The actual paper (thanks, p_ing!) seems more measured than most, and consistent with what economists like Acemoglu have been saying - possibly up to 1% per year improvement in productivity, all going well.