The arguments on safety of the AI especially, but it is a general comment on the arguments surrounding that sort of issue in general.
edit: To clarify. It is a fact of fuzzy decision systems that they diverge from being useful exponentially or worse in number of steps. You may not like this fact, but that is how things are, and when you are fuzzily thinking of the fuzzy concepts, the resulting effort is useless if not worse than useless (which it easily can be if its inaccuracy is under-estimated).
Was this intended in response to something in particular?
The arguments on safety of the AI especially, but it is a general comment on the arguments surrounding that sort of issue in general.
edit: To clarify. It is a fact of fuzzy decision systems that they diverge from being useful exponentially or worse in number of steps. You may not like this fact, but that is how things are, and when you are fuzzily thinking of the fuzzy concepts, the resulting effort is useless if not worse than useless (which it easily can be if its inaccuracy is under-estimated).