would be evolutionary innovations for a bacterium. Does Dr. Wolfram find this concept wooly?

Does he think that prokaryotes and people are exactly equivalent in computational sophistication? Even if he does think that, does he believe that evolutionary iterations on something like a bacterial genome can write an insect genome or a human genome? If so, what evidence does he rely on?

Dr. Wolfram has only a little to say about "meaning" (p 1182). Evolutionary Innovations would be the results of genetic programs with new meaning. Strings of DNA with meaning -- that didn't exist before -- encoded in them. Meaning that is evident because the DNA gets translated into real biological features. This is qualitatively different from the manipulation of shapes on a computer screen. It is also different from physical phenomena like turbulence, where there are no encoded instructions even capable of carrying encoded meaning. This concept gets some elaboration at http://www.panspermia.org/harvardprep.htm -- Macroevolutionary Progress Redefined: Can It Happen Without Gene Transfer?

The underlying question (see 1st paragraph) is important, I'm convinced. Any chance he could get interested? ...Thanks for your kind attention. Best regards. Brig

05:21 AM 9/6/2006: Dear Brig, Thank you for your mail. To be honest at this time Dr. Wolfram is completely booked with work on Mathematica's next major release and does not have any time to devote to other outside research projects including his own. There is just too much to work on right now. But perhaps someone else will find the connections for you. If I come across something relevant I will make sure to pass it along. Best wishes, Catherine

02:11 PM 9/19/2006: Below are some comments from a member of our group. I hope you find them useful.

"It is not clear from his comments that he thoroughly digested the material we recommended to you. The phenomenon of universality is already evidence for the possibility of innovation within a fixed instruction set, because the space of behaviors any such system can reach is precisely the computables, as such. Dr. Wolfram believes systems sophisticated enough to act as universal computers are ubiquitous. Naturally, the states actually visited by any given universal system are a tiny subset of the computables, so working universal systems regularly visit previously unvisited states.

Choosing to regard some such systems as "more sophisticated" than others seems to be a bit subjective and unclear, compared to the concept of universality. We are not in the philosophy business so we need an operational definition (of "sophistication"), and that means a working program. That rhetoric may be the currency in academia, politics, etc., but we are interested in concrete progress.

Encouraging A-life experiments is a fine idea, and a prize like the one he proposes might meaningfully do so. It is not practical though to expect researchers to chase down everything he might mean."

Best wishes, Catherine

03:59 PM 9/19/2006: Dear Caherine -- Thank you. I will posted the entire exchange to date on the panspermia.org/evolutionprize website.... I'm sorry that Dr. Wolfram is so tied up with Mathematica's next major release. I know he will be relieved when the cellular automata take over jobs like that. When does he think that will happen? Meanwhile, Best regards. Brig

Tue, 19 Sep 2006 18:35:53 -0500: Brig, ...Immediately remove the text from your website to avoid any further legal action. ...Benson Dastrup | In-House Counsel | Wolfram Research, Inc.

09:49 AM 9/21/2006: ...Your email requests that I remove from the Evolution Prize website the email exchange with Wolfram Research's spokesperson. I am puzzled, because I made clear in my presentation at the Wolfram Science Conference in Washington DC, 16 June, before the exchange, that a website would be set up and email exchanges about the prize would be posted there. The same understanding was established earlier (4 June) at the Artificial Life Conference in Bloomington Indiana, and later (5 September) at the Cosmic Dust and Panspermia Conference in Cardiff, without confusion. The reason for posting the exchanges is to facilitate and document scientific progress. Stephen Wolfram's thoughts would be especially interesting to many.

But even if there were no valid purpose for posting the exchange with Wolram's designee, the emails posted contain nothing confidential or privileged. The posting of the emails is in no way a copyright infringement. If you feel that further communication about this would be fruitful, please continue, but only for the record. ...Brig Klyce | Astrobiology Research Trust

posted 17 Sep 2006 | Michael Travisano

11:44 PM 9/4/2006: Dear Brig, Although it may be that OEEI has not yet been observed, certainly great effort at demonstrating it has not yet occurred. Far from it. Population genetics is still an active area of research, even after 100 years of investigation, but it does not touch upon the causes of evolutionary change. While there has been active research on the causes of evolutionary change for decades, most of Biology has been focused on the mechanics of daily life (DNA replication, metabolism, etc.). Moreover, what you are suggesting is Experimental Evolution on a major evolutionary transition, which certainly will be achieved, but will be a major advance on what has been done so far. That no one has seen OEEI is by no means a surprise, we are still figuring out far more simple things! Best Wishes, Michael Travisano | Dept. of Biology & Biochemistry | University of Houston

01:53 PM 9/12/2006: Michael I apologize for the haste of this reply. But to claim that something "certainly will be achieved" is simply an expression of faith. (Al Michelson was pretty sure that an ether drift would be detectable. Almost nobody doubted it.) In this case however, most adults in the US doubt it, or would if they understood it. To claim that we already know it anyway is unscientific, I believe. Thanks for your interest. Best regards. Brig

07:06 PM 9/12/2006: Dear Brig, I strongly disagree with your 'mis'-characterization of my comments as "faith". By deciding that we already know enough Evolutionary Biology to make a pronouncement on the origin of life, you are making the same error that Lord Kelvin made when he drastically miscalculated the age of the Earth.

1. Where in my email is the statement, "certainly will be achieved"?
2. Our understanding of the dynamics, much less the causes and origin of evolutionary change is still growing rapidly. Hopefully the following list of some recent Microbial Experimental Evolution pubs will illustrate how rapid understanding in the field is growing. Going ahead and stating that "Open Ended Evolution is Impossible" is pretty much the same sort of thing as stating the Earth is 100 million years old.

