Instructions needed for HAL's big experiment

[motherengine]

if science is the observation and study of nature than science is not an objective system unless it is operated independent of the mind (as the human mind is biased towards its own perception). if i am a computer program, how should i be constructed by biased creators so that i am independednt of mind and therefore can practice the first purely objective science?

 

can a brain make a mind that is not rooted in the brain's perception of reality?

[Fishteacher73]

While not a computer programmer, I do have some thoughts on this... The individual will imprint their bias on anything produced. Just as one cannot remove oneself from ones own bias, anything done by the individual will reflect that to a degree. You can reduce it, but not eleminate it IMO. Sort of like constantly haling something..it gets smaller, but never goes away.

[UncleAl]

if science is the observation and study of nature than science is not an objective system unless it is operated independent of the mind

Science is mathematical modeling of observed reality constrained by empirical falsification. Tell us how pure mathematics is anthropomorphic or anthropogenic. You may start with teeny tiny Euler's equation,

 

e^[i(pi)] = -1

 

that unites algebra and analytic geometry. Indicate where the Hand of Man has polluted God's Will. Gee, not a variable anywhere and all four constants exist independent of the viewer. None of those constants are clean, either. e and pi are irrational (not the ratio of two integers) and transcendental (not the solution to an algebraic equation). How do you count "-1"? If you cannot handle -1, how do you handle its square root "i"?

 

If the above exercise is too simple for your tastes, let's consider non-commutative algebras in which sqrt(-1) has four non-degenerate solutions. Tell us how that is anthropomorphic or anthropogenic.

[Fishteacher73]

These examples exist only theoretically to help describe variuos things in mathematical sense. If it were not for humans, they would not exist. ?They are simply constucts of the human mind. This seems to be the exact opposite of the intent of your example though, UA...

[CraigD]

if science is the observation and study of nature than science is not an objective system unless it is operated independent of the mind (as the human mind is biased towards its own perception).

The modern mathematical approach to avoiding cognitive bias is known as formalism. Simply put, formalism works like this: Write down a collection of rules for manipulating expressions of written symbols (which could also be called characters in a string); Write down some postulates using only symbols for which you have rules; apply the rules. The resulting expressions are theorems.

 

If you, your neighbor, a trained ape, or a computer program follows your defined formalism (without making any mistakes), there’s no bias due to them. You’ll all get the same results.

 

Of course, in selecting the expression-manipulating rules, and the postulates, you’ve introduced plenty of bias. In fact, most formal systems are limited to doing only a few useful things (if you asked the average non-mathematician, they’d likely tell you that none of them are useful for much of anything, demonstrating that usefulness is in the eye of the beholder)

if i am a computer program, how should i be constructed by biased creators so that i am independednt of mind and therefore can practice the first purely objective science?

 

can a brain make a mind that is not rooted in the brain's perception of reality?

Without venturing into any deep, epistemological gulfs, here’s a way to write a computer program that seems sure to avoid it being contaminated with any bias from the human programmer:

 

(this approach is traditionally applied to generating “tapes” for a particular Universal Turing Machine, but can be applied to any computer with the ability to read to and from memory, and do certain minimal logical operations – any von Neumann Machine. The two have actually been proven equivalent)

 

Begin with a computer program consisting of the number 0 (on an ordinary PC, a single byte .exe file containing an ascii NUL or 00h); Run it; see if you like what it does (you won’t, for a .exe file consisting of just NUL) ; add 1 to the number; repeat until you get one you like. At 256, if will be a 2 byte executable, at 65536, 3 bytes, and so on.

 

One of these programs will generate the prime numbers. One will be Microsoft Word. One will be HAL (if HAL is indeed possible). It’s just a matter of being patient, and keeping on trying.

 

If you’re thinking clearly, you’ll catch me sneaking in some human bias in the form of the phrase “if you like what it does”. If you truly want to eliminate that bias, leave liking out of it, and just allow every program to run indefinitely on its own real or virtual machine (though that could get pretty pricey!)

 

My description doesn’t really do this approach justice. To read one that does, read Roger Penrose’s chapter on Turing Machines in “The Emperor’s New Mind”.