By request, this has been moved from the "Should preaching be banned [or whatever]" thread in which I wrote the following:
Speaking of constrained optimization, I found myself watching "Faith Under Fire about a month back (Don't ask why that show or even that channel, UPN.). So they were arguing about intelligent design and evolution, and the evolution proponent rightly asks, "If your god is so brilliant, why did he do such a bad job designing so many of the things we find in nature?" He launched into a polemic about how the human eye could be made better.
Now the YEC/ID guy didn't blink. When his adversary was finished, he calmly asked, "Have you heard of something called constrained optimization? Engineers have recently discovered this and think it might explain why the human eye is the way it is."
Now Lee Strobel (third-rate evangelical apologist and host of Faith Under Fire) was duly impressed with this answer and wrapped up the program without giving the evolutionist a chance to respond. The irony, which had so many levels in this case, nearly caused me to spontaneously hurl my television set through the wall.
I will give the reasons I found the response ironic below, but first let me present the question.
Assuming that the xtian god created the universe, did this god solve a constrained optimization problem? Does optimization theory show that the eye is "well-designed" or better "optimally designed" given some constraints?
Ok now the ironies.
First, on the most superficial level, constrained optimization was not discovered recently by engineers. Anyone who has taken a multivariate calculus course has almost certainly been introduced these types of problems. In fact, Joseph-Louis Lagrange developed one of the most famous and common methods of optimization with constraints called the Method of Langrange Multipliers around 1755 (He was 19-years-old!).
That Lee Strobel was impressed by this argument did more to reveal his fascination with polysyllabic phrases than the strength of the argument.
On the next deeper level, I agree with DrScribbles that if the structure of the human eye is a product of a constrained optimum, then this would support the theory of evolution more than the idea of special creation. Optimizing an objective function subject to some natural constraints could model natural selection quite well. Species adapt to their environments to achieve the best chances of reproducing. This process cannot happen just anyway. For example, there are constraints from time, other species, and laws governing how organic compounds can change or how much energy an species needs to stay alive. In fact in economics, I've encountered genetic algorithms that have been used to find optima of stochastic objective functions (generally not well behaved ones to be sure) and were originally designed to model natural selection/evolution.
Finally, and most ironic in my opinion, is that the implication of claiming your god had to solve a constrained optimization problem when designing the world is that it necessarily implies he faces constraints! Is there really any reason to think that the designer of reality faced physical constraints when designing the human eye? Imagine a neurophysicist/mechanical engineer meeting the creator and heckling him about it?
Engineer: "So uh god, the human eye is quite a puzzle. Why did you create it in such a shoddy fashion. I could give you a few pointers."
God: "Settle down, hot shot. I did the best I could. You see I was running low on photocells, and there's no way to cram the optic neurons in there without making the human forehead bigger. And how do you suppose I do that?"
Pretty silly huh? Of course, it doesn't make sense to think of a god with the awesome power of creatio ex nihilo as being constrained by the reality he makes. He makes the rules! If not, then it would imply that he lives within some greater reality with rules that bind or constrain his powers.
So I think the ID apologists, "constrained optimization" excuse was little more than a quick bit of obfuscation. He knew it would give the impression that he had some impressive mathematical theory to back him up and took the risk that his philosopher opponent either wouldn't recognize the phrase or wouldn't have time to formulate a clear response.
Originally posted by telerionI went to school just down the road from you and spent two years as a researcher at NASA Langley working on developing a particular algorithm of this sort. It's good to see another optimization guy here - we're few and far between. You may be the first that I've encountered in the wild, not in an academic or professional setting.
I've encountered genetic algorithms that have been used to find optima of stochastic objective functions (generally not well behaved ones to be sure) and were originally designed to model natural selection/evolution.
Your points are valid. I'm guessing that was just written into the script as something that people wouldn't even try to understand but would nonetheless be impressed with. Maybe the screenwriter got mixed up about who was supposed to make the point with that argument, since it's obviously in favor of a natural, undesigned product.
Dr. S
P.S. I think I may have misunderstood the nature of the program. Is it a live debate? At first I thought it was some kind of drama. I wonder if some of us RHPers could try out for it if it is a debate. I guarantee a ratings boost. Could you just imagine a televised version of this forum? What a show it would be! We could have alerts, bannings, recs, and everything else that makes this forum great.
Originally posted by DoctorScribblesWell met. I think we could get televised. We have the personalities here at RHP. All we need is an affable, but incredibly biased, host. It would be public access at first, but maybe someday, with luck and determination, we could reach the apex of biased opinion television and get a half-hour spot on FoxNews.
I went to school just down the road from you and spent two years as a researcher at NASA Langley working on developing a particular algorithm of this sort. It's good to see another optimization guy here - we're few and far between. You may be the first that I've encountered in the wild, not in an academic or professional setting.
