Intelligent Design - theory, examples, implications

[Buffy]

Interesting post, thank you. You're at Sunnyvale and your knowledge of intercepts... do you work for Aerospace, the AF, a contractor?

Thank you for acknowledging that neural networks are strong evidence for the ability of randomly evolving systems to produce intelligent behavior!

 

Cheers,

Buffy

[TeleMad]

Concerning the evolvable hardware I mentioned earlier, at the bottom is a link to one of the pages: it describes the core part of the system, an FPGA or field-programmable gate array. In simple terms, the logic that each gate performs can be determined not by someone having to manually reconfigure it "at the shop", but instead electronically "in the field". Conceptually, if you send a 1 as the configuration bit it might set itself up as an OR gate whereas a 0 might cause it to set itself up as an AND gate. Then a large array of these individual field-progammable gates are arranged into an array. Since their individual configutations can be set electronically using one or a few bits, the whole array's 'phenotype' can be specified using a bit string "genotype". So the researchers basically have the system create a small set of random genotypes and then physically instantiate them in hardware, testing each one for function. The ones with the best function are retained and then mutated to form the next "generation". This process of random mutation and selection of most fit continues and eventually a fully functioning circuit was produced, without the researchers specifying the intermediate steps to achieve the function (they were found by mutation + selection).

 

PS: The work mentioned at the site is from 1996 so I'm pretty sure they've advanced on this since then.

 

http://www.cogs.susx.ac.uk/users/adrianth/ices96/node2.html#SECTION00020000000000000000

[TeleMad]

What happened to the flagellum?

 

Here's a main theme I mentioned earlier that must be addressed: the type III secretory system which may be a simpler, functional precursor to the bacterial flagellum, just one that serves a different function.

 

” … molecular studies of proteins in the TTSS have revealed a surprising fact – the proteins of the TTSS are directly homologous to the proteins in the basal portion of the bacterial flagellum. As figure 2 (Heuck 1998) shows, these homologies extend to a cluster of closely-associated proteins found in both of these molecular "machines." On the basis of these homologies, McNab (McNab 1999) has argued that the flagellum itself should be regarded as a type III secretory system. Extending such studies with a detailed comparison of the proteins associated with both systems, Aizawa has seconded this suggestion, noting that the two systems "consist of homologous component proteins with common physico-chemical properties" (Aizawa 2001, 163). It is now clear, therefore, that a smaller subset of the full complement of proteins in the flagellum makes up the functional transmembrane portion of the TTSS.”

(http://www.millerandlevine.com/km/evol/design2/article.html)

[lindagarrette]

Thank you for acknowledging that neural networks are strong evidence for the ability of randomly evolving systems to produce intelligent behavior!

 

Cheers,

Buffy

There is much complexity but nothing random about neural networks. As yet, they have not produced anything much more intelligent than my washing machine. I read the Generation5 web site you provided which was excellent! Thanks for the link. Linda

[Buffy]

There is much complexity but nothing random about neural networks. As yet, they have not produced anything much more intelligent than my washing machine.

Actually, your washing machine uses Fuzzy Logic, which is usually not tied to a neural network (look up the master Lotfi Zadeh who invented it for more info). FL tends to use traditional algorithms. NN is in its infancy, and one thing has become clear about it is that it takes a lot of evolution to do more complex things. The main area of interest in NN turns out to be image recognition and there's lots of interesting research in this area thats starting to work rather well, although its still in the lab, not your appliances...yet! As to randomness, actually randomness is critical to the learning and goal seeking component of NN. If you don't use random values, NN's learn MUCH more slowly, because they're dependent on your (you being the God in the system) concepts about what might be the right next direction to go in...and you'll usually turn out to be wrong! NN produces unbelievably counterintuitive solutions...that work! Moreover, don't complain that computers produce only psuedo random numbers: good algorithms take input from atomic particle counters that produce "real" random numbers.

 

Cheers,

Buffy

[Lolic]

Thank you for acknowledging that neural networks are strong evidence for the ability of randomly evolving systems to produce intelligent behavior!

We went from interesting post to someone putting interesting words in my mouth. :)

I would recommend anyone read the following:

Forecasting with Neural Networks – A Review

Raymond J. Ballard

Texas A&M University-Commerce

Where he talks about neural networks needing to be trained etc. I don’t think there is a free lunch here, or free intelligence, or non-intelligence creating intelligence. More a case of fast computer cycles able to explore many possibilities and possibly coming up with a good fit…if properly trained.

 

“This learning ability of the neural network is accomplished through intense training of the network by providing it with numerous, reliable, and correct examples. It is essential that the examples provided to the neural network be accurate and complete since this is the only way to reach a desirable output result. As this learning process takes place, the many examples are combined to produce a "training set" of data. For each example presented to the network, there is a mathematical value representing both the input data and the preferred output data.”

