[killthebuddha]

Perfect reasoning, with certain assumptions, is perfectly deterministic, but that does not at all imply that it's predictable. In fact we have extremely strong evidence to the contrary (e.g. we have the halting problem).

[stevenhuang]

This sounds confused. Why do you think the halting problem is relevant to predictability? Undecidable problems != Unpredictable problems