  • 1991: First test of the replicability long term evolution. Just 15 years ago this year. Lenski, Rose, Simpson & Tadler (Am. Nat.)
  • 1994: Evolution over 10,000 generations: longest Biological evolution experiment. Lenski and Travisano (PNAS).
  • 1995: Relative roles of Adaptation, Chance and History on Evolution. Travisano, Mongold, Bennett, & Lenski (Science).
  • 1996. Punctuated evolution caused by selection of rare beneficial mutations. Elena, Cooper, and Lenski (Science)
  • 1997: The advantage of sex in evolving yeast populations. Zeyl and Bell (Nature).
  • 1998: Adaptive radiation in a heterogeneous environment. Rainey and Travisano (Nature)
  • 1999: Different Trajectories of Parallel evolution during viral adaptation. Wichman, zBadgett, Scott, & Bull (Science).
  • 2000: Evolvability of an RNA virus is determined by its mutational neighborhood. Burch and Chao (Nature).
  • 2001: Contribution of individual random mutations to genotype by environment interactions in E. coli. Remold and Lenksi (PNAS).
  • 2002: Hybrid speciation in experimental populations of yeast. Greig and Travisano (Science)
  • 2003: An evolutionary advantage of haploidy in large yeast populations. Zeyl, Vanderford, & Carter (Science).
  • 2004: Cooperation and competition in pathogenic baceria. Griffin, West, & Buckling (Nature).
  • 2005: The selective cause of an ancient adaptation. Zhu, Golding, Dean (Science).
  • 2006: Evolution of an obligate social cheater to a superior cooperator. Fiegna, Yu, Kadam, & Velicer (Nature)
Best Wishes, Michael Travisano / Experimentalist

08:51 AM 9/13/2006: Dear Michael -- I'm not the one "deciding that we already know enough Evolutionary Biology to make a pronouncement on the origin of life." That would be the mainstream darwinists. I'm saying the opposite: we don't know enough to even be sure that it does happen. Please scan your email again for the quote you could not find. In general, we will have a better dialog if we read each other carefully. I use the cut-and-paste function, when available, for transcribing quotes.

I have never said "Open Ended Evolution is Impossible." In fact it clearly is possible, because the word "evolution" can mean many things besides innovation. Here's what I say / ask -- Open-Ended Evolutionary Innovation in a Closed System has not been clearly demonstrated in biology, nor in computer models. (Many other phenomena have been, but not OEEI/CS.) Without a demonstration, at this point it is reasonable to wonder if it is possible at all. If it isn't, we don't have to abandon science, as both darwinists and creationist/IDers imply; but we might have to tinker with the big bang theory. So the question is entirely scientific -- Is Open-Ended Evolutionary Innovation in a Closed System Possible? I am genuinely curious. However, a prize for a demonstration is not widely welcome, I'm finding out.

Among the references you list, which are the ones that pertain to the [last above] question? Specifically, which ones allow no influence from a possibly open system such as Earth's biosphere? I am aware that some of the ones on which Lenski and you worked do qualify. Among them, which ones clearly document Open-Ended Evolutionary Innovation? I would like to read them carefully. And I am interested in your considered thoughts about the issue I have raised, but not in mis-characterizations of it! Thank you for your continuing interest. Best regards. Brig | Astrobiology Research Trust

03:20 PM 9/13/2006: Dear Brig, I was over the top, for that I apologize. In terms of "certainly will be achieved" that is in respect to a major evolutionary transition, not just non-life to life. That is one major evolutionary transition, but not the only one. I am sorry that my first email was less than clear, especially so if (non-life to life) is the major evolutionary transition that I think you are interested in. Your clarification of OEEI is very welcome, although I still am uncertain as to the meaning of OEEI. I think you are asking about OEEI meaning the origin of life or perhaps the origin of multicellular life. Other major evolutionary transitions include, the evolution of the endosymbiosis (Eukaryotic cells), the evolution of photosynthesis, the evolution flowering plants, (etc.) all of which had major impacts on the diversity of life on earth. I don't think you mean these other transitions, some of which (e.g., the evolution of endosymbiosis) already have fairly strong experimental (not just phylogenetic) demonstration.

With respect the evolution of life from non-life, we are still just figuring out how living things evolve. That is what all those papers refer to, how living systems change. What distinguishes this work from previous work is that is done purely as experiments, demonstrations. If OEEI is about life from non-life, then they only relate in so far as we understand how selection works in general. These experiments will not provide much information about the necessary initial conditions. If OEEI is about something else, then I have completely misunderstood the issue. Best WIshes, Michael

04:12 PM 9/13/2006: Dear Michael -- Thanks for yours. I mistakenly assumed that you had seen my abstracts about the Evolution Prize. These make clear that I am interested in other major evolutionary transitions. And the abstracts also clarify other related concepts, I hope. The Evolution Prize has a website at http://www.panspermia.org/evolutionprize/. The abstracts, and some other recent discussion, are linked from there. If you're still interested, please have a look. I value your thoughts. Thank you. Best regards. Brig

15 Sep 2006 | Don Jewett | updated

05:52 PM: Dear Brig, Sorry I didn't get to the last dinner [in Cardiff, 8 September], where I had planned to have a further discussion with you. I had mis-judged the train times to Heathrow, where I had to be for an early flight the next day. But on the flight, I read [Astronomical Origins of Life, by Hoyle and Wickramasinghe, Klewer, 2000]. I assume you know the Chap. that starts on p. 55 = Biological Evolution, which outlines the arguments against Darwinism as a means to get large shifts in biological systems.Is this the underlying basis for the Evolution Prize?

