I certainly agree with your assertion that any computable problem can be converted to a Turing machine!
My hypothesis asserts that an algorithm (as defined above) cannot arise from null.
Hugs!
Well this is certainly true insofar as you have to have some positive Kolmogorov complexity for any kind of computation to occur. Quite a bit more if you want something interesting to happen.
That said, one could question your hypothesis by questioning your assumption that our universe and/or whatever is outside our universe ever had a null information substrate. We have no reason to believe this was the case (or the opposite for that matter), and the very fact that you can't bootstrap machinery from a null Kolmogorov complexity makes it doubtful. I see no reason to invoke another degree of freedom by positing a bootstrap from a null substrate (remembering that timeline causality in the conventional sense is merely an artifact of our cozy corner of the universe), since positing a non-null substrate is just as reasonable and doesn't make mathematicians do the monkey dance. My weapon of choice here would be Occam's Razor.