Can we trust robots to make moral decisions?