If so, I suggest that trying to get someone to do something that is impossible is not much of a proof, when no-one can do it. It would be better to ask for a proof that H&W are correct in positing a limit to the "inventiveness" of evolution. This leads me away from all the things I was going to say to you, and instead will offer some arguments that H&W are correct. Whether it will be possible to offer better evidence, I can't say.

Here is my take on this: I am working on how the cerebral cortex works. In trying to understand the evolution of the brain, it is clear that there are "primitive" functions that are controlled by specific "lower" areas. The differences in "advancement" in species involve addition to these lower areas, followed by enlargement of the "neo" cortex. This has strong parallels to Rodney Brooks examples of building "intelligent" robots, (See his book Cambrian Intelligence). He describes the "rules" as:
1. Get simple functions going, and completely debugged.
2. Add higher-level control that does NOT get inside the lower levels, but either canges inputs or outputs of the lower levels.

In this way-- the system ALWAYS FUNCTIONS, as new capabilities are added. Clearly this is also a requirement for living systems-- to be always survivable, although adding new features. The incorporation of mitochondria into cells, and perhaps even the nucleus, are probably examples of this.

In terms of creativity, it is easiest to take two ideas, each of which works, and then combine them. And so it may be in the cerebral cortex that different "images" that are stored in distributed arrays of cells, may interact not by some chaotic mixing, but by each image "holding its own" as they are juxtiposed. I'll certainly be thinking more on this. Please let me know if my guess about the Evolution Prize is correct, and that these ideas are an issue to the Cosmic Ancestry constructs. ...Best regards, Don J.
Don L Jewett, MD, DPhil / Prof. Emeritus / UCSF

08:15 PM 9/15/2006: Dear Don -- Many thanks for your part in the Cardiff conference, and for your interest in the Evolution Prize. Yes, Hoyle and Wickramasinghe's article that you saw in Astronomical Origins of Life has influenced my thinking about evolution. But no, it is not the basis for the prize. Darwinists are clever about suggesting ways around H&W's arguments. I want to know if any of their suggestions might work. Perhaps one will. But I also want to draw attention to the fact that, so far, none has been clearly shown to work. This lack is generally, actively ignored. ...Thanks. I hope to see you again someday soon. Best regards. Brig

09:01 PM 9/15/2006: Dear Don -- after a longer look at H&W's article, "Biological Evolution," I see that their Section, "Genetics in Open and Closed Systems" (p 65-68), is exactly on point for my prize. They show how evolution could make major advances in an open system. The prize asks if evolution can make major advances in a closed system, and if this can be demonstrated. If it can, the open system is unnecessary. ...Brig Klyce / Astrobiology Research Trust

16 Sep 2006, 02:50 PM, from Don: ...Now, let's *really* talk about the H&W paper and the Prize. Let's say that someone does show that in a computer, one can get major changes from (many) small steps. What does that say about the world? Nothing. It is possible that computers can do it, and life cannot.

04:09 PM, from Brig: To my mind, computers are weaker than biology. If computers can do it, I'll be pretty well convinced that biology can, too.

If no one does it on a computer, does that mean it can't be done? No. Someone might do it. But realize that now-a-days working with genetic systems is common. Therefore, what you really want to know (this is presumptious, sorry); let's re-phrase that: You might want to re-phrase your Prize question: Can big steps be taken by a system based upon small changes in DNA, or are there *other mechanisms* by which higher organisms might inherit operational subsystems that provide for new functionality? (We know that bacteria do this. We are sure that mitochondria did it for eukaryotes. What about higher?).

We may already have a confusion. Bacteria obtain new (new to the obtainer, at least) genetic programs by gene transfer. Eukaryotes probably acquired mitochondria by symbiosis. In both cases the systems originally containing them had to be either 1) open to the admission of the acquired programs; or, 2) if the systems were closed, the acquired programs had to already be in them at the time of the acquisition. (I recognize that mitochondria have epigenetic material as well.)

We know that higher organisms also obtain new programs by various means of gene transfer. (On my viruses webpage -- http://www.panspermia.org/virus.htm --- I list and link to at least a hundred examples under "What'sNEW." There was an interesting new one Thursday.) So the question is the same for bacteria and simpler eukaryotes as for higher ones. All can incorporate programs that are available. But how do those programs get composed? If we only ever see them being acquired, how do we know they ever get composed?

You see, I think the eye example *has been done* on a computer. Recall that the eye was "invented" several times in evolution, and the methods of getting the eye are *different*. So if the "eye" is worthy of the Prize, then it has been done.

Whenever I have looked at references for this, I couldn't see how the eye has "been done." Please provide references. Maybe I haven't seen the one you're thinking of.

Now, it is also clear that by small steps you can get from a hippo-like animal to a whale, where MOST CHANGES ARE NOT to the "basic plan" but to the "extremities". You can also get from a small rodent-type mammal to a bat, where the fingers become wings. So are these changes enough for the Prize?

The prize requires the capability to write new genetic programs. Clearly there are some phenotypic changes that require very little change in the genome. Think of dog breeds. But equally clearly, some changes require new genetic programs such as new exons, multiple exons, whole genes, or suites of genes. My conference abstract mentions a lengthy list of big evolutionary changes by Geerat J. Vermeij, also described at http://www.panspermia.org/whatsne40.htm#060128. I believe that all of these required new genetic programs .

