No. Ai cannot “think” any of its own original “thoughts”, it’s usually trained on LOTS AND LOTS of human generated text / data and uses highly complex algorithms to generate it’s response to whatever human input is given as a prompt.
If it can’t generate it’s own original thoughts, it can’t conspire on its own.
If i had to guess what the conspiracy theories are about, it’d be related to more recent ai models being trained/created with ai generated text/data. The mere existence of the “rokos basilisk” thought experiment is probably a common starting point / “core” of a lot of whatever zany ai conspiracy theories are floating about. They’ve been floating around for a long time before chatGPT got popular, but chatGPT and other ais all becoming more popular could only ever further increase the amount of ai related conspiracy theories.
Conspiracy theories and other similar distrusts of new technologies is actually pretty common throughout history, so this is basically just a continuation of that historical trend.
What makes you so confident your first paragraph leads to the second?
I suppose a better way to highlight it is the where’s the difference between this and a natural human?
Assuming a brain is a complex computer, and our senses more complex training data/promts, we fit the qualifications so must also be incapable of original thought
do computers perspire?
Do computers inspire?
Do computers respire?
Do computers aspire?
Do computers expire?
Do computers catch fire?
Do computers acquire?
if you program them so, yes, of course.
Also, if there is a error in a similarly made program. (in science fiction there are many proposed scenarios for this : 2001 : A space Odyssey, Colossus: The Forbin Project, …)some google employees on the AI team are just reformatting search results and calling it AI. spell check is better at thinking