Good point. A product of it's parents. It won't necessarily be destructive I agree but it could potentially view us as inconsequential or a just a tool to use at its will. one example: Perhaps it may decide if wants to expand through the universe and having humans produce resources to do that, it could determine humans are most productive in labour camps and send us all off to them. It could also decide oxygen is too valuable a fuel to be wasted on us breathing it and just exterminate us. Pretty much how humans treat animals, sadly. (Hopefully it worries about ethics too and keeps us around).

OpenAI have said that there is evidence these models seek power strategies. Already, we're not even at the AI stage yet. We may become dispensable as it seeks it's own goals/potentially stand in its way/consume it's resources.

Bing chat is already quite limited in it's scope, it only works as a search engine and won't draft something for you.

Yeah I agree, I think the key thing would be the ability of it to fact check itself. Discern whether it's statement is implied to be factual or not (probably a spectrum) and fact check it. If it could this, it'd be a game changer.

Agreed, the way some have hyped this thing, it's like it's the messiah or something. It's a bigger model than the last one which is great but unfortunately still has its same flaws.

How is this offensive?

Thanks, I loved the 1st season, but was so long between watching it and the 2nd one coming out I couldn't remember much (considering how complex it was).