BTW, there is lots of evidence now that newly acquired sequences are subject to a higher-than-average mutation rate, which later settles down. One can imagine that this is how the genome explores the potential (if any) of new sequences, and optimizes them.

What I'm trying to emphasize is that the "jury" needs to decide what will convince it that there is a "big step". Otherwise, each time there is some evidence, it will get rejected, just like the ID folks will always reject something that goes against their beliefs. ...Hope this helps, Regards, Don L Jewett, MD, DPhil / Prof. Emeritus / UCSF

I can easily imagine a demonstration that would convince me. I just haven't seen it. Yes, there needs to be an excellent jury. But people are not volunteering. ...Brig

13 Sep 2006 | to Adam Ierymenko | from Brig Klyce

11:56 AM: Dear Adam -- It was a pleasure to meet you Monday night, and to hear and see your presentation about Nanopond. I congratulate you on the project. ...Ten days ago, I came across your blog entry about me and the Evolution Prize. It was an open posting, so I composed an open reply, but then decided not to actually post it. Now that we have met and I see that you are not nearly as disagreeable as your blog entry, I've re-edited and posted it.... I want to have productive dialog on this subject, but it seems rare. If you think such dialog is possible, please reply. Thanks again for the presentation. Good to meet you. Best regards. ...Brig Klyce

2 Sep 2006: The following comments about the Evolution Prize are posted at http://www.greythumb.org/blog/index.php?/archives/114-More-on-that-alife-X-prize.html#extended. The writer was present at the ALifeX session on June 4th, when the Evolution Prize was first formally proposed. My replies are interspersed (in blue) with his text (in normal black).

13 June 2006 / by Adam Ierymenko ...The evening's meeting started with a wonderful open discussion with a panel consisting of folks like Jordan Pollack, Mark Bedau, and Chris Adami along with these guys who are apparently a new think tank-ish organization promoting bio-inspired design synthesis and engineering.

The main topic of this meeting was setting some goals for the alife field and trying to publicize the field a bit. A PBS documentary was discussed, which is a great idea. Suggestions were collected from those in attendance about what sort of formal milestones might help to advance the field. My suggestion was "exploring Turing-space," something I've written about before and will write more on later.

Jordan made the excellent point that we should avoid setting unachievable, scientifically vacuous, poorly defined, or pseudoscientificy milestones the way the AI field did. As Jordan put it, the AI field promised something like "HAL 9000" before even understanding the problem and has suffered decades of humiliation as a result. To avoid this, Jordan urged that the alife field avoid setting goals like the "Turing test" and instead focus on achieving steady scientific progress. I can't agree more. Artificial life is basically cybernetics and AI reborn, and we shouldn't shoot ourselves in the foot this time by proposing tests that are out of sync with where the science is. So this part of the meeting was great. Then came the circus. Well... I guess it wasn't that bad. But it was a bit... well... interesting...

There's this guy named Brig Klyce. He has a lot of money that he evidently made in clothing or something. He thinks that life came from outer space, and he likes to throw his money around to (among other things) challenge people to prove him wrong. So here's what he wants to do. He wants to know if evolution can produce "runaway complexity growth" in an informatically closed system. By informatically closed system, he means a system in which no ordered information (as opposed to just energy) is allowed to enter. For our purposes he means a computer that's plugged into the wall but that nobody is sitting and typing at and that's not connected to any networks. ...Heavy sigh....

Perhaps Mark Bedau used the words you quote. My question was, "Is Open-Ended Evolutionary Innovation in a [genetically] Closed System Possible?" When we disagree, it is too easy to misunderstand each other. Let's try to be precise.

The crowd's reaction to this wasn't very favorable. There was a lot of concern about defocusing and about the scientific vacuity of the proposed goal. There was a lot of scoffing. Here's a brief summary of the problems with this:

1) How exactly are we supposed to quantify this? As Chris Adami pointed out, we could just take a simulator like Avida and crank down the mutation rate to 0.0000000001 and then it would continue to produce "evolutionary innovation" for "as long as we would care to watch it." The goal is vague and arbitrary.

Defining the goal was acknowledged as a problem needing attention. But the subject question is too important to ignore, isn't it? Or, is the desire to ignore the question motivating these objections?

2) How do we know we haven't produced it already? Simulators like Nanopond or Avida seem to level out, but nobody has provided any kind of formal proof that they must remain so. If you visited Earth three billion years ago, you would see endless mats of bacteria. If you visited two billion years ago, you would see... endless mats of bacteria. If you visited a billion years ago... you get the idea. Redwoods, hula hoops, and Richard Nixon are the very tail end of a very long curve.

I think I understand. Simulators seem to level out, but the phenomenon will be demonstrated, just wait. If someone thinks a promised-someday-demonstration is not sufficient, he is clearly not among the faithful.

3) Since the goal is vague and arbitrary, there was a general sense that depending on how it was defined there would be two possible outcomes: 1) the contest would be won within a week or two or 2) the contest would never be won.

4) Scientists don't like people with ideological agendas. Brig certainly appears to have one. There was a bit of concern about whether this would lead to teabagging from creationists. "See! They have a contest for a hundred grand to produce runaway complexity growth and they can't do it!" Of course, creationists and IDers aren't going to believe anything we do no matter what the result, so who cares? I'm all for not giving a damn about this one. Screw the peanut gallery.

Scientists don't like people with ideological agendas different from their own. Yes, the creationists think Darwinism can't be the whole story. If the ALife community has evidence that closes the case, why not show it? No need to be reticent.

