|
Author
|
Topic: The Characterization of Intelligent Causation
|
Danpech
Member
Member # 163
|
posted 11. June 2007 14:38
I cannot help but observe that I am an intelligent causal agent. Intelligent causation is within my power, as is my ability to observe its nature or structure.
But, as is hinted at http://www.iscid.org/boards/ubb-get_topic-f-24-t-000004
there is a fundamental distinction between myself as intelligent causal agent and a Ground of Being as intelligent causal agent.
IP: Logged
|
|
Daniel Smith
Member
Member # 3004
|
posted 11. June 2007 19:39
IF:
quote:
Of the list that you provided which do you think I should start with?
I'd recommend the 2 books by Michael Denton to start. I haven't read all the books on that list, but I have read several and I think Denton is the most balanced and unbiased of those I've read.
IP: Logged
|
|
LifeEngineer
Member
Member # 3446
|
posted 12. June 2007 08:32
INTELLIGENT CAUSATION AND LOGIC MACHINES
If you decide to analyze intelligent causation from a hard science mathematical perspective, you inevitably end up modeling and analyzing intelligent causation in terms of information processing, logic machines and computers.
When you model and analyze intelligent behavior in terms of logic machines or computers you are dealing with systems with two types of inputs- 1) environmental stimuli and 2) computer programs or processing algorithms. Modeled and analyzed from this perspective, systems have one type of output labeled environmental responses.
For those who have difficulty getting beyond naïve realisms, it might be worth noting at this time that inputs, outputs, and processing algorithms used in scientific analysis are all abstractions defined by scientists and, if you like to think of it that way, created in the minds of scientists for communication between scientists. In the old days, the abstract inputs, outputs, and programs to computers might have been abstract representations of holes in punch cards. In studying neurons the inputs and outputs might be neuronal impulses and chemical factors impacting these impulses. The point, or side issue here, is that if you wish to get a hard science understanding of the relationship between intelligence and materialism, start by learning to understand the relationship between abstract scientific variables and functions and the phenomena these variables and functions model.
If you are going to scientifically analyze intelligent causation in terms of logic machines and information processing, the first step is to identify some form of behavior that 1) clearly involves intelligent causation and 2) is relatively easy to formally observe, model and experiment with. You could probably write an entire book on the subject of “What is an appropriate starting point for observing and measuring intelligent causation?”
There are at least four major problem involved in identifying an appropriate subject for the study of intelligent behavior. First, behaviors that are clearly intelligent, like writing a Nobel Prize winning novel, are too complex to study. Second, it is not certain that very simple forms of behavior are in fact intelligent behavior. Third, it is not always clear that the intelligent behavior being modeled or simulated is the same as the intelligent behavior being observed. Fourth, there are a lot of misleading and inappropriate expectations about what intelligent causation ‘should be like’ (as opposed to the traits identified by actually observing and studying intelligent causation).
It turns out, with the benefit of hindsight, that there are actually all sorts of examples of intelligent causation that are simple and easy to study. Feedback loops, or at least programmable feedback loops turn out to be the basic or elementary form of intelligent causation. But you only discover that after you have performed formal analysis of more complex forms of intelligent causation.
As I have discussed elsewhere, I started by studying information processing in neurons. If nervous systems or at least human nervous systems are recognized as capable of producing intelligent behavior or intelligent causation, then by the principles of scientific reductionism, individual neurons must also be capable of intelligent causation. I later studied human decision making which clearly involves intelligent causation and which, surprisingly is much easier to observe and analyze. As I recently became aware, gene regulatory processes provide an excellent opportunity to study intelligent causation as it impacts evolutionary change.
Again, in selecting a form of intelligent causation to study formally using the information processing approach, you need to know 1) that the behavior being studied involves intelligent causation, and 2)that it is practical to model input, output, and processing algorithms.
Once you have an appropriate subject for study, and once you can begin to model the input, output and processing algorithms associated with the behavior, you can begin to observe the characteristics of intelligent causation. Note that in this analysis, as in all real scientific analysis, the properties and characteristics of intelligent causation are identified based on observing the behavior and by analyzing models of the behavior. The more typical soft science approach is to start with a set of preconceived notions or beliefs, and then to attempt to force the observed data to fit the preconceived beliefs.