 

“There are, however critics who point out the disadvantages of using neural networks as forecasting tools. First, the design of the neural network is a very complex procedure that still relies mostly on trial and error. This is due to the necessity to determine the appropriate input variables and the necessary level of training of the system. The training process is time-consuming, and must be continuously repeated to account for changes in values of the variables. Without the recurring retraining, the accuracy of the neural network will decline. In addition, because the neural network can only produce accurate results if provided with an accurate training set, it is necessary to use large volumes of precise examples in the training phase.”

[Lolic]

Surely you see how the logic is messed up now.

Response to above:

I basically agree with your point.. This is however a multifaceted issue. One of the issues is why is there so much bad information about Darwin and evolution in the text books when scientists know it’s not correct? This to “me” a main point of “Icons of Evolution” .

 

Buffy has posted some interesting information in regard to text book updating, but I do’t think it answers the questions that most interests me and the IDers. We know text books are updated rapidly with new editions. Keeps the knowledge updated and keeps the sales dollars rolling in. It’s a win for the publisher. But in this case, new editions are met with old data, old theories that people in the field know are not correct, yet they allow it to continue.

 

The Chinese scientist said more or less, in America you can knock the govt but not Darwin.

Dr. Wells gives other interesting examples of that issue in his book. One example is where European scientists aren’t’ afraid to discuss things that American scientist won’t. The example he gives has research funding as the factor.

 

This information makes me realize I was taught some bad information and bad science in school. Ok, mistakes can happen. We are all human, but 20-30 years later much of the same incorrect information is still being used.

 

For this point alone, I’m very glad the “Discovery Institute” has the courage to take on the issue. People like to belittle their motives, but motives…who cares. Everyone has motives…many of them selfish. At least they present good evidence of why things aren’t correct as they are and support the removal of incorrect information and bad theories.

[Lolic]

From a letter to the editor recently....

Intelligent Design incorrectly explained in Opinion piece

Letter

March 08, 2005

To the Editor:

Re: "Intelligent Design ... Brilliant!" Opinion, March 3

Perhaps I should give Josh Katz some latitude for his poorly-researched article -- after all, exactly what is taught in our children's schools will always be a touchy subject. I am reminded of a passage from an article by J. Dunphy, published in The Humanist: "I am convinced that the battle for humankind's future must be waged and won in the public school classroom by teachers who correctly perceive their role as the proselytisers of a new faith; a religion of humanity ... The classroom must and will become an arena of conflict between the old and the new -- the rotting corpse of Christianity, together with its adjacent evils and misery, and the new faith of humanism."

But high tensions notwithstanding, his mischaracterizing of Intelligent Design has such a blatant disregard for facts as to require a response. Intelligent Design is no more a religious belief than Evolution itself. It is a scientific theory which holds that many features of the universe and living things are best explained by an intelligent cause, and not as the result of an undirected, chance-based process. Intelligent Design theory is a testable hypothesis subscribed to by many educated scientists today; moreover, it was a belief in intelligent design, and thus a rational universe, that led to science as we know it. When Darwin posited his theory, little was known about the complexity of living organisms and many thinking people accepted his writings as fact. But much has happened since then. The advent of the microscope, the discovery of DNA, the "failure" of the fossil-record -- these all make it impossible to give Darwinian Evolution the free ride it once had. In the words of H.P. Yockey, from the Journal of Theoretical Biology: "One must conclude that, contrary to the established and current wisdom, a scenario describing the genesis of life on earth by chance and natural causes which can be accepted on the basis of fact and not faith, has not yet been written." But must we take it on faith?

Hannah Maxson '05

[Buffy]

Where he talks about neural networks needing to be trained etc. I don’t think there is a free lunch here, or free intelligence, or non-intelligence creating intelligence. More a case of fast computer cycles able to explore many possibilities and possibly coming up with a good fit…if properly trained.

 

“This learning ability of the neural network is accomplished through intense training of the network by providing it with numerous, reliable, and correct examples. It is essential that the examples provided to the neural network be accurate and complete since this is the only way to reach a desirable output result....

 

“There are, however critics who point out the disadvantages of using neural networks as forecasting tools. First, the design of the neural network is a very complex procedure that still relies mostly on trial and error. This is due to the necessity to determine the appropriate input variables and the necessary level of training of the system. The training process is time-consuming, and must be continuously repeated to account for changes in values of the variables. Without the recurring retraining, the accuracy of the neural network will decline. In addition, because the neural network can only produce accurate results if provided with an accurate training set, it is necessary to use large volumes of precise examples in the training phase.”