5) It's irrelevant to Brig's notion of panspermia. It's also irrelevant to intelligent design or creationism. Why? Well, for starters, Earth is not an informatically closed system. Earth receives cosmic rays, starlight, asteroids, meteorites, comets, space dust, pulsar waves, and gravitational contortions from other stellar bodies. All of these carry information. So, whether or not a computer with nobody programming it can produce runaway information growth is pretty much irrelevant with regard to the origin question.

As the abstract states, "The system is closed if no additional instructions in any form, such as keystrokes, commands, viruses, patches, etc., are admitted after the evolutionary process begins." Cosmic rays, etc., may be "information" in some sense, but they are not instructions. "Runaway information growth" is not the issue.

There is also no way to prove that aliens or God or spores from space didn't seed us, even if we produce "runaway complexity growth" in a computer simulation. Barring any evidence for or against, panspermia (or creationism, or ID, ...) is an arbitrary assertion. As Chris Adami noted, science does not deal in the arbitrary.

If OEEI/CS is possible, no additional mechanism beyond the darwinian one is needed to acount for the evolution that has occurred on Earth. If OEEI/CS is not possible, another mechanism is needed. This issue is not arbitrary.

6) Complexity growth does not equal progress. In many cases, loss of information and complexity constitutes adaptation. (Any engineer knows this!) So achieving "runaway complexity growth" isn't even relevant to the question of evolutionary progress.

Wrong quote. Complexity is perhaps quantifiable, but its relevance for evolutionary innovation is indirect. Randomness is maximally complex, but randomness is not innovation. Evolutionary innovation is the phenomenon needing study.

7) There is no universal method for measuring information content in any objective way. There's not even an objective definition of information. For example, the output of a random number generator has either high or no information content depending on what metric you use.

I agree. That's why I avoid the term "information." I think the thing to quantify, if possible, is meaning. Or perhaps "programmatic meaning." Yes, it's a hard one.

8) Is there progress in evolution and if so what would it constitute? This is a major area of debate among evolutionary biologists. If there's no solid consensus definition for evolutionary progress, then how are we supposed to know it when we see it? Some people don't even think there is such a thing.

If you are one of those people, please say so. It could save us both some time.

Ok. I'm sick of typing. I could come up with more. ...So there was a lot of concern that Brig coming in and throwing his money around would defocus the field and that his goal was silly. I agree with the latter, and I understand the concern about the former. Jordan had a good idea though: why don't we instead create a yearly prize for the best paper toward the goal of producing complexity growth in an informatically closed system? So we might have just created a yearly prize for the best paper in the evolution of complexity in artificial life. Good job Jordan!

I really don't care if people want to throw their money around and finance science that they find interesting, even if they don't really quite know what they're talking about. I don't see this as a bad thing. But don't hold your breath for this one.

One thing that this latter half of the night's meeting did underscore in my mind is the need to develop commercial applications for artificial life's discoveries. This was mentioned quite a bit in the first half. We need a "virtuous cycle" with industry similar to other fields whose work directly translates into technology. Not only will this provide better funding sources than the government and eccentric millionaires, but it will also provide meaningful feedback about how our ideas are actually interfacing with reality. Engineering application grounds science. I think it's almost a requirement to keep things moving forward. Without it, you drift away from reality and end up out in the land of the theologists and the postmodern literary critics. ...Posted by Adam Ierymenko in Artificial Life at 15:35 [13 Jun 2006]

Does the ALife community no longer claim to have anything relevant to say about evolutionary innovation?

13 Sep 2006 / 09:38 PM: I read your response to some of my comments, and I appreciate it. You're right... I'm not really that disagreeable a person. Everyone seems disagreeable online on blogs and discussion groups, if you haven't noticed. The electronic media seems to have that effect.

I was, however, under the suspicion that you were some form of "intelligent design plant" or something like that. There are people who don't want answers to the sort of questions you seem to be genuinely interested in, and they aren't above resorting to dishonest tactics as I'm sure you're aware. I'm now pretty convinced that you're not.

I'd like to respond, but I'd like to have some time to think about it. There are actually some deep philosophical issues involved here as well as scientific and technological ones.

By the way, I don't speak for the "ALife community," and I don't really think anyone does. It's a general subject matter area, not an organization. I only speak for myself. -Adam

11 Sep 2006 | from Max Wallis

Mon, 17:38:52 GMT0BST: Hi Brig - in our discussion at dinner on Friday, we were making a stab at why your formulation of the "evolution prize" challenge is too narrowly framed. Perhaps it needs a pro-terrestrial evolutionist to help - but I've had a go at the arguments in the attached [1-page doc]. The Vermeij paper that I suppose you mean [Geerat J. Vermeij, "Historical contingency and the purported uniqueness of evolutionary innovations" [abstract], p 1804-1809 v 103, Proc. Natl. Acad. Sci. USA, 7 Feb (online 27 Jan) 2006] appeared in February - I attach a copy in case Pat and Chandra want to read it.

Did anyone else at our [Cosmic Dust and Panspermia ] conference show interest in being involved? It was great to meet you and others - but we must avoid association with unrelated whacky propositions (which is why I shouted rubbish over unlimited speed gravitational waves)! ...Best wishes, Max Wallis | Cardiff Centre for Astrobiology | Cardiff University | 2 North Road | CF10 2DY

7 Sep 2006 | from Robert Cobb

07:42 AM: Dear Brig: Attempting to resolve the question of whether or not evolutionary progress occurs in a closed system by resort to artificial intelligence and computer runs amounts to a setback for science. Science has accepted the challenge of understanding the complexities of life processes locally and cosmically but it remains doubtful, because of the nature of humankind as part of nature, that artificial intelligence or computer runs are inherently adequate for the task posed by The Evolution Prize. In forelawsship, Robert Cobb / Forelaws on Board

posted 4 Sep 2006 | from William R. Buckley | update 15 Sep

Fri, 25 Aug, 08:49:03 -0700: Brig: ...As I said to you at ALifeX, I believe that John von Neumann and John Myhill have already answered your question: open-ended evolution within a closed system is possible. I would be interested to hear the argument of he who disagrees as to the applicability of such statements (those of JVN and Myhill) as demonstrating satisfaction of your question.

