RE: Denying evolution - couple of curiousities (Full Version)

All Forums >> [Theology] >> Science & Origins



Message

Jhud -> RE: Denying evolution - couple of curiousities (5/27/2008 4:24:20 PM)

quote:

NO they don't. The anti-virus industry describes them has having biological virus qualities, but in practice the computer viruses in the wild are not adapting. Other programmers are changing the virus source code.... ultimately, the biological analogy of a "computer virus" is an incomplete description. Useful to a point, but 100% not useful in a discussion about biological evolution.


Actually, polymorphic computer viruses do indeed adapt by progammed mutation of their code.



drj11 -> RE: Denying evolution - couple of curiousities (5/27/2008 4:25:26 PM)

quote:

ORIGINAL: swan42

quote:

ORIGINAL: Jhud

quote:

Exactly. Computer viruses evolve because of the two step process of change and selection.


Apparently you aren't much of a programmer. Computer viruses adapt because they were programmed to do so, plain and simple.

NO they don't. The anti-virus industry describes them has having biological virus qualities, but in practice the computer viruses in the wild are not adapting. Other programmers are changing the virus source code.... ultimately, the biological analogy of a "computer virus" is an incomplete description. Useful to a point, but 100% not useful in a discussion about biological evolution.


Exactly. At present, the adaptability some types of computer viruses may seem to have, really isnt a good analogy for biology or evolution, except on some superficial levels. All this talk is taking the analogies way to far.

Metamorphic and polymorphic viruses do not adapt or gain new traits through their reproduction, not even novel ones. Its the exact same algorithm every time. In the presence of selective pressures (AV software) if your system manages to wipe out all but one copy of the virus that it didn't recognize, it cannot reproduce with any more success than the copies that got wiped out. Each copy of itself has the same chance of being recognized by the AV system. It cannot pass on its resistance to its descendants, like a real virus or bacteria.



Jhud -> RE: Denying evolution - couple of curiousities (5/27/2008 4:28:03 PM)

quote:

Oh ok.. then can I conclude you do not have a problem believing the improbable is possible?
If so, then we are just left with arguing over little details like which event is more probable than another and whether one event which is less probable than a 2nd improbable event occurred even though it is more probable for the 2nd event to have occurred. It's just details after that. There's no disparity, just details and more details.


Perhaps, depending on what one believes is true about the universe. If one for instance believed in a multiverse, nothing could reliably be said to be 'impossible'.

For those of us who believe this is the only universe there is, I think some things are impossible, I simply haven’t figured out a way of convincing certain skeptics of that yet.



swan42 -> RE: Denying evolution - couple of curiousities (5/27/2008 4:33:02 PM)

quote:

ORIGINAL: Jhud

quote:

NO they don't. The anti-virus industry describes them has having biological virus qualities, but in practice the computer viruses in the wild are not adapting. Other programmers are changing the virus source code.... ultimately, the biological analogy of a "computer virus" is an incomplete description. Useful to a point, but 100% not useful in a discussion about biological evolution.


Actually, polymorphic computer viruses do indeed adapt by programmed mutation of their code.


Polymorphic computer viruses do not mutate their (analogous to DNA) code, they mutate (randomize) their encrypted format to avoid detection by anti-virus checksum and signature algorithms.

Example using ROT as an example code decryption engine.

Generation 1:
ROT13 (actual code) -> actual code generate new random encryption module using a ROT algorithm

Generation 2:
ROT9 (actual code) -> actual code

The net result is the analogy is still extremely weak for purposes of discussing evolution. If a virus was programmed to truly mutate its code it would not successfully propagate enough.

The challenge anti-virus software is faced with is to only focus on the very small decryption module without raising too many false positives.



Jhud -> RE: Denying evolution - couple of curiousities (5/27/2008 4:38:19 PM)

quote:

Polymorphic computer viruses do not mutate their (analogous to DNA) code, they mutate (randomize) their encrypted format to avoid detection by anti-virus checksum and signature algorithms.

Example using ROT as an example code decryption engine.

Generation 1:
ROT13 (actual code) -> actual code generate new random encryption module using a ROT algorithm

Generation 2:
ROT9 (actual code) -> actual code

The net result is the analogy is still extremely weak for purposes of discussing evolution. If a virus was programmed to truly mutate its code it would not successfully propagate enough.


Well, you are assuming DNA mutates in accordance with the common evolutionary paradigm; I think increasingly it is being shown that that the adaptation of DNA is much more like that of polymorphic program than it is like the mechanisms proposed by NeoDarwinians.

But more to the point, as you seem to have some knowledge in the area - why couldn't a computer code evolve, if, as we are often reminded around here, all that is needed is an imperfect replicator?



drj11 -> RE: Denying evolution - couple of curiousities (5/27/2008 4:40:57 PM)

quote:

ORIGINAL: Jhud

quote:

Polymorphic computer viruses do not mutate their (analogous to DNA) code, they mutate (randomize) their encrypted format to avoid detection by anti-virus checksum and signature algorithms.