If you search the literature, or if you talk to experts like aiguy or IF (or ID experts) or if you talk to the academics at institutions like Santa Fe Institute or Duke or Harvard, you are not likely to run into any serious discussion of the topics outlined above. Lots of people seem to place a lot of importance on the literature and on the opinions of academic experts, but when you get down to the nitty-gritty of real scientific analysis, you will not find directly relevant materials and you won’t find any meaningful discussion of relevant topics.
IP: Logged
|
|
LifeEngineer
Member
Member # 3446
|
posted 12. June 2007 09:07
IF,
Quote: How do you suggest hard science be used to figure out the events in the deep past? In what way are todays methods to figure out events in the deep past wrong?
What a strange question. I can only assume that you are confusing historical or descriptive soft science explanations with hard science scientific explanations.
Soft science descriptive explanations are something like a series of snapshots made into what appears to be a continuous movie. Based on the fossil record, for example, someone might create a description of the changes from land mammal to whale or the apparently gradual change in the neck length of the evolving giraffe. While it is possible to assign names to the change processes involved in the observed transformations, such names do not constitute scientific explanations of the change processes.
Real science or hard science explanations of past events always start by developing testable predictive theories of change processes. By definitions, such theories must be currently testable. Once valid predictive theories have been developed, hard science methodologies might be used to project or extrapolate known processes backwards in time. These backward predictions might then be tested against historical or fossil data to determine the validity of the predictive theories.
Except possibly among soft science practitioners, there should never be any confusion about how science evaluates events in the past.
IP: Logged
|
|
Melvin H. Fox
Member
Member # 1684
|
posted 12. June 2007 10:09
LifeEngineer,
You wrote:
quote:
Once valid predictive theories have been developed, hard science methodologies might be used to project or extrapolate known processes backwards in time. These backward predictions might then be tested against historical or fossil data to determine the validity of the predictive theories.
This is a problem for deep past extrapolation. Both historical and fossil data are indirect data; the further back in time one goes, then the more questionable the data becomes. For example, in order to use radio-carbon dating to determine the HARD age of a fossil, we need to KNOW how much carbon was in the animal [or at least in the atmosphere] at the time that animal died. If we do not KNOW this information, then any age determined by radio-carbon dating is a SOFT age of the fossil and therefore that data is questionable.
Specifically now, in the testing process how should hard-science overcome the absence of direct data?
Also, your terminology, “known processes” sounds very uniformitarian when used in the context of the deep past. Is hard-science bound to the assumptions of uniformitarianism [the principle that the same processes that shape the universe occurred in the past as they do now, and that the same laws of physics apply in all parts of the knowable universe – Wikipedia]?
-Mel
IP: Logged
|
|
LifeEngineer
Member
Member # 3446
|
posted 13. June 2007 08:15
Mel,
You bring up a number of key issues regarding scientific analysis and ‘deep past extrapolation’.
First, deep past extrapolations are based on the use of theories and assumptions. As a general rule or principle, real science assigns far more credibility to theories and assumptions that are independently and openly testable.
Second, the validity of all deep past extrapolations are subject to the limits of Goodman’s paradox. In simple terms, this paradox points out that we can never actually know what happened in the unobservable past. An hypothesis asserting that the world was created last Wednesday or 6000 years ago (with the appearance of greater age) is just as compatible with the evidence as theories that suggest the earth is much older.
Third, real science or hard science extrapolations are based on the ‘current best set of logically consistent testable assumptions and theories’. Real science can not and does not attempt to produce perfect extrapolations. Hard science only attempts to recognize the current best extrapolation. In hard science, it is not sufficient or even relevant to point out that the current accepted extrapolation based on a currently accepted set of theories and assumptions is imperfect. It is a major soft science misconception to believe that you can 1)legitimately reject a set of theories and assumptions without offering a testable predictive alternative or 2)legitimately continue to accept a set of theories incompatible with the evidence because you don’t like the alternatives presented. Biologists and geneticists attempt to hand on to their discredited theories because they don’t like the teleological alternatives. Some YEC supporters attempt to reject all dating models because the data appears imperfect. Some people reject global warming theories because in their view the available data is imperfect. Again, science does not produce perfect extrapolations, but ‘current best meeting scientific standards’ extrapolations.