Well, I'm still waiting for refutation of the point. The excerpt here is exactly what I mentioned previously: the "hard" part comes in trying to short circuit the process, to allow training with the lowest amount of training. We're talking about doing pattern recognition on warhead dummies with only a month or so of "training" so yes, there's a lot of intelligence in that. The more fundamental research however has shown that time is indeed the key factor. The mention above about producing an "accurate training set" is necessary for the short-circuiting of the process. The statement that "Without the recurring retraining, the accuracy of the neural network will decline" is only valid if there is no feedback mechanism built in, which is very common in "use infrequently" systems like missle interceptors. In systems where there is constant input and constant learning--auto-pilot systems like the ones being used by the car manufacturers for automated roadways and the DARPA contest to get a self-driving vechicle to go across the Mojave desert--these systems *do* constantly improve. And these are with systems that are far simpler than even a shark brain, and have had training measured in months rather than millions of years.

 

As you can see, it can be misleading to jump to conclusions about comments out of context!

 

Automotonically,

Buffy

[Buffy]

Buffy has posted some interesting information in regard to text book updating, but I do’t think it answers the questions that most interests me and the IDers. We know text books are updated rapidly with new editions. Keeps the knowledge updated and keeps the sales dollars rolling in. It’s a win for the publisher. But in this case, new editions are met with old data, old theories that people in the field know are not correct, yet they allow it to continue.

Who is allowing whom? The publishers actually don't have much to gain by spending money to make these books accurate: it costs money, and any change is challenged, usually by the creationist/ID community for "promoting" Evolution, and leads to long, drawn out approval processes that may mean losing a multi-million dollar deal, while leaving it the same means they can simply say, "you didn't object before" which embarrases the critics into silence. This is wrong, but its how it works. The scientific community? You'll see just as many people complaining about this among the scientists, leading to some cynical opinions that the errors are left in to throw a bone to the anti-evolutionists so they can keep publishing these articles proving a conspiracy rivaled only by the coverup of UFOs. The scientific community has been trying to shame both the publishers and the school boards into doing the right thing, but they have little or no influence. Again, this is wrong, but its how it works. If you wish to see conspiracy theories in this, fine, but you still sound like the "aliens at Area 51" folks.

 

No more secrets,

Buffy

[James Putnam]

There you have it: "There is no known cause for intelligence," therefore it could not have been evolved or created by "mechanical forces." The two counter-arguments to this are:

1. There are in fact theories and evidence, which are explained away by ID as being "our inability to perceive the universe objectively because of photons deceiving us," or by simply ignoring the evidence.
   
2. Even where the evidence is not entirely conclusive, it is amusing to see the statement that this "must be the result of intelligent design" being refuted when new theories and more evidence ends up showing exactly how these "miracles of science" occur.
   

I actually applaud your approach James, because at least you try to come up with theories of your own to attempt to prove some theories of physics wrong, instead of simply changing the subject when faced with evidence. I think we'd all like to understand better what the flaws are that make the development of intelligence by mechanical means "impossible" or even "improbable." The explanations in Evolutionary theory of how intelligence evolved are well-developed and need not be represented here by me, but misusing statistics (which was my point in this thread), or endlessly asking "why?", or using the circular reasoning of "there must be a cause of intelligence (or unity of the universe), and since science does not provide a cause it is inadequate, therefore there must be a cause that did not come from mechanical causes or evolution and therefore there the universe was intelligently designed" does not refute these theories or the evidence that exists.

 

Cheers,

Buffy

 

Thank you buffy for your response. I suggest that the explanations in evolutionary theory of how intelligence evolved is only the acknowledgement that we recognize the evolution of intelligence. It is the reporting of results. It is the cataloguing of observations. Yes intelligence evolved. Explanation is not included. It is the case that some levels of intelligence won out in the survival contest against other levels of intelligence. This information tells us nothing about why intelligence began or why it advanced. (Here I mean advancement as in: How did DNA both begin and continue to progressively add to the quality of intelligence?) We know both occurred. The questions is why? If there is an explanation that shows a link between theoretical physics and the emergence of intelligence, I would like to hear it.

 

I argue that proof by association is not valid. For example, we define the universe as being mechanical. We see that parts of the universe assemble together into life and intelligence. I would challenge an attempt to use this apparent association to prove that mechanics gives rise to life and intelligence. I would argue that that which cannot predict the emergence of life and intelligence has been disproved by the fact of the emergence of life and intelligence. What do you think?

 

James

[James Putnam]

Excellent point James. Experience shows us that intelligence comes from a mind, so after observatin after observation, it's logical to deduce that other signs of intelligence came from a mind unless proven otherwise.

 

There certainly are other paths that lead to ID besides the teleological or design arguement. Care to branch into the Cosmological or Moral Law issues?

