I think the argument is simple because the general reporting on AI hasn’t even met that bar. It’s little more than breathless repetition of advertising copy. There are basic questions that aren’t getting asked. It’s a computer program. Why should this program be treated as potentially sentient when others aren’t? Because its complicated, complex? Ok (seems convenient, but ok) …except you can very easily make a complicated, complex program in the same way only train it to spit out absolute useless noise, and neither the program or the hardware have a preference for one or the other.