Alternatively, nothing will change too much because we don't really care. Watson was a fun Jeoparody contestant, but not a permanent addition to the show. Watch

Alternatively, nothing will change too much because we don't really care. Watson was a fun Jeoparody contestant, but not a permanent addition to the show. Watching robots do stuff is boring: if you think religion and sports are going to get disrupted by AI, I'd wager you're not considering why they're popular in the first place. If you presuppose anything, you have to try to justify it. That's what ChatGPT is doing - it has no model of "collapse" or understanding of the outside world - it's just taking your prompt ("assume the world ends by AI") and tries to come up with a believable justification. It's testament to how indiscriminate and willingly wrong it can be. The fact that AI unconsciously oscillates between fact and fiction makes it unsuitable for world-domination, at least without serious assistance. We have multiple adversarial nation-states with nuclear armaments and advanced manufacturing tools that could wipe out all human life. Something tells me an erudite text-generation tool isn't going to end the world in my lifetime.