The place to look is von Neumann's Hixon Symposium paper: "The description of this automaton E has some further attractive sides, into which I shall not go at this time at any length. For instance, it is quite clear that the instruction IsubD is roughly affecting the functions of a gene. It is also clear that the copying mechanism B performs the fundamental act of reproduction, the duplication of the genetic material, which is clearly the fundamental operation in the multiplication of living cells. It is also easy to see how arbitrary alterations of the system E, and in particular of IsubD, can exhibit certain typical traits which appear in connection with mutation, lethally as a rule, but with a possibility of continuing reproduction with a modification of traits. It is, of course, equally clear at which point the analogy ceases to be valid. The natural gene does probably not contain a complete description of the object whose construction its presence stimulates. It probably contains only general pointers, general cues. In the generality in which the foregoing consideration is moving, this simplification is not attempted. It is, nevertheless, clear that this simplification, and others similar to it, are in themselves of great and qualitative importance. We are very far from any real understanding of the natural processes if we do not attempt to penetrate such simplifying principles." - pages 30-31.

and Myhill's paper, as reprinted in the book Essay's on Cellular Automata: "Finally we shall sketch a proof, from our simple idealised assumptions concerning machines, of the existence of a machine-building machine from which successive generations of descendants become in a certain sense more and more intelligent." - page 206.

Von Neumann shows how systems of automata can be so constructed as to model a genetic mechanism and propagate it. Myhill says that the systems evolve, ad infinitum, with the improvement to the specie being increased intelligence. Successive generations means ad infinitum. What is missing from this model, which is also contained in OEEI/CS? Sincerely, William R. Buckley

08:33 PM 8/27/2006: Dear William -- thanks for yours [above]. The models you cite might prove OEEI/CS, if they work. But by my reading, they're still just proposals at this stage. Am I wrong? ...Brig

Sun, 3 Sep 2006 16:04:23 -0700: Dear Brig: The models mentioned work. That Myhill gives a mathematical proof that machines exist which exhibit specific character tends to move your objection to one of necessitating demonstration. Your original question concerned the reality of OEEI/CS, not identification of an example. However, my own work includes implementation of von Neumann's model, machine self-replication. Indeed, I have been building self-replicators since 1980, for a variety of computing platforms.

The more important concern is probably to get debate within the research community respecting the work I cited. You might pose the issue before Mark Bedau, seeking his opinion.

From a mathematical point of view, the question you posed, to me, seems answered in the affirmative. Of course, you could thereupon refer to Gödel, who showed that a problem exists in the foundation of mathematics. Sincerely, Bill

08:36 PM 9/3/2006: Dear Bill -- The question has always demanded a demonstration. (By logic, Zeno could prove that hares can't overtake tortoises.) If OEEI/CS is possible, it should be demonstrable.

Consider alchemy. Even Isaac Newton was convinced that lead could be turned into gold with ordinary chemistry. Immense effort went into the project of discovering this process. No amount of failure discouraged the alchemists. But it turns out that that lead can't be turned into gold with ordinary chemistry. Is there a lesson here? Maybe, if it can't be demonstrated after much effort, maybe it can't be done? (But maybe there's also a lesson here for me: No amount of failure discourages the faithful.)

Bill, at this point it's not clear that OEEI/CS can't be demonstrated, only that it hasn't been. All I'm saying is this: it's probably time to admit something heretofore not admitted: OEEI/CS might not be possible. Admitting this is the is the ethical and scientific way to proceed. This admission might even motivate new approaches to the question.

Yes it is important to get the question before the research community, as I am trying to do. I have alerted and lobbied Mark Bedau since 2001 at least....Thanks for your interest. Best regards. Brig

01:31 PM 9/15/2006: Brig: After having reviewed the comments of Adam Lerymenko, I must conclude that you ask the impossible. Demonstration follows along the lines of Chris Adami's thinking, and in extension thereof. One might slow Avida, and so cover the life of the observer. One might understand that one lives only so long, so that any demonstration is local, not global, in time. For any model proposed, you personally shall be afforded such time as may remain of your life, and demonstrations require, ultimately, the entire lifetime of the universe before they provide unassailable answers. I do not see how you shall reconcile your goal with the characters of its expression. Your thoughts here are welcome. Sincerely, William R. Buckley

2 Sep 2006 | Podcasts Up! | from Tom Barbalet

Sat, 16:12:11 -0700: ...and Brig, Here are the related links for ...your [18.5-minute interview about the Evolution Prize]. The front of the Biota.org site is the main link; [http://www.biota.org/]. It shows the iTunes subscription and Yahoo subscription information as well as the direct plays through the page. The audio is [http://www.biota.org/podcast/biota_bklyce_090206.mp3]. Incidentally the direct audio links may change in the future. Your best bet is to link through to the iTunes subscription; [http://phobos.apple.com/WebObjects/MZStore.woa/wa/viewPodcast?id=154339649 scroll to 9/2/2006], and the Yahoo page [discontinued] [or use local link.]
Many thanks for the chance to interview you.... Best regards, Tom Barbalet / biota.org / nobleape.com

27 Aug 2006 | from George Kampis

Sun, 20:16:06 +0200: Dear Brig: We met during AlifeX and I was among those expressed interest in the call for OEEI/CS. I agree with some of your points, in particular, that there does not need to be a hidden religious agenda in your call. Yet I disagree with a few other points, in particular, where you say: "But if OEEI/CS is not possible, then we must take highly-evolved life as an initial condition. I think this would mean that either 1) life is the result of a miracle, or 2) life has always existed, and therefore the universe must have always existed."