The uniformitarian assumption or the permanent and universal causal relationship assumption is an essential requirement for all hard science theories. Failure to adhere to this assumption or requirement results in theories that are trivial and not openly testable. However, this requirement or assumption does not preclude recognition of discontinuous or dynamic processes and operations. Most immediately relevant, intelligent causation, such as that associated with evolution, is always discontinuous. Evolution, as is well documented, does not occur slowly and steadily, but it rapid spurt with sometimes very long periods of stasis.
Even in the field of physics, the uniformitarian assumption is not as simplistic as it might at first appear. For example, the expanding universe evidence supporting the straight line extrapolation that produces so called big band theory, may also be compatible with a cyclical expansion/contraction model and theory. It is at least theoretically possible that some of the processes used in dating materials might be variable or controllable rather than constant. It may be confusing to many, but science does require the use of the uniformitarian or permanent and universal assumption. This required assumption, does not, however, preclude recognition of discontinuous processes.
To repeat the point I made to IF, the hard science concepts and principles and standards used in the analysis of past events appear to be well established. The problem is not the lack of established principles, but the fact that the principles are not well understood and often not properly applied.
IP: Logged
|
|
IF
Member
Member # 1904
|
posted 13. June 2007 08:32
LE, quote:
...{Intelligent Causation}... If you search the literature, or if you talk to experts like aiguy or IF (or ID experts) or if you talk to the academics at institutions like Santa Fe Institute or Duke or Harvard, you are not likely to run into any serious discussion of the topics outlined above. Lots of people seem to place a lot of importance on the literature and on the opinions of academic experts, but when you get down to the nitty-gritty of real scientific analysis, you will not find directly relevant materials and you won’t find any meaningful discussion of relevant topics.
How about in the philosophical magisterium? Or could it be that it is too big and vague a question for our current hard won, important but "trivial" knowledge synoptically speaking! Maybe, the current process is slowly (much too slowly for you and many others) gathering data so that a paradigm shift your way is inevitable! The main thing is that the search for any and all trivial answers continue in the hope that as they accumulate a new perspective on those big vague kinds of questions result.
IP: Logged
|
|
LifeEngineer
Member
Member # 3446
|
posted 13. June 2007 16:13
OBSERVING INTELLIGENT CAUSATION
It seems kind of strange, but for all the interest in and discussion of the subject of intelligent causation, there is surprisingly little evidence of scientists actually attempting to observe and model intelligent causation as it occurs in nature. Given the apparent lack of good data and good observations, I am going to start here by discussing some of the more obvious features of intelligent causation.
PRODUCING CORRECT OR GOAL COMPATIBLE RESPONSES
A useful starting point in observing intelligent causation is to consider the output of intelligent systems that are operating successfully and effectively. If we look at the output from properly functioning neurons or properly functioning gene regulatory processes or from properly functioning human decision making, we quickly discover that properly functioning intelligent systems routinely generate the proper or appropriate or goal directed output.
CHANGING BEHAVIOR
A second readily observable or obvious feature of the behavior of intelligent systems is change. Given the same or very similar inputs at different points in time, an intelligent system will produce different outputs (But, again, in a properly functioning intelligent system the changed output will still be correct, appropriate or goal-compatible.) There are lots of different terms or labels assigned to these changes in behavior including learning, adaptive change, and even evolution.
One of the major technical challenges associated with the analysis of intelligent behavior is the question of how to address or model changes in behavior. The problem or issue of modeling or simulating complex adaptive intelligent behavior is one familiar to anyone who has tried to write programs to simulate such behavior. As most people who have worked with this type of issue know, it is very easy, almost trivial to simulate intelligent processing under a very specific set of conditions. There is generally a specific identifiable stimulus that can be described as producing a specific response as the result of a specific processing algorithm.
However, attempts to simulate intelligent causation using such simple models inevitably fail badly under realistic conditions. In order to simulate even seemingly simple forms of intelligent causation under realistic environmental conditions, it is necessary to develop complex feedback and control processes. These adaptive feedback and control processes, if they are even to begin simulating real capabilities of real world intelligent systems, will involve very complex input and very complex processing algorithms.