Your points ...[text shortened]... uld be! We could have alerts, bannings, recs, and everything else that makes this forum great.
Originally posted by DoctorScribblesI can just picture no1marauder throwing a chair at someone...
P.S. I think I may have misunderstood the nature of the program. Is it a live debate? At first I thought it was some kind of drama. I wonder if some of us RHPers could try out for it if it is a debate. I guarantee a ratings boost. Could you just imagine a televised version of this forum? What a show it would be! We could have alerts, bannings, recs, and everything else that makes this forum great.
Originally posted by NemesioSEASON ONE
I can just picture no1marauder throwing a chair at someone...
Episode 1: no1marauder and pcaspian debate gay marriage laws, with side commentary from RBHILL and darvlay.
Episode 2: STANG and chancremechanic discuss U.S. foreign policy
Episode 3: Live from Vatican City, Nemesio and ivanhoe explore the finer points of Leviticus, moderated by the pope.
Episode 4: Live from Jerusalem, Nemesio and ivanhoe explore the finer points of Leviticus, moderated by Feivel.
Episode 5: Bbarr and the good Doctor paint each other's philosophies as extreme caricatures, and then defend such depictions using impeccable rigor and unprecedented rhetoric.
Episode 6: RWilis gives a monologue on the tenets of "weak atheism". Nobody shows up to rebut him since we have all either accepted it or given up all hope of convincing him of its non-existence.
Episode 7: A all-versus-one battle featuring kirksey against all other RHP Christians who say he is going to hell.
Episode 8: The season finale, an all out battle royale between the full Wolfpack and the reunited SSS, featuring pradtf coming out of retirement, referreed by poor Russ.
Dr. S
Originally posted by DoctorScribblesLMFAO!
SEASON ONE
Episode 1: no1marauder and pcaspian debate gay marriage laws, with side commentary from RBHILL and darvlay.
Episode 2: STANG and chancremechanic discuss U.S. foreign policy
Episode 3: Live from Vatican City, Nemesio and ivanhoe explore the finer points of Leviticus, moderated by the pope.
Episode 4: Live from Jerusalem, Nemes ...[text shortened]... he reunited SSS, featuring pradtf coming out of retirement, referreed by poor Russ.
Dr. S
I thought all optimization problems had constraints! 😀
Originally posted by telerion
On the next deeper level, I agree with DrScribbles that if the structure of the human eye is a product of a constrained optimum, then this would support the theory of evolution more than the idea of special creation. Optimizing an objective function subject to some natural constraints could model natural selection quite well. Species adapt to their environments to achieve the best chances of reproducing. This process cannot happen just anyway. For example, there are constraints from time, other species, and laws governing how organic compounds can change or how much energy an species needs to stay alive. In fact in economics, I've encountered genetic algorithms that have been used to find optima of stochastic objective functions (generally not well behaved ones to be sure) and were originally designed to model natural selection/evolution.
I'm not sure if I'm going OT here, but IMO one of the key problems evolution faces is with questions of irreducible complexity. In terms of optimisation, that's like a solution consisting of a spike in the middle of a "valley", so that regular "hill-climbing" algorithms would never find the solution. Even if you assume mutation in your genetic algorithm, it is still extremely unlikely that a random mutation would result in precisely the right solution.
Finally, and most ironic in my opinion, is that the implication of claiming your god had to solve a constrained optimization problem when designing the world is that it necessarily implies he faces constraints! Is there really any reason to think that the designer of reality faced physical constraints when designing the human eye? Imagine a neurophysicist/mechanical engineer meeting the creator and heckling him about it?
This is like the old question, "If God is all-powerful, can he create a rock large enough that he cannot lift it?"
Clearly, (and I think Doc pointed this out sometime) God could be constrained by logical laws such as non-contradiction. Further, there may be non-physical constraints (e.g. beauty, spiritual constraints) in the design of the human eye.
Originally posted by lucifershammerThe problem here is that non-contradiction and non-physical restraints are human perceptions. If god did exist, then there is nothing to say that he could not perceive in an entirely different way, unfathomable by us. If god was required to operate within the boundaries of human perception when creating the universe, surely this sets finite levels upon him. The only constraints we can offer are those we suffer from ourselves and I doubt god would suffer from these as well. Would the perception of a human not lead to the perciever thinking as a human, acting as a human, being a human?
Clearly, (and I think Doc pointed this out sometime) God could be constrained by logical laws such as non-contradiction. Further, there may be non-physical constraints (e.g. beauty, spiritual constraints) in the design of the human eye.
Originally posted by StarrmanFirst, the non-contradiction - I don't think it's a constraint due to human perception. A.(not A) will be false even in alien logic. It seems to be a more fundamental constraint, perhaps by definition.
