Main      Site Guide    
Message Forum
Re: IIRC/response
Posted By: gremlinn, on host 24.165.8.100
Date: Friday, May 12, 2006, at 20:20:29
In Reply To: Re: IIRC/response posted by Sam on Friday, May 12, 2006, at 18:52:12:

> > He would have had to go out of his way to deliberately replicate code just to make it look like evolution had occurred.
>
> This is kind of skewed reasoning, thinking of God in human terms. Nothing is "out of his way" for God to do. Create the universe or light a match; one's about as labor-intensive as the other.
>

I'm not sure it's fair to take that position in an argument. Well, maybe it is fair, maybe even true, but if so you're just closing it off to further discussion. We can't talk or think about anything except in human terms. It's not as if I was trying to prove that God *couldn't* have done something. I was trying to compare degrees of plausibility on our terms, because that's all we can do. Oh well, moving on.

> Moreover, it's even flawed in human terms. I was trying to think of a good analogy, and forgive me that the best I could come up with was PokerBot. The easiest way for me to write PokerBot (a computer program that runs in RinkChat and conducts poker games, supporting many different poker variants) was to first code in a base class, which contained all the functionality that would make it easy to build upon to write a suite of individual variants. So I built in code for different types of initial bets, for example -- antes, bring-ins, and blinds. No variant yet uses all three, but it's easier to build in all three up front, to support whatever combinations I might want to use down the road.
>
> Each variant, then is derived from this base class -- or one of the other variants -- and the only coding I do is to write in what things make that variant distinct. My code for Omaha says to deal four initial cards and to evaluate hands at the end a little weirdly, but otherwise it's just "do what Texas Hold'Em does." When the object is built at runtime, the code for antes and bring-ins and double boards and all that IS THERE, but totally unused.
>
> It would be *more* work to build Omaha up from scratch, even if I did it by copying the common pieces from other variants where I'd already written them.
>
> My silly analogy is, I'm sure, not any more of an accurate representation of the way God thinks and works. But it illustrates the point that the minimal ground-up work required to accomplish a particular task -- whether accomplished by an intelligent being or not -- isn't always the simplest or most natural way to get it done.

I should have waited until I saw the page that wintermute linked to, but now that I've skimmed through that and remembered some things, I've got some more to add.

Your analogy is not complete unless you add in a description of bugs in the code. Let's say that someone is looking at all your bot code as a whole, wondering if you wrote out each one individually or copied bits of functionality. Then he finds a bug in one bot's code -- a strange typo that will cause an error if a certain case happens in the game, but through testing that hadn't happened yet. He then looks at the corresponding code for other bots and finds the SAME typo in every single bot that uses the same routine. (Moreover, say there's ample evidence that you are correct virtually every other time you use similar syntax/expressions/statements). That's pretty solid evidence that you copied bits of code rather than starting from scratch with each one.

It is known that the process of transcribing DNA to pass on to offspring is not perfect -- I think everyone can agree that mutations of this sort take place (that's how bacteria acquire resistance to antibiotics eventually, after all). Often one little error like this will be enough for that whole section of genetic material to cease functioning, and usually that's a bad thing and doesn't help the creature's chances.

Sometimes the changes are helpful and get passed on. Sometimes the changes will mess up the functionality somewhat but in a way that's not really important but which *can* nevertheless be favored by natural selection because of proximity to a gene or part of a gene that's favored on its own. [Note: I'm not very knowledgeable about this last point, but it's discussed somewhere on the page linked to by wintermute.] Geneticists have looked for these errors and found that the same exact "typos" occur in the same places in genes on different organisms, with more error code in common between organisms which were considered closely related by methods independent from genetic research. They interpret this as very strong evidence that the errors occurred in the DNA of a common ancestor and were passed on to descendants via different evolutionary branches.

So far, so good, both sides might say. You could argue that because *we* consider something an error in the genetic code, maybe God doesn't, and he has his reasons, and that's why certain errors appear uniformly across groups of organisms -- because they were in the blueprint code common across those groups, and that's how it was meant to be. I maintain that the deeper you delve into it, the less plausible it looks. But plausibility is subjective, so if you don't buy it so far, I won't bother looking for more detailed theory.

One more thing I'd point out are the occurrences of entire sections of genetic code being transcribed but inserted into the wrong places. That doesn't quite match up with the PokerBot analogy, because if you accidentally copied/pasted an entire section of code, the bug would probably rear its head immediately. But with DNA, such an occurrence wouldn't bring life to a crashing halt unless it happened to interrupt a critical sequence which was in place there already. You'd just have this totally useless sequence of code in entirely the wrong spot.

Personally, I interpret this as even stronger evidence that there are numerous ways that DNA can be seriously mangled over time and (importantly) leave trails about how and when it happened. It also makes it less plausible to me that an intelligent designer purposefully made blueprints with multiple kinds of (seeming, to us humans) errors. Not only did he reuse a lot of the same blueprint sections, but he randomly cut out sections, made some photocopies, and taped the copies over the old blueprints in random places for no reason discernible to humankind.

All to test our faith the more, perhaps. I dunno.

Post a Reply

RinkChat Username:
Password:
Email: (optional)
Subject:
Message:
Link URL: (optional)
Link Title: (optional)

Make sure you read our message forum policy before posting.