A few months ago I had a mini-epiphany about this whole issue of Intelligent Design. It's not actually very useful in the political debate, but I found it interesting in an abstract philosophical way.

I believe in Strong AI. (Well, "believe in" is a loaded term. Let's say "subscribe to the notion of".) Strong AI says, roughly, that any entity or system that behaves like an intelligent consciousness (i.e., that can pass a Turing test) is, in fact, an intelligent consciousness. Another way of looking at it is that there is no distinction between "mind" and "brain"; consciousness isn't something metaphysical apart from the physical brain, it's simply an emergent property of a system, whether it be a functioning human brain or a computer running sufficiently complex AI software. (I guess the bumper sticker would be "Cylons are people too!")

So what does this have to do with Intelligent Design? Simple: the results of the process of evolution over many millennia are indistinguishable from the products of intentional design by an intelligent consciousness; therefore, the process of evolution is an intelligent consciousness. Well, perhaps it's a little weird to ascribe "consciousness" to a process that just performs one specific task (creating new forms of life) and couldn't actually participate in a Turing test. But my point is that it is intelligent, because it produces things that can only be produced by intelligence, and there's no need to posit something metaphysical apart from the physical biosphere of Earth in order to explain humans and finches and platypuses and Venus flytraps.

From: (Anonymous)


"the results of the process of evolution over many millennia are indistinguishable from the products of intentional design by an intelligent consciousness;"


if this were true, so many women wouldn't die in childbirth.

From: [identity profile] dougo.livejournal.com


Hey, nobody's perfect. :)

Actually, I think that's a pretty difficult design project, to make human skulls big enough to hold a brain but small enough to fit through a vagina. There's something about this in How the Mind Works. (Whoa, Google Print rules!)

From: [identity profile] http://users.livejournal.com/little_e_/


nonsense. just have a zipper going down the front of the stumach. we have all this empty space between our ribs and pelvis where we could give birth without damaging ourselves, but evolutionary needs have created a specimum where the eggs and sperm are excreted at the excretion end of the digestive tract. this is clearly an inefficiency caused by our evolutionary pathway, but there's no way in hell we're going to randomly mutate to give birth elsewhere in our bodies.

From: [identity profile] dougo.livejournal.com


I suspect there are good reasons why that wouldn't work. But in general, I think it's clear that given enough time, anything that is physically possible (and better for survival) will eventually be "designed" by evolution. But I'm guessing that some (many?) Intelligent Design proponents wouldn't like the idea that the design process is not "finished" yet.

From: [identity profile] st-rev.livejournal.com


That only applies if you assume that evolution is hyperintelligent and can make huge leaps. I think the conclusion here is that evolution is a hacker and only makes incremental improvements.

From: [identity profile] dougo.livejournal.com


It's also sloppy and lazy: it doesn't bother to clean up vestigial organs because it's more trouble than it's worth.

Oddly, most of the top hits in a Google search for "vestigial" are "Darwinism refuted" sites. Augh.

From: [identity profile] st-rev.livejournal.com


A stupid search algorithm can look intelligent if 1) the space it's searching is sufficiently complex and 2) it has a ridiculously huge amount of time and processing power.

From: [identity profile] artname.livejournal.com


Not to mention this whole "walking upright in gravity" thing. The mother has to NOT deliver the baby when she's upright and walking, and then a few hours later the baby has to come on out.

From: [identity profile] mshonle.livejournal.com


Actually, for birthing it's *best* to be upright (squatting, in a pool of water, or with bolstering from others...).

From: [identity profile] artname.livejournal.com


Right. Upright's great for birthing, but it really sucks to have to spend the nine months before that being upright and *not* birthing.

From: [identity profile] lordjulius.livejournal.com


> But my point is that it is intelligent, because it produces things that can only be produced by intelligence

I'm not sure I see your reasoning here. (By the way, this is also an argument used to demonstrate the existence of God. God exists because the universe could only have been created by God. QED.)

Random chance can, in quantum mechanical theory, give rise to an exact copy of our species. The odds are quadrillions to the quadrillions to the quadrillions against. But it can happen. Does that make random chance intelligent?

I think your logical premise (derived from Strong AI) is faulty. (You almost say as much when you mention that the process of evolution could not participate in a Turing test.) You claim a priori the results of evolution are indistinguishable from the products of intentional design by an intelligent consciousness. I can claim that the results of evolution are indistinguishable from the products of an unintentional design by environmental and genetic factors. The point is we have no way of knowing if our intelligence arose from intelligent design or unintentional design.

From: [identity profile] dougo.livejournal.com


Does that make random chance intelligent?

Yes.

You claim a priori the results of evolution are indistinguishable from the products of intentional design by an intelligent consciousness. I can claim that the results of evolution are indistinguishable from the products of an unintentional design by environmental and genetic factors.