It may be easy to show that a single neuron produces an output response in reaction to an edge in a particular part of the visual field, but simulating the processing associated with such a simple behavior can be extremely complex. Similarly, gene regulatory process may produce simple ‘produce protein or don’t produce protein’ responses, but the complexity of such intelligent processing is probably beyond the current capabilities of the best human systems designer. Human decisions like ‘pick the best brand of peanut butter’ may appear simple, but simulating such complex behavior is a major challenge for scientists.
To repeat, if we go out and observe even apparently simple forms of intelligent causation in the real world, we find that such intelligent causation is highly dynamic and producing realistic models and simulations of the dynamic and adaptive aspects of intelligent causation is extremely difficult. [Except, of course, to the soft science practitioners who can wave a hand and make all the real world complexity disappear.]
IP: Logged
|
|
IF
Member
Member # 1904
|
posted 13. June 2007 16:36
LE,
quote:
To repeat, if we go out and observe even apparently simple forms of intelligent causation in the real world, we find that such intelligent causation is highly dynamic and producing realistic models and simulations of the dynamic and adaptive aspects of intelligent causation is extremely difficult. [Except, of course, to the soft science practitioners who can wave a hand and make all the real world complexity disappear.]
Read "GEB" by Hofstadter, I think you will enjoy it!
IP: Logged
|
|
IF
Member
Member # 1904
|
posted 14. June 2007 07:54
LE,
quote:
Real science or hard science explanations of past events always start by developing testable predictive theories of change processes. By definitions, such theories must be currently testable. Once valid predictive theories have been developed, hard science methodologies might be used to project or extrapolate known processes backwards in time. These backward predictions might then be tested against historical or fossil data to determine the validity of the predictive theories.
Why don't you think that the current scientific methods (as applied to the fossil record) follow your template? Also, the study and understanding of DNA should provide the kind of hard science experimental tests and exciting insights, with regards to both ontogeny and phylogeny, that you advocate, no?
IP: Logged
|
|
LifeEngineer
Member
Member # 3446
|
posted 14. June 2007 08:41
IF,
QED:Read "GEB" by Hofstadter, I think you will enjoy it!
Thanks, I’ll give it a look. However, based on the reviews, I doubt if it would add anything to the discussion here.
The relationship between philosophy of science literature (and the relationship of any scientific literature) to actual scientific problem solving is an interesting one. To begin with, there are clearly two broad classes of readers of the literature. The first and by far the largest group of readers are those who read this literature as a source of authoritative information. Many of this type of reader appears to operate on the misconception that if they have read a piece of literature and if they can reference it then they actually understand what the author is saying.
The second type of reader, people who actually attempt to use materials contained in the literature to solve scientific problems, have a very different outlook on the literature. To begin with, scientists actually trying to solve problems quickly learn that 99% plus of the literature is irrelevant and useless drivel. Second, hard science analysts realize that useful literature is of two basic types. First, one useful type of literature is literature that discusses specific hard science solutions to real hard science problems. Any new problem addressed by any scientists will involve many different components, and it is likely that other scientists will have already resolved the issues surrounding some components. A productive scientists builds on the actual useful solutions produced by other scientists.
The second useful type of literature involves speculation on potential solutions for unsolved components of the larger unsolved problems. Issues like the mind/body problem, the ‘what is intelligence?’ problem and the problem of formulating predictive theories in the life sciences have been around unsolved (or apparently unsolved) for a very long time. Philosophers have produced a lot of interesting speculation (and a lot of useless speculation) on these topics. [Note that when you have a complex multiple component problem, it is generally very easy to come with solutions that work for any one component. The challenge is not finding solutions for each component, but in finding a set of logically compatible solutions that can be used to solve the bigger problem being addressed.]
The real problem solving scientist generally does not view the literature as a source of solutions, but rather as a source of ideas for formulating solutions. For the most part, the scientist looks at the literature as a record of failed efforts to solve the problem he or she is currently addressing.
Once a scientist finds a solution to a previously unsolved problem, he may go back to the literature and look at previous efforts to address the issue. Once you know a solution to a problem, it is relatively easy to recognize which parts earlier scientists and philosophers got right and which parts they got wrong.
The two main points to make here are that 1)most people seem to use references to the literature as magical incantations without really understanding the literature, and 2) even the best parts of the literature report failed efforts to address unsolved problems.
