After the Government's Global AI Safety Summit - and the Fringe events around it - what does the road ahead to responsible AI look like?