The cover article in August 22th-28th issue of The Economist is a discussion of whether LLM's of AI do or will have feelings. It uses the term "J-space" to suggest that some models might be conscious.
All of this is way, way above my pay grade, but seemingly very important to those working on the cutting edge of AI. Reading it, I'm reminded of how HAL in 2001, A Space Odyssey went rogue, only to be disabled by the remaining crewman pulling the plug that supplied his power.
I doubt that anyone "will pull the plug" on AI. So folks like me will have to tighten our seat belts and hang on.