IP: Logged
|
|
LifeEngineer
Member
Member # 3446
|
posted 14. June 2007 09:50
IF,
Quote: Why don't you think that the current scientific methods (as applied to the fossil record) follow your template? Also, the study and understanding of DNA should provide the kind of hard science experimental tests and exciting insights, with regards to both ontogeny and phylogeny, that you advocate, no?
The analysis of the fossil record is simply descriptive. There are no testable scientific theories addressing the change process.
I assume that much of the geological science involved in dating materials is based largely on hard science methodologies. I suspect that a part of the confusion regarding dating methodologies arises because hard science geology gets mixed in with soft science evolutionary biology. The other side of the problem, however, is soft science concept used by YECs, as well as Darwinists, that any scientific theories can be subjectively rejected if they can be shown to appear to be imperfect. The hard science standard should be ‘one failure falsifies if a replacement theory is available’. Note that academic sciences create this ‘subjective rejection of theories that don’t explain everything’ when they refused to recognize teleological theories.
A very good argument can be made that genetics is the most absurd and logically unsound of pseudo-science that ever existed. It is certainly far worse than alchemy and astrology combined. Genetics is a based on the continued use of a set of assumptions that are known to be false and known to produce misleading interpretations of the data. Among the central assumptions of genetics known to be false are 1) the assumption that a genotype phenotype mapping can or does exist (there is ample evidence that phenotype of far more complex than genotypes), 2) the assumption that genotype changes can cause phenotype changes-the genetic determinism assumption (again the evidence clearly shows that genetic data contains only a tiny fraction of the information required to produce phenotypes), 3) the assumption that genetic changes and evolutionary changes can be explained by non-intelligent searches, and 4) the assumption that there are no intelligent goal directed control processes involved in genetic change.
The only really interesting thing about the science of genetics is that it continues to be supported by academia and the academic sciences. Genetic science is clearly the most impressive modern illustration of the ‘Emperor’s new clothes’ phenomenon.
IP: Logged
|
|
IF
Member
Member # 1904
|
posted 14. June 2007 11:55
LE,
quote:
Thanks, I’ll give it a look. However, based on the reviews, I doubt if it would add anything to the discussion here.
Which review? It's well written, even very cleverly written, and it won the Pulitzer Prize! From amazon.com:
"Twenty years after it topped the bestseller charts, Douglas R. Hofstadter's Gödel, Escher, Bach: An Eternal Golden Braid is still something of a marvel. Besides being a profound and entertaining meditation on human thought and creativity, this book looks at the surprising points of contact between the music of Bach, the artwork of Escher, and the mathematics of Gödel. It also looks at the prospects for computers and artificial intelligence (AI) for mimicking human thought. For the general reader and the computer techie alike, this book still sets a standard for thinking about the future of computers and their relation to the way we think.
Hofstadter's great achievement in Gödel, Escher, Bach was making abstruse mathematical topics (like undecidability, recursion, and 'strange loops') accessible and remarkably entertaining. Borrowing a page from Lewis Carroll (who might well have been a fan of this book), each chapter presents dialogue between the Tortoise and Achilles, as well as other characters who dramatize concepts discussed later in more detail. Allusions to Bach's music (centering on his Musical Offering) and Escher's continually paradoxical artwork are plentiful here. This more approachable material lets the author delve into serious number theory (concentrating on the ramifications of Gödel's Theorem of Incompleteness) while stopping along the way to ponder the work of a host of other mathematicians, artists, and thinkers.
Topics Covered: J.S. Bach, M.C. Escher, Kurt Gödel: biographical information and work, artificial intelligence (AI) history and theories, strange loops and tangled hierarchies, formal and informal systems, number theory, form in mathematics, figure and ground, consistency, completeness, Euclidean and non-Euclidean geometry, recursive structures, theories of meaning, propositional calculus, typographical number theory, Zen and mathematics, levels of description and computers; theory of mind: neurons, minds and thoughts; undecidability; self-reference and self-representation; Turing test for machine intelligence. "
It covers most of your topics and may give you an even better perspective so that your presentation to the rest of us will be better understood!
