AI would never be able to hijack the political process. AI Super-PACs dominating elections would be a clear milestone in the progression of the "gradual disempo

AI would never be able to hijack the political process. AI Super-PACs dominating elections would be a clear milestone in the progression of the "gradual disempowerment" theory of loss of human control. How could a person dedicated to their denial of the possibility of gradual disempowerment spin this as "good actually?" Am I reduced to "technological improvement is always good"? Or there some logically smaller step I can take?