Talk:Avida

From Wikipedia, the free encyclopedia

I removed the reference to natural selection as being without direction by definition. Since natural selection does have direction (requirements such as food, water, and mates) this statement was untrue. - Anonymous user, 12:41pm 09/16/2005

[edit] irreducible complexity?

What is this talk about irreducible complexity? Is this a promotional article for Intelligent Design?

This expirement was partialyl aimed at showing darwinan evolution's capailbiy to to give rise to complex systems ,but the intial stem conditions were not Absoulutly maximized.talk2me

But the concept of "irreducible complexity" stems from the ID/creationist fraction, which is not an accepted scientific theory. I don't think it's good practice to refer to it. Lukas 02:15, 16 January 2006 (UTC)

I removed a link that had nothing to do with Avida, and the last paragraph, which also had relatively little to do with the program. I don't know enough about the official guidelines to know where the line is officially drawn on arguments in articles, but I know enough to see that this article is in a poor state. How does one place a marker asking for quality review? Arguments about Science versus Creationism should be kept to the talk pages of those articles, not in computer science. Cathal 21:51, 25 January 2006 (UTC)

Sorry if I'm treading on toes/speaking out of turn, I'm new. But I do know a reasonable amount about Avida, have run it, and have read some of the papers relating to it. The significance of Avida is that is a program that has been specifically written to address whether or not life/complex systems can evolve. To disregard this in writing about it, and to ignore the claims that the writers make in this context, is to miss the whole point of what this entry ought to be about, surely.--Exiled from GROGGS 12:10, 27 January 2006 (UTC)

Not a problem. Lots of people are new. Are you referring to a particular edit that was made recently, or are you making this point in general? If the latter, just Wikipedia:Be bold, i.e. go ahead and change it! - Samsara contrib talk 15:21, 27 January 2006 (UTC)

What I added (and stood for a month or so before it was deleted) was: Further, the objection to the real evolution of irreducible complexity made by proponents of Intelligent Design is ultimately based on its improbability. The irreducible complexity expressed by the equals operator in Avida is improbable, but not so improbable as to be beyond the search capability of Avida as it generates millions of virtual organisms, and more probable than the universal probability bound. It has not been demonstrated that real irreducibly complex biochemical systems (such as the bacterial flagellum, the blood clotting cascade) are sufficiently probable for it to be reasonable to assume that they could arise by chance. I'll put it back, as it seems a reasonable critique - factual etc. - of what the authors of Avida are hoping to achieve. I realise that the issue of IC and so on are strongly tied up with Intelligent Design - but this program was written specifically to cast light on the evolution debate. Thanks for your encouragement.--Exiled from GROGGS 18:27, 27 January 2006 (UTC)

Well, I can see the weakness of that passage, in that it skims over the weakness of irreducible complexity arguments. To my mind, it is impossible to show that any given biological system is irreducibly complex, so it is not clear that this is the relevant question. I think I've read the paper you are referring to, and it seemed a weak paper to me at the time, because all it shows (if I remember correctly) is that a particular predetermined sequence of symbols will arise in finite time.
The passage you added also does not read well to me, but I'm happy to hear second opinions on this. - Samsara contrib talk 18:45, 27 January 2006 (UTC)
I do not think that the assertion that Avida was written to address antievolution arguments like "irreducible complexity" comes anywhere near "well-substantiated knowledge". Consider the Digital Evolution Lab's own introduction to Avida. There is a section on "Scientific Motivation" which has five elements, none of them being to respond to antievolution arguments. The emphasis given in this article to "irreducible complexity" is out of line with having the article be about Avida. The concern about results from Avida addressing "irreducible complexity" warrants at most a sentence that would link to the irreducible complexity article, but I don't see how any more than that can be sustained and still meet NPOV. Certainly the current paragraph on IC amounts to advocacy of a particular antievolution criticism of Avida, not well-substantiated knowledge. On the one hand, there are about four articles in Science and Nature concerning the results obtained from research using Avida, and set against that there are some websites that say that they were wrong. Wesley R. Elsberry 05:48, 9 April 2006 (UTC)

Avida has little to do with IC - that one specific paper is all that Avida has to do with IC. I've revised that section to give what I feel is the proper amount of emphasis, and to be less advocative. I'll be coming back later to try to expand the parts of the article that are actually about Avida. Michael Ralston 16:39, 19 April 2006 (UTC)

The issue of Avida vs Irreducible Complexity was significant from 2003-2005 - see Lenski's paper "The evolutionary origin of complex features" - and Adami also presented a seminar at AAAS specifically addressing this issue. However, from the point of view of balance, if no claim is being made as to Avida's ability to refute IC, then there's no need to point out that this is disputable, I guess.Exiled from GROGGS 13:27, 14 October 2006 (UTC)

This statement can be false: "but this program was written specifically to cast light on the evolution debate", while the statement that Avida demonstrates that interlocking complexity can be demonstrated to arise in an instance of evolutionary computation is true. The former is, indeed, false, and the latter is, indeed, true. --Wesley R. Elsberry (talk) 05:42, 13 December 2007 (UTC)

I am hesitant to edit -- I've never worked with Wikipedia before -- but this sentence is factually inaccurate:

"Nature' published "The Evolutionary Origin of Complex Features" in 2003, in which the evolution of a mathematical equals operation is constructed of at least 19 simpler, precisely ordered logical operations."

Actually, control conditions reported in that paper showed that _no_ particular simpler logic operations were necessary to the evolution of the "equals" operation. And by no stretch were 19 required: the Lenski, et al, experimental procedure only employed a maximum of 9 logic operations, including 'equals', so there were only 8 simpler operations available.

The anonymous commenter is confusing "logical operations" in Avida with "instructions". Avidian instruction sets are where one looks to find the steps referred to. The "19" figure refers to the number of instructions needed to accomplish the task EQU in the hand-coded Avidian.
We ran the functional-genomic analyses on all 23 pivotal genotypes. The number of instructions required for EQU ranged from 17 to 43, with a median of 28 instructions. Notice that one evolved type apparently needed only 17 instructions to perform EQU, whereas our shortest hand-written program used 19 instructions.
There were 9 "logical operations" which an Avidian in the Lenski et al. experiments could perform to receive a merit bonus. Each logic task required an input/output instruction and one or more instructions from the Avidian instruction set. To see that logical operations and instructions are separate concepts, consult the supplemental page showing the shortest known hand-coded Avidian program sequence of instructions, and compare that to the list of one and two-input logical operations.
Further, the assertion that "_no_ particular simpler logic operations were necessary" to evolve EQU does not fit with this statement from the "Discussion" section of the paper:
The most complex function, EQU, evolved only when several simpler functions were also useful.
The only way that I see to interpret the assertion as having a point is if one stresses "particular", as in identifying one particular simpler logical operation as necessary to the evolution of EQU. However, one cannot read the assertion as saying that the paper establishes that EQU can evolve directly without rewarding any other simpler logical operation; the paper itself explicitly says that is false. --Wesley R. Elsberry (talk) 05:33, 13 December 2007 (UTC)