 

Hi Lolic,

 

Thank you. It is my position that life and intelligence did evolve by natural causes. However, I argue the real natural causes are not those defined by mechanical theory. What is your viewpoint? Please give me some indication as to what you mean by Cosmological or Moral Law issues? I do not mean this in a confrontational sense. It is only that there are very many viewpoints and, the English language doesn't always easily indicate to me what each person means. Cosmological sounds mechanical, but, perhaps you do not mean it that way. Moral Law sounds almost Biblical. Is this how you mean it? Are you referring to conclusions that you believe can be inferred from studying real natural properties of the universe? If you have past posts where you already explained your position, please point me to them.

 

James

[lindagarrette]

I would argue that that which cannot predict the emergence of life and intelligence has been disproved by the fact of the emergence of life and intelligence. What do you think?

 

James

I don't understand your point. Where did "intelligence" come from? There must have been something (or nothing) before it existed. If it evolved, it must have evolved from something else. Right? If there was an intelligence before there was an intelligence, it must be turtles all the way down.

[bumab]

I think he was stating the obvious argument.

 

If A produced B, then any theory of A must include the production of B, or it's already invalid.

[James Putnam]

I don't understand your point. Where did "intelligence" come from? There must have been something (or nothing) before it existed. If it evolved, it must have evolved from something else. Right? If there was an intelligence before there was an intelligence, it must be turtles all the way down.

 

Intelligence comes from intelligence. The process of evolution gives the impression that intelligence grew from lower states to higher states. However, I think the full potential for all possible intelligence had to have been provided for since the beginning of the universe. This is analogous to arguing that the full potential for the mechanical properties of the universe has existed since the beginning of the universe. My position is that a mechanical interpretation of the nature of the universe is inadequate to describe or provide for either the realization of intelligence or its potential. I see no time when mechanics could be used to represent any level of intelligence.

 

The turtles represent the fact that we cannot acount for original cause by scientific means. Original cause, by definition, must itself be uncaused. In a mechanical interpretation this problem can be expressed as: The first cause of motion must itself be unmoved. We cannot give alternative answers for these questions, by scientific means, without resorting to those 'turtles'. What good would it serve to propose an infinite series of causes for causes? Or, what is the benefit of trying to prove that 'nothing' was the original cause? Are those answers superior to allowing for the an uncaused intelligence?

 

Uncaused simply means we cannot find a cause for intelligence within the universe and, we cannot scientifically determine a cause for intelligence before the universe. Any effort to propose a cause must show how the cause predicts the emergence of intelligence. I do not believe that anyone can do this. Even if a cause could be logically proposed, then we are faced with the question of: What caused that cause? Why would this be an improvement over the argument that intelligence, insofar as human logic is concerned, is uncaused?

 

My answer to your question is that intelligence evolved from intelligence. Just as the properties that caused matter to evolve from a disassociated state into a state of association. The same argument would apply for all properties. Intelligence existed at its full potential from the beginning of the universe. In the physical sense, it was dissasociated. It evolved into a state of complex association leading to its realization in complex form.

 

It is my position there are fundamental properties of intelligence analogous to the proposed mechanical properties. In either case, the fundamental properties evolve toward complex results. For me, the mechanical properties are clearly artificial and cannot be substantiated by empirical knowledge. Intelligence is clearly a real empirical property of the universe. I see mechanics as being wholly incapable of leading to intelligence. I see fundamental intelligent properties as a reasonable idea for the means by which complex intelligence can evolve.

 

James

[TeleMad]

Yes intelligence evolved. Explanation is not included. It is the case that some levels of intelligence won out in the survival contest against other levels of intelligence. This information tells us nothing about why intelligence began or why it advanced. (Here I mean advancement as in: How did DNA both begin and continue to progressively add to the quality of intelligence?) We know both occurred. The questions is why?

 

No, the question for evolution is NOT why. WHY is a metaphysical, philosophical concept. Evolution is science: it deals with HOW.

[TeleMad]

I would argue that that which cannot predict the emergence of life and intelligence has been disproved by the fact of the emergence of life and intelligence. What do you think?

 

Let's see, dolphins can't predict the emergence of life and intelligence, therefore dolphins have been disproved. Let's see, Einstein's theory of special relativity can't predict the emergence of life and intelligence, therefore special relativiy has been disproved.

 

Okay, enough of looking at the statement in isolation.

 

Your larger argument is that intelligence must have existed before the universe because evolution cann't predict the emergence of intelligence. Well, evolution cannot predict the emergence of eyes either, so eyes must have existed before the universe. Hey, you know what, evolution can't predict the emergence of humans either, so I guess humans existed before the universe.