What's the connection between the link and the url, you may ask?
Well, have you heard, or perhaps made, the insightful claim that all languages are essentially the same since they all boil down to machine code?
In one sense this is correct, but in the same sense that everything in the world is the same because there are really only colliding atoms (quanta changing energy levels, or what have you). Yet, we still think there is such a thing as a "thing."
Why is this? To cut a long story short, I submit the above linked Kolmogorov complexity is mathematical proof of the existence of "things." To bring this back into the CS realm, pg has said the only way to get the power of lisp in another language is to essentially recreate lisp in said language. So, if it is coherent to equate a language to a set of specific algorithms or programs, then it is a short hop to say certain programs are irreducible (i.e. Kolmogorov complex) and are therefore not just machine code.
No matter the method or number of numbers a computer cranks through, it cannot recognize these special programs without beginning with them (yes, very similar to Godel's undecidable sentences). And thus, one distinct k-complex computer language cannot eclipse another without already containing the k-programs of, and therefore being, the other.
Well, have you heard, or perhaps made, the insightful claim that all languages are essentially the same since they all boil down to machine code?
In one sense this is correct, but in the same sense that everything in the world is the same because there are really only colliding atoms (quanta changing energy levels, or what have you). Yet, we still think there is such a thing as a "thing."
Why is this? To cut a long story short, I submit the above linked Kolmogorov complexity is mathematical proof of the existence of "things." To bring this back into the CS realm, pg has said the only way to get the power of lisp in another language is to essentially recreate lisp in said language. So, if it is coherent to equate a language to a set of specific algorithms or programs, then it is a short hop to say certain programs are irreducible (i.e. Kolmogorov complex) and are therefore not just machine code.
No matter the method or number of numbers a computer cranks through, it cannot recognize these special programs without beginning with them (yes, very similar to Godel's undecidable sentences). And thus, one distinct k-complex computer language cannot eclipse another without already containing the k-programs of, and therefore being, the other.
What say you?