Deconstructing Dembski #3: The Man with the Golden Arm

In which we cover all but the last section of the first chapter of Dembski's book "The Design Inference"

Category:

Education

Tags:

• randyhelzerman
• kolmogorov
• complexity
• chaitin
• intelligent
• design
• evolution
• creationism
• hairy
• math
• dembski
• the
• infe

[w0r1dpeace]

Randy where the hell are you?

[randyhelzerman]

LOL hanging around

[schrodcat]

Do you really need Kolmgorov complexity for Caputo's paradox? I'd call it entropy. First, his sequence is astoundingly improbable (low entropy). Next, you don't compare it with "any" given sequence like in the previous video, but with the typical sequence (H and T in similar numbers of occurrences, in "any" order, i.e. high entropy). In other words, the probability that a sequence HHH... (more than 40) is generated by a biased coin is practically 1.

[randyhelzerman]

Excellent question, schrodcat. First, low probability alone isn't enough, because ANY sequence of 41 heads and tails has the same probability as any other (under the uniform distribution). Second, a "typical" sequence will have high entropy, that's very true. The fact that caputo's sequence has low entropy would raises eyebrows. The reason I wanted to use Komogorov complexity instead of entropy is that there is a direct connection between a string of one's and zeros being fabricated(cont)

[randyhelzerman]

(cont, to schrodcat) and its Kolmogorov complexity--namely, the shorter the instructions for fabricating it, the lower its Kolmogorov complexity is. Since we suspicion that Caputo fabricated the numbers, this is a natural fit. (BTW, with very high probability, strings with high entropy also have high kolmogorov complexity, so the entropy of the string is a good estimate of the kolmogorov complexity)

[AmusedChild]

1. Couldn't one say that the "fishiness" was caused by predicting, not the random pattern of Dems and Repubs, but the nonrandom choosing of Dems by a Dem?

2. Pattern = sustained reproduction; and is this not the definition of life? (Chem was decades ago!) Aren't we really asking what "randomness" is?

3. Bring on the programming languages! I love them - and I loved BASIC!

[EvoBiologist]

It doesn't matter how many times you throw a coin. 10 is just an arbitrary number to show how unlikely each specific event can be. Use 20 or 20,000, it doesn't matter. Patterns exist whether or not we examine them, just as a tree falling in the forest makes sound waves whether or not someone is there to hear them.

[randyhelzerman]

Hi Mike1977a1, I guess I don't understand your question :-)

[Mike1977a1]

Ermm, if i toss a coin 10 times and find a pattern in the result, why choose ten times?

Isnt it us who sets the rules & finds patterns, but theres nothing in the world to determine why i should go with 10 sets. i dont want to believe that theres absurdity in mathematics and geometry too, i always hoped they where totally determinable!