Reviewing the 2003 experiment, Mary Cummings, a human-computer interface expert, noted that even the published results indicated that none of the signs of “overload” that the researchers were testing were consistent across all three variations of the test
