S2Focus

(...and not a particularly bright fencepost, mind you...).

I've been popping on over to the theologyonline.com e/c forum every now and then to plumb the depths of bob b's stupidity. (I'm posting as "caerbannog" over there).

If you check out the most recent exchanges between bob b and myself in the "No Free Lunch" thread, you'll see that I haven't hit bottom yet.

<a href="http://www.theologyonline.com/vbulletin/showthread.php?s=&threadid=2502&perpage=15&pagenum ber=7" target="_blank">This link</a> will take you to the end of the thread where bob b's stupidity really comes through in all its hilarity. .


(edited to change "...the most exchanges between bob b..." to "...the most recent exchanges between bob b...")

[ September 18, 2002: Message edited by: S2Focus ]</p>

Camaban

There's a bottom?

Skydancer

Since our last report, bob b's posts have reached rock bottom. He is now attempting to dig a hole.

Dr.GH

Whooeeyyy

That boy is flyin' close to the ground.

Camaban

I meant that with him, from what I've seen, I'm not sure that there actually is a bottom.

whenever you think he has to have hit it by now, you just found out how wrong you were.

RBH

S2Focus,

A thought struck me as I read the quoted material from Wolpert & Macready in thread you provided the link to. The NFL theorems are at base about optimization and the average performance over all search algorithms over all fitness landscapes. Your W&M quote says
This is just a nascent thought at the moment, but evolution, and RM&NS in particular, is not really a global optimization process. It isn't 'looking for' an "extremum of a cost function." It's a local optimum finder: it doesn't give a damn about global optima. Why under those circumstances should one even think about NFL?

RBH

[ September 18, 2002: Message edited by: RBH ]</p>

S2Focus

You are absolutely right -- there's no good reason to discuss the supposed implications of NFL with regard to evolution. But Dembski has been trying to flog evolution with his (mis)interpretation of the NFL theories, and has been able to convince a lot of loud-mouthed simpletons (like bob b) that he's onto something big.

But trying to argue your very valid point to a typical Dembski follower would be rather like trying to get a pig to read and understand Principia Mathematica.

So I've been trying a less subtle, "dumbed-down" (and admittedly sloppy) line of reasoning in an attempt to keep my argument from sailing over poor bob b's head. Basically, I've been trying to do something more along the lines of getting a pig to read "The Cat in the Hat" (without much success).

RBH

Prolly 'cause you won't let the pig move its lips.

RBH

HRG

Your point about local vs. global optima is certainly a good one; but IMHO it is even more important that the NFL theorem averages over all (or a large class of) conceivable cost/fitness/whatever functions. However, the class of fitness functions which are relevant for evolution is much smaller and NFL cannot be applied to it.

Regards,
HRG.

Jesus Tap-Dancin' Christ

You know, this is sad. Arguing to a brick wall, yet somehow losing ground.

Who put that netwarrior link up? This guy classifies as a boulder.