Figure 2: Japanese participants in our Experiment 1 and orig-inal U.S. participants in [37] show the same response patternto the question “How much blame does the [agent] deservefor [not] opening the chute?”.
I think it's really interesting that despite the cultural differences in perspectives on robots the authors cite (helpers vs. logical machines we should fear), the basic trend of blame judgment in figure 2 is roughly the same for both Japanese and US participants. One difference that I thought was interesting that I think can be attributed to this cultural perspective was the similarity in Action bars for Japanese participants in both human and robot cases, whereas the one for the US differs by around > 5%. Disqualification rates are also lower in experiment 2 for Japanese participants. Are Japanese robots just so anthropomorphized to the extent that Japanese participants feel comfortable assigning similar moral blame to them as a human, despite the blame being technically belonging to the designers?