Re: A Critical Question

[EMAIL PROTECTED] (Daryl McCullough) writes:
>Neil W Rickert says...
>>[EMAIL PROTECTED] (Daryl McCullough) writes:

>>>The modern foundation for thermodynamics is statistical mechanics.
>>>Statistical mechanics defines entropy in terms of information, and
>>>defines temperature in terms of entropy.

>>It is my impression that the entropy from information theory is
>>different from the entropy from thermodynamics.  They are analogous,
>>in that they satisfy the same mathematical equations.  But they are
>>different things.

>Yes, but there is a common idea connecting them: in both cases,
>entropy quantifies the relationship between gross (statistical)
>knowledge of a system state and detailed knowledge. In both cases,
>the entropy is the expected number of bits of information required
>to specify the details given the gross knowledge.

I don't have any problems with that.

I tend to think that the Shannon concept of information is the wrong
one for cognitive science.  /dev/random supplies entropy, but you
won't be much informed by reading its output.