I personally feel that universities need to rethink some of their "tool obsession" at the student selection stage because it leads in part to scientists (especially in the UK) who lack appreciation of the "non-scientific" parts of society.

But, even if we say we have to retain a hyper-specialist stream in schools to educate future physicists etc. it's time we recognised that future "artists/politicians/whatever" need a different curriculum to help them relate to science.

At one silly level, my glib statement about this is always "we should be teaching statistics instead of calculus." That's true in that some sense of risk analysis and understanding expression of polling etc. would be useful, but it's still rather a tool oriented response.

Really, in a way, it would be best to ask some artists etc. who managed to develop a good relationship with science not just how to spark their interest, but about the pathway that education could take that would work for them.

I don't say this will solve everything for everybody, but I am sure that the way we do things now really isn't the best way.