Example using ROT as an example code decryption engine.

Generation 1:
ROT13 (actual code) -> actual code generate new random encryption module using a ROT algorithm

Generation 2:
ROT9 (actual code) -> actual code

The net result is the analogy is still extremely weak for purposes of discussing evolution. If a virus was programmed to truly mutate its code it would not successfully propagate enough.


Well, you are assuming DNA mutates in accordance with the common evolutionary paradigm; I think increasingly it is being shown that that the adaptation of DNA is much more like that of polymorphic program than it is like the mechanisms proposed by NeoDarwinians.


Even something as simple as the sickle cell mutations stretch the analogy past the point of usefulness.



Jhud -> RE: Denying evolution - couple of curiousities (5/27/2008 4:43:27 PM)

quote:

Even something as simple as the sickle cell mutations stretch the analogy past the point of usefulness.


I tacked this question onto the last post a bit late, and I am curious about what you think:

But more to the point, as you seem to have some knowledge in the area - why couldn't a computer code evolve, if, as we are often reminded around here, all that is needed is an imperfect replicator?



drj11 -> RE: Denying evolution - couple of curiousities (5/27/2008 4:57:11 PM)

quote:

ORIGINAL: Jhud

quote:

Even something as simple as the sickle cell mutations stretch the analogy past the point of usefulness.


I tacked this question onto the last post a bit late, and I am curious about what you think:

But more to the point, as you seem to have some knowledge in the area - why couldn't a computer code evolve, if, as we are often reminded around here, all that is needed is an imperfect replicator?


The thing that prevents one from drawing a complete coherent analogy between computer viruses and biology is that there really isn't any mechanism for inheritance, at least not yet. Perhaps there could be one day, given some progress in the area. Each time the virus iterates, it is not passing its own successful variation on to its descendant, thereby making them more able to successfully reproduce. They are simply randomly generating a new signature.

Both polymorphic viruses and metamorphic viruses simply use obfuscation techniques to hide themselves. Sort of like a CAPTCHA phrase that attempts to obfuscate letters and numbers from spam bots by distorting the characters and adding random artifacts in the background. The CAPTCHA will never evolve into a new language or alphabet, even though each time you view one, it is different.



swan42 -> RE: Denying evolution - couple of curiousities (5/27/2008 4:57:56 PM)

quote:

But more to the point, as you seem to have some knowledge in the area - why couldn't a computer code evolve, if, as we are often reminded around here, all that is needed is an imperfect replicator?


Computer code can evolve; but hardly anyone can wield analogy properly without both biological expertise and computer science expertise.

Computer code that is intended to evolve is structured to evolve and each generation is analyzed according to an externally or internally applied fitness test. This is the basis for the biologically inspired (though in no way identical) fields of genetic algorithms (data mutates) and genetic programs (code mutates). These approaches work better than classical optimization algorithms when the fitness test is not well understood or difficult to describe.

Often, the genetically inspired approach yields unexpected solutions because even the classical optimization techniques have a tendency to reach a local maximal fitness without ever searching for maximal fitness further along the fitness optimization landscape.

As I mentioned before, computers are not perfect when faced with external influences like cosmic radiation; but we don't keep these broken programs around for a 2nd or 3rd generation to have another go. Instead, either ECC memory corrects the error, or non-ECC memory lets your computer crash. In these cases, 'mutation' occurs, but there is no structure imposed that allow for selection and a 2nd generation.



swan42 -> RE: Denying evolution - couple of curiousities (5/27/2008 5:02:01 PM)

The classic, "Classical" optimization technique is Newton's Method.

http://en.wikipedia.org/wiki/Newton%27s_method

A genetic algorithm can arrive at solutions when Newton's Method fails.



drj11 -> RE: Denying evolution - couple of curiousities (5/27/2008 6:11:54 PM)

Another issue that renders the analogies imperfect between biological and computer viruses, which I'm not quite sure how one could reconcile, is the affect on the hardware/morphology. Real biological viruses can have a real physical impact on the host. Something like HIV will destroy your immune system, whereas a computer virus cannot infect chips and transistors and cause them to blow up and spawn new viruses. Computer viruses really cannot impact the circuitry on your motherboard or CPU. Except in the rare cases of boot sector and bios viruses a simple wipe of the hard drive will render your system as good as new. But in either of those two rare cases, flashing the bios and wiping of the mbr will do the same.

In the biological world there isn't really any clear separation between the "software" and the "hardware".



Page: <<   < prev  4 5 6 7 [8]



Forum Software © ASPPlayground.NET Advanced Edition 2.5 ANSI