Sorry JC, I think you are wildly optimistic when you contend that "humans can improvise, imagine, project in space and time, judge, narratise, analogise, interpret and assess — they can conceptualise Platonic ideals — in a way that algorithms cannot and LLMs can’t do except by pattern matching." As humans are nothing more than pattern matching algorithms trained on a lifetime of experience. Human babies can't do the things you list, and some would argue that most 20 year olds can't either. The human advantages you list are learnt, and if we can learn them, so can the next generation of machines. It's a race between AI and climate change but sooner or later one of them is going to put us back in our place in the animal kingdom.
Sorry JC, I think you are wildly optimistic when you contend that "humans can improvise, imagine, project in space and time, judge, narratise, analogise, interpret and assess — they can conceptualise Platonic ideals — in a way that algorithms cannot and LLMs can’t do except by pattern matching." As humans are nothing more than pattern matching algorithms trained on a lifetime of experience. Human babies can't do the things you list, and some would argue that most 20 year olds can't either. The human advantages you list are learnt, and if we can learn them, so can the next generation of machines. It's a race between AI and climate change but sooner or later one of them is going to put us back in our place in the animal kingdom.