[LolWolf]

I disagree. Why can we not just get really really good at measuring the universe and describe everything from there? If you agree with physics as a decent approximation of the universe, everything else is collective/emergent (though statistical, sure) behaviour; so there is no 'natural line' of impossibility at any level. In fact, I'd argue it's the opposite, but maybe I'm misguided and highly biased (as a physicist).

I guess my question is: what's your reasoning behind the statement that such a line exists?

[algesten]

Humans behaviour. A gazillion synapses interconnected together first directly in our brains and then by being part of a context.

I think it's like building a model accurately predicting the grains of sand around the world. Sure, theoretically we can get really really good at measuring everything, so we pretty much know where all those grains are, but I suspect we're not close.

Instead we must do, as you say, "decent approximation" – and this is where my line comes in. I think the crudeness of our current models to predict human behaviour are equivalent to roughly knowing where the Sahara desert is.