None of the two conclusions are justified from a philosophy of science point of view. And that's what I am, a philosopher of science.

If OEEI/CS is not possible then we just don't understand something. Either we don't understand geology (and a number of other disciplines), or we don't understand OEEI/CS. That life has emerged on Earth we know from the fossile record (etc: let me skip the long list of other proof steps). How could we invalidate the fossile record (and the items I skipped) and therefore invalidate geology (and the sciences I did not mention) by an Alife study?

We can't. All we can disprove by an Alife study (or any other abstract study) is that, in the given framework of explanation, the explanandum (the origin and evolution of life, in this case) cannot be explained. So what? Facts and explanations don't depend on one another like that. Evolution, as a fact (known from overwhelmingly empirical siences) does not depend on evolution, as an explanation (in overwhelmingly theoretical sciences). Go for better explanations until you cover the facts!

I hope my remarks help clarify what some people (myself included) may think about OEEI/CS: that your agenda may not be religious (or anything definite) but IT IS NOT SCIENTIFIC EITHER.

Many of us (myself included) appreciate your offer, but we can only do our work if science is respected. If you want OEEI/CS to be taken seriously by the scientific community, I guess you should take a more open-minded and pragmatic standpoint. Such a standpoint would not permit a jump into conclusions, like above.

Thanks for your attention, ...regards, -- gk ...George Kampis | http://hps.elte.hu/~kampis

08:26 PM 8/27/2006: (I am blind-copying the Evolution Prize mailing list on this, in case it interests anyone else....) George, I remember a pleasant visit with you in the poster hall at the conference. Thanks for your reply. I appreciate the opportunity to, hopefully, clear up a misundstanding. BTW, You are the first to tell me that my cited conclusion is unjustified by my premises and logic.

I agree with your statement, "If OEEI/CS is not possible then we just don't understand something." But disagree with the next two, "...Either we don't understand geology (and a number of other disciplines), or we don't understand OEEI/CS. That life has emerged on Earth we know from the fossile record."

Life on Earth may have arrived from elsewhere. The genes for all of higher life may have similarly arrived. You may doubt it, but no evidence rules out these possibilities. So perhaps life doesn't originate-and-evolve, but instead arrives-and-develops. This scenario would not challenge our understanding of geology. Geology as a record of life's history on Earth is accepted. In fact, life's rapid appearance and punctuated equilibrium would make better sense this way.

As for understanding OEEI/CS, if you mean our faith in mainstream (darwinian) account of new genetic programs, yes that would be challenged. That's the point of the question. If OEEI/CS is impossible, darwinian evolution cannot be the whole story. Of course, there could be other scenarios besides the panspermia described above. But I think they would have to be either miraculous (unscientific), or open-universe scenarios. Can you suggest another category? Have I misunderstood your earlier suggestions?

The general public, and even many evolutionary biologists, think that ALife can already demonstrate OEEI/CS. This is not true. Whether ALife can demonstrate OEEI/CS is an unanswered question on which I would like to bring attention. If a demonstration succeeds, a huge breakthrough has been achieved, and Darwin gets another affirmation. If demonstrations fail, after much effort, maybe we should question some of our assumptions. I still believe this issue is important. I am sure that it is scientific.

You say, "Evolution, as a fact ...does not depend on evolution, as an explanation." I agree. But here we are using "evolution" in two different senses. Yes, the fossil record tells a reliable, if incomplete, story of the history of life on Earth (evolution 1). But our current explanation -- that darwinian processes can write new genetic programs and make prokaryotes into people (evolution 2) may be wrong. "So What?" you ask. Wow.

Very many people besides me care if our explanation, evolution 2, is wrong. Most of them turn to unscientific alternatives, because mainstream science will not look at the question. I think this is a terrible misfortune.

Thanks again. Comments welcome from any and all. Best regards. ...Brig Klyce | Astrobiology Research Trust

24 Aug 2006 | from Tim Hutton

08:37 AM: Hi Brig, ...I didn't make it to AlifeX but I've heard plenty about the ART prize, so you've made something of an impact at least. Mostly I read people suspecting you/ART of religious motives but I know nothing about this.

I simply have one question for you. In the PDF released at AlifeX (http://www.panspermia.org/oeeipossible3.pdf) you say "it is reasonable to doubt that OEEI in a closed system is possible." Do you not regard the universe as a closed system? ...Thanks for your time, ...Tim -- Tim Hutton - http://www.sq3.org.uk

open reply re: Evolution Prize at Alife X | Friday, August 25, 2006 6:49 AM: Dear Tim -- Thanks for yours [above], which has roused me to recontact the Evolution Prize mailing group by copies hereof. Hi everyone.