IP: Logged
|
|
IF
Member
Member # 1904
|
posted 14. June 2007 14:22
LE,
quote:
The second type of reader, people who actually attempt to use materials contained in the literature to solve scientific problems, have a very different outlook on the literature. To begin with, scientists actually trying to solve problems quickly learn that 99% plus of the literature is irrelevant and useless drivel. Second, hard science analysts realize that useful literature is of two basic types. First, one useful type of literature is literature that discusses specific hard science solutions to real hard science problems. Any new problem addressed by any scientists will involve many different components, and it is likely that other scientists will have already resolved the issues surrounding some components. A productive scientists builds on the actual useful solutions produced by other scientists.
A third kind, me, is one who wants to do a broad search of ideas before doing a deep search of any individual one. (Hence, IF, i.e. IdeaForager)
quote:
The two main points to make here are that 1)most people seem to use references to the literature as magical incantations without really understanding the literature, and 2) even the best parts of the literature report failed efforts to address unsolved problems.
I would like to add at least a third point which is to allow the rest of the world access to thoughts/ideas/solutions/controversies/etc., so that folks are not bamboozled into following someone's explanations on sentiment or word of mouth alone.
quote:
The analysis of the fossil record is simply descriptive. There are no testable scientific theories addressing the change process.
Aren't you forgetting about the predictions of transient forms? For example, in the paper today is the alleged discovery of a larger than expected feathered dinosaur? If you have followed the hypotheses/theories of the evolution of the feathered wing this has great bearing.
quote:
Note that academic sciences create this ‘subjective rejection of theories that don’t explain everything’ when they refused to recognize teleological theories.
I thought that the explanations of facts via theories were supposed to fit together at least reasonably well rather than stand isolated from other related facts and theories.
quote:
A very good argument can be made that genetics is the most absurd and logically unsound of pseudo-science that ever existed. It is certainly far worse than alchemy and astrology combined. Genetics is a based on the continued use of a set of assumptions that are known to be false and known to produce misleading interpretations of the data. Among the central assumptions of genetics known to be false are 1) the assumption that a genotype phenotype mapping can or does exist (there is ample evidence that phenotype of far more complex than genotypes),
Link or reference?
quote:
2) the assumption that genotype changes can cause phenotype changes-the genetic determinism assumption (again the evidence clearly shows that genetic data contains only a tiny fraction of the information required to produce phenotypes),
Giantism, dwarfism, etc. are genetic anomalies that allow us to extrapolate/imagine iterative changes in other portions of the genotype into a phenotype change, no?
quote:
3) the assumption that genetic changes and evolutionary changes can be explained by non-intelligent searches, and 4) the assumption that there are no intelligent goal directed control processes involved in genetic change.
I thought that the concept has been successfully demonstrated by computer modeling!
quote:
The only really interesting thing about the science of genetics is that it continues to be supported by academia and the academic sciences. Genetic science is clearly the most impressive modern illustration of the ‘Emperor’s new clothes’ phenomenon.
Wow! Have you read other "experts" that share your view?
IP: Logged
|
|
LifeEngineer
Member
Member # 3446
|
posted 15. June 2007 07:15
IF,
Quote: Hofstadter's great achievement in Gödel, Escher, Bach was making abstruse mathematical topics (like undecidability, recursion, and 'strange loops') accessible and remarkably entertaining. Borrowing a page from Lewis Carroll (who might well have been a fan of this book), each chapter presents dialogue between the Tortoise and Achilles, as well as other characters who dramatize concepts discussed later in more detail.
It is a mistake to confuse literature and entertainment with formal science. Entertaining and articulate presentations of complex concepts serve a useful role in educating people and in exposing people to new ideas and new terminology. But such entertaining presentations should not be confused with real science.
In the last 50 or so years, we have learned a great deal about intelligent causation, primarily as the result of developing computer systems to simulate complex intelligent processing. But while we have made great progress on the applied side of the scientific analysis of intelligent causation, there has been essentially no progress, at least from the academic side, in solving the theoretical part of understanding intelligent causation.
The issue here, IMO, is the formal hard science analysis of intelligent causation. The better written and easier to read forms of literature may provide some background materials, but they do not provide solutions to the scientific issues being discussed.
It is also important that we make a distinction between 1)the individuals who actually produce literature and 2)the individuals who purport to be experts because they can claim to have read the material and can quote a reference to the material. Way too many people use references as ‘magical incantations’ in an attempt to pretend they actually understand the materials presented.
IP: Logged
|
|
|
Printer-friendly view of this topic
