Theory of Everything - View Single Post

ccristi99 · **Computational irreducibility -** 12-05-2004, 07:01 PM

WM, I read about your computational irreducibility and I must say I don't fully agree, or maybe I don't fully understand it. To "compute" means to build a model of a system and then evolve it based on its rules. Irreducibility is therefore relative to the model used. Change the model and you can always reduce it further. Example:
If addition is the only operation allowed in your model then the following expression cannot be reduced: 2+2+2+2. As you were saying, you just have to perform all three operations to get to the result. However you can invent multiplication and just do 2*4 (one operation) for the same result.

Some people will argue that multiplication is more expensive, but that's just how it's implemented now in computers. In school we learned both addition and multiplication for small numbers as lookup tables, which makes them equally fast.

	(#5 (permalink))
ccristi99 In Training Status: Offline Posts: 4 Thanks Given: 0 Thanked 0x in 0 Posts Join Date: Aug 2004 Rep Power: 0	Computational irreducibility - 12-05-2004, 07:01 PM WM, I read about your computational irreducibility and I must say I don't fully agree, or maybe I don't fully understand it. To "compute" means to build a model of a system and then evolve it based on its rules. Irreducibility is therefore relative to the model used. Change the model and you can always reduce it further. Example: If addition is the only operation allowed in your model then the following expression cannot be reduced: 2+2+2+2. As you were saying, you just have to perform all three operations to get to the result. However you can invent multiplication and just do 2*4 (one operation) for the same result. Some people will argue that multiplication is more expensive, but that's just how it's implemented now in computers. In school we learned both addition and multiplication for small numbers as lookup tables, which makes them equally fast.