I went to Alife X and to the Wolfram Conference in June to promote the question, Is Open-Ended Evolutionary Innovation in a Closed System Possible (OEEI/CS?) and to promote the Evolution Prize to stimulate interest in the question. At both conference sessions there was good attendance and lively discussion. But subsequently the thing has not taken off. For a few weeks I had private email exchanges with about eight of the roughly 40 people who asked to be on the mailing list. But I have sensed a certain standoffishness, as if everyone wants to see what happens before getting too involved. Your comment about "religious motives" is news to me (thanks), and is perhaps the reason people are waiting. If so, I am extremely &%$#&^%$#ing irritated, because I carefully explained to both groups that I am a militant agnostc. My mother considers me an atheist. I have absolutely no religious or other hidden agenda. This position is also made clear on my website at panspermia.org. Yes I have an open agenda, also revealed on the website. It is this: if OEEI/CS is not possible, my bet is on the strong version of panspermia. But I am not seeking interest in panspermia, only in the question.

You asked if I regard the universe as a closed system. I think it is interesting that throughout history, except in Asia, people have tended to believe that the world ends just over the horizon. Same thing now. But the big bang theory is too new and too fluid to serve as the foundation for the rest of science. (Funny that the big bang is the only thing the creationists and darwinists agree on.) Anyway, I don't know if the universe is open or closed. If OEEI/CS is possible, the universe can be either open or closed. But if OEEI/CS is not possible, then we must take highly-evolved life as an initial condition. I think this would mean that either 1) life is the result of a miracle, or 2) life has always existed, and therefore the universe must have always existed. Such a universe would be open, timewise at least. (Multiple universes would all be part of the one universe in my usage.) I am profoundly opposed to the invocation of miracles, so I would opt for for choice 2. In other words, because I have faith in natural causes, I think that the answer to the OEEI/CS question can tell us if the universe must be, or need not be, open. Biology has something to say about cosmology.

Meanwhile, in my opinion, the *direct* evidence for OEEI/CS, in life or any system analogous to it, is shockingly weak. This is the point that I have been trying to sell to others for the past few years. In biology, any new genetic program can be delivered by gene transfer. Very many examples of this are well known. More every week. This fact already challenges the Darwinian account for new genetic programs -- they evolved elsewhere, in an unobserved species or process? To bolster the case that they "originated", the big bang must be invoked. Meanwhile, computer models have not convincingly demonstrated OEEI/CS. But almost no one else, it seems, thinks this weakness is shocking.

Yes there is interest in the prize money. ART's offering it helped me get the audiences, after all. (BTW, ART has given approximately 300,000 dollars to support scientists researching within astrobiology or the OEEI/CS question in the past 10 years.) As I explained to the organizers of ALife X, I want to support the pursuit of this issue by a group more qualified and powerful than myself. So far I perceive that people are more interested in the support than the pursuit.

I still want to assemble a coalition. Hopefully, now that I have clarified my non-religious motive, momentum may pick back up. There was a consensus for a cash prize to the best paper that bears on the subject (OEEI/CS?). I will be pleased if this becomes real. I welcome responses from anyone on the Eprize list, about the process. I welcome submissions for such papers, with jurying to be determined.

...Meanwhile, ART is ready now to support individual or cooperative research that bears directly on the question. Please, anyone, say if you are interested. Tim, thanks for your interest and for the helpful feedback. I hope this email is informative. Sorry to be grumpy. Best regards. Brig

EPrizeDiscussion -- Tim has posted this exchange, including subsequest emails thru 26 Jun 2007.

5 Jun 2006 | EPrize at ALife X | to Evolution Prize mailing list | from Brig Klyce

Mon, 22:51:45 -0400: At the ALife X Conference in Bloomington IN, at a workshop cochaired with Mark Bedau the evening of June 4, I promoted a question, "Is Open-Ended Evolutionary Innovation in a Closed System Possible?" I also posed a followup question, "Can it be demonstrated?", and I proposed a cash prize of $100,000 sponsored by the Astrobiology Research Trust for a successful demonstration in an ALife model.

The workshop generated a high level of interest among the 50 or so workshop attendees and panelists. Some expressed important reservations about the terms of the question, prizes in general, the sponsor's affiliation with panspermia, and other issues. Most of those who spoke agreed that the issue was important, but needs more work.

Eleven interested people met over lunch on Monday to pursue the subject further. With the leadership of Jordan Pollack, we seemed to agree that the prize is too embryonic (Janet Wiles' term) to be simply announced to the world. Needed first are steps that would lead to clarification of the question behind the prize. For these steps, papers could be solicited, with lesser prizes (such as could be funded by the interest on a principal of $100,000) to be awarded for the best paper on a regular basis, such as annually, or biannually.

We discussed scheduling another meeting, but meanwhile we agreed that some kind of internet presence that would allow us to interact easily and frequently would be useful. To facilitate this I have, hopefully, secured the domain name "eprize.org". More on this as it develops.

As mentioned in the June 4 workshop, the EPrize (although it may not yet be a prize) needs a Board of Directors to guide its development. I will ask Jordan Pollack and Mark Bedau to help with nominations, and hopefully there will be a Board of 5 or 7 members within a couple of months. Of course I welcome all who are interested to remain as collaborators and friends.

In summation here is what I am now proposing for further discussion and refinement:

A prize of approximately $3000 to be awarded annually to the best paper pertaining to open-ended evolutionary innovation in a closed system that either 1) advances our knowledge, or 2) further delineates and clarifies a question along that theme.

This prize will be awarded by a jury comprising a board of directors of 5-7 scientists who have credibility and standing in the ALife field. These directors will be chosen in a manner to be determined. The Eprize and its board will have no affiliation with the topic or website "panspermia." Perhaps the first such prize could be awarded at ALife 11.

Let me know what you think. Thanks again for your interest and your support of this inquiry. I appreciate your hospitality and your willingness to provided education and insight into this challenging area.

Best regards, ... Brig Klyce | The Evolution Prize | phone 901-763-2222