Researchers from the University of Chicago Booth School of Business conducted a series of fascinating social experiments in which they observed people’s reactions to AI and robot preachers in temples and churches.
These experiments revealed that people are not likely to trust robots as religious leaders, and they may even stop giving donations to religious groups that use robot preachers and AI to spread their message.
“Our research arrives at a point where automation is pervading every job industry, and it suggests that some professions may not be automated so easily. Robots may struggle in professions, like priests or monks, that require high levels of credibility,” Joshua Conrad Jackson, lead researcher, and a professor at Chicago Booth told ZME Science.
Testing robot preachers
It seems pretty crazy to even think about it, but then again, an AI winning an art competition would have seemed crazy a few years. As AI is taking on more tasks that were previously considered human-only, it’s not even clear what tasks are truly restricted to humans. AI preachers? Let’s give it a go.
Joshua and his team wanted to see how persuasive and effective robot preachers can be. They conducted three separate experiments. The first experiment took place in Japan where they used Mindar, a humanoid robot designed to teach the learning of Buddhism and bring Japanese youth closer to their religion. Yes, this already exists.
Mindar was developed by scientists at Osaka University and it cost 106 million Yen (~$755,000). The robot has human-like facial features and during the experiment, it preached sermons from Buddhism to hundreds of people who visited the Kodai-Ji Buddhist temple in Kyoto.
The second experiment was conducted in a Singapore-based Taoist temple. Here one group of devotees listened to a human preacher and another group heard sermons delivered by Pepper, a robot developed by SoftBank Robotics. It is a popular robot in Japan, where it was first made commercially available in 2015.
Believers are not fond of robots
At the end of the first and second experiments, the researchers asked 359 and 239 devotees, respectively, about their views on the robot preachers. Most participants in both countries considered robot preachers less credible than their human priests.
“In both our field studies, participants who were exposed to a robot (vs. human) preacher donated less to the temple, were less likely to agree to circulate flyers for the temple and were less willing to circulate the message of the sermon. We found these results even when the content of the sermon was identical in the two conditions,” said Jackson.
However, what’s more interesting is that when the researchers asked the devotees to rate robot and human preachers out of five for credibility (five for highest credibility), it was almost a tie as the average rating of robots and human preachers were 3.12 and 3.51.
“This suggests that there are a lot of people out there who think robots could be effective preachers, but there are more people who aren’t convinced,” Jackson added.
It’s noteworthy that the experiments were carried out in countries that are usually very supportive of emerging technologies. In other countries, the robots would have likely been even less effective.
What about Christian sermons?
Finally, the researchers went for a third experiment and this time it involved a generative AI program and 274 followers of Christianity in the US. They divided the followers into two groups and asked the members of both groups to read a sermon online.
They told the participants in the second group that an AI program produced the sermon; but to members of the first group, they said that a human preacher wrote the sermon. When they asked all of them about their experience, the members in the second group responded that the sermon was less credible as it came from an AI.
In other words, it’s not so much what the AI is actually doing — it’s how people perceive the AI. This is essentially an attitude difference, and attitudes towards technology can change relatively quickly.
“We conducted our research in Singapore, the USA, and Japan, so we cannot make claims to other cultural contexts where attitudes towards robots or religious authorities are different,” Joshua told ZME Science.
However, these findings indicate that, for now, robots and religions don’t go too well together.
“Robots and AI programs can’t truly hold any religious beliefs so religious organizations may see a declining commitment from their congregations if they rely more on technology than on human leaders who can demonstrate their faith,” he added.
The study is published in the Journal of Experimental Psychology.