Considering research “paths” are often measured in decades I’d be massively impressed if a scientist could guess what would be profitable 20 years from now.