A Cambridge philosopher has warned that there is currently far too little evidence to determine whether artificial intelligence is conscious, arguing that humanity does not yet understand consciousness well enough to test for it in machines.