I agree with both of these claims, because the results are irrelevant of intent. Put another way, the definition of intent depends on the definition of intelligence, and I'm defining intelligence as "that which appears to be intelligent".

From: [identity profile] lordjulius.livejournal.com


I still don't understand what your criteria are for defining the appearance of intelligence.

I thought the Turing test defined the appearance of intelligence. Evolution cannot pass a Turing test.

If you define 'intelligence' as "the appearance of intelligence", then it's just a tautology. In my opinion. (I think I understand what you are trying to say, however.)

Doesn't Godel, Escher, Bach have something to say about this? Do you remember the part where they are analyzing an ant colony (or some such) which appears to have the attributes of intelligence, but is actually composed of many unintelligent bits?

From: [identity profile] novalis.livejournal.com


I think [livejournal.com profile] dougo is saying that generating some being which can pass the Turing Test is a valid way of passing the Turing Test. This implies that cp passes the Turing Test, if any program can pass. I think Turing's original definition is more narrow than this; and that narrowness prevents the test from contradiction our intuition.

From: [identity profile] dougo.livejournal.com


That isn't what I was saying, although that's an interesting tangent. I think what I was saying was that there is a different intelligence test that's analogous to the Turing Test but doesn't require a dialogue. Call it the Scopes Monkey Test, or something. But I don't have a good definition of how something passes this test, other than "the ID people think this looks like it was designed by an intelligence". Which is somewhat of a tautology. But it's hard to avoid tautologies when discussing unfalsifiable claims. (And natural selection is sort of a tautology itself: things that are better at surviving survive.)

From: [identity profile] dougo.livejournal.com


Also, I'm not sure copying counts as "generating". Sexual reproduction by intelligent beings isn't intelligence. Generations of reproduction and mutation from non-intelligent beings into intelligent beings is an intelligent process, but this doesn't require producing intelligent beings—producing an eye from a non-eye is similar.

From: [identity profile] dougo.livejournal.com


I only have Metamagical Themas. My copy of GEB was stolen from the UCB CSUA before I got through all the Achilles & Tortoise dialogs. Looks like Google Print has them both though. (Haven't figured out how to link to a specific page yet.)

From: [identity profile] mrmorse.livejournal.com


I'm willing to buy Strong AI. I'm also willing to that reality is computationally equivalent to a Turing machine. (See A New Kind of Science by Wolfram.) I still think there's a missing logical leap in declaring that therefore, evolution is equivalent to Intelligent Design.

I think the problem is that you still have to assume that intelligence is necessary for evolution. Put another way, if there is an Intelligent Designer, computational equivalence and Strong AI lead to the conclusion that the designer is reality itself. But I don't think it runs the other way.

Along the lines of GEB, electrical impulses in the brain are a necessary component of human intelligence. However, the impulses are themselves not a result of human intelligence.

Even if we conclude that reality is itself intelligent, evolution may be the functional equivalent of the electrical impulses. In fact, I think that model is more consistent with intelligent reality than the conclusion that evolution is the output of that intelligence.

From: [identity profile] dougo.livejournal.com


Along the lines of GEB, electrical impulses in the brain are a necessary component of human intelligence. However, the impulses are themselves not a result of human intelligence.

Did you mean to say "human intelligence is not a result of the impulses themselves"?

Even if we conclude that reality is itself intelligent, evolution may be the functional equivalent of the electrical impulses. In fact, I think that model is more consistent with intelligent reality than the conclusion that evolution is the output of that intelligence.

You're right, the process of evolution isn't enough by itself, it needs an environment to work in. And varying some parameters of this environment might change its results significantly, e.g. increasing or decreasing the amount of cosmic rays that cause mutations. Maybe the analogy is that evolution is the software while reality is the hardware. On the other hand, software is just a particular arrangement of electrons in the hardware, so I'm not sure there's any reason to draw that line of separation.

From: [identity profile] ketzl.livejournal.com


I like this line of thinking, but must disagree.

intelligent, because it produces things that can only be produced by intelligence

But that's the rub, of course. The process of evolution isn't obviously intelligent and seems to have produced new forms of life even so. There's lots of simpler examples of systems that seem to have no discernable special qualities but whose emergent properties are surprising and seem to speak of a higher intelligence, though-- fractals, various physical processes, the perfect spirals of a snail's shell, etc. Would you say those are intelligent too?

Here's my own version of your thinking: I believe in Strong Divinity. Any entity or system that produces results worthy of being worshipped is divine. So however, whoever or whatever is reponsible for our existence gets my prayers.

From: [identity profile] dougo.livejournal.com


Do you expect an answer to the prayers? Or is worshipping just a one-way (or zero-way) communication?
.