The problem here is that non-contradiction and non-physical restraints are human perceptions. If god did exist, then there is nothing to say that he could not perceive in an entirely different way, unfathomable by us. If god was required to operate within the boundaries of human perception when creating the universe, surely this sets finite levels upon hi ...[text shortened]... tion of a human not lead to the perciever thinking as a human, acting as a human, being a human?
Second, the non-physical constraints - clearly, we humans do not know what they are (at the moment) and, I guess, we won't in our lifetimes. Hence, these non-physical constraints might be unfathomable to us.
Originally posted by lucifershammer
I thought all optimization problems had constraints!
Well, in some sense, solutions to these problems are constrained by their own objective functions. For example, maximize, by choice of x, the function 3x - x^2. Now your answer can't be 15 because there is no x for which 3x- x^2=15 is true.
However, this is not what is meant by constrained here. Generally speaking, constraints restrict certain values of the objective function from being optima. The argmax of the problem above (an example of unconstrained optimization) is 3/2, which gives a value of 9/4 for the max of the objective function. Suppose however that we introduce a constraint. Say x must be less than or equal to 1. Now we have corner solution at x=1, which yields a constrained maximum of 2.
So if we are consistent with mathematical terminology, then no, not all optimization problems have constraints.
Originally posted by lucifershammer
I'm not sure if I'm going OT here, but IMO one of the key problems evolution faces is with questions of irreducible complexity. In terms of optimisation, that's like a solution consisting of a spike in the middle of a "valley", so that regular "hill-climbing" algorithms would never find the solution. Even if you assume mutation in your genetic algorithm, it is still extremely unlikely that a random mutation would result in precisely the right solution.
I think this is perfectly on topic. Behe's idea of irreducible complexity has hit it big in creationist and ID circles, but as far as I know it is much discredited in the scientific community. I would assume this is because at its crux, irreducible complexity is not science. It is "goddunit" on speed. It says, "At this point, I can no longer see how these simple systems could have arisen naturally. Therefore it must have been created supernaturally." This boils down to "I give up," a response that is the perhaps the antithesis of science. Maybe some one with more specific expertise in evolutionary theory and ID could offer some insight here (prn perhaps?).
Ok now your analogy to optimization. Certainly there are problems that simple "hill-climbers" will always mess up. They will invariably get stuck on some local extremum and miss out on the big one. In my original post, I mentioned stochastic objective functions as one example of where simple techniques can go wrong. Nevertheless, we don't throw our hands up and say that these problem can never be solved except by a god. We develop better techniques for finding the extrema (e.g. pattern search techniques).
This analogy however does not apply to an all-knowing, all-powerful god. The reason we, humans, could have trouble finding the maximum in your example is because we don't know the spike is there. If we knew that the spike was there, we could find its max by sufficiently restricting the region of our search. This is why unless we have sufficient information about the objective function at every point, we cannot be certain that any located extrema are global and not local. This dilemma could not apply to an all-knowing god since it would know the exact nature of every possible objective function.
Originally posted by lucifershammer
This is like the old question, "If God is all-powerful, can he create a rock large enough that he cannot lift it?"
Clearly, (and I think Doc pointed this out sometime) God could be constrained by logical laws such as non-contradiction. Further, there may be non-physical constraints (e.g. beauty, spiritual constraints) in the design of the human eye.
If engineers have pointed out how the eye is poorly designed, then this implies that they have knowledge of a design that would be better. In fact, some have pointed to eyes of certain other species that do not suffer the same problems. If this is the case, then it cannot be that these particular designs some how run against laws concerning what is logically possible.
As for non-physical constraints, these amount to "God made the human eyes inefficiently because that's what he wanted to do." Sort of an unknown cause argument. I cannot really argue against it. Maybe that's the case. Nevertheless, I think non-physical considerations get away from ID, which makes its case on the perceived efficiency of design for the physical world, not on notions of divine aestheticism or efficacy in the "spirit world." ID says look at how well suited the human eye is for its function, gathering and channeling information about light for the brain. We see things around us that we have made for a specific function (bridges, buildings, watches). These were constructed by intelligent designers (engineers, architects,watchmakers). Eyes then were designed by some sort of ultra-intelligent "eye maker."
So could we include non-physical constraints? Sure. Mathematically we could write a model with specific "beauty" constraints so that we find the current structure of the human eye to be optimal. Unfortunately, these exogenous "beauty" constraints would be driving our results. This would amount to little more than a mathematical example of "Because he wanted to."
So far I think ID looks like this:
How did life arise in the universe?
Goddunit.
Why did it God do it this way?
Because he wanted to.
By the way, evolution through natural selection endogenizes the constraints since the constraints arise as results of the environment.
This is kinda interesting. Hopefully we attract some more people.