As I noted, there's plenty of things to worry about, and your example fits that mold. But it has nothing to do with the "AI will kill us all in 10 years" claim of the rapid self-improvement doomer crowd.