This idea of solidifying false memories through repeated exposure is part of what could make false AI images such a risk, says Prasad, who is now a graduate student of psychological and brain sciences at Dartmouth.
“I think misinformation, especially when presented as the truth repeatedly, could definitely lead to the creation of false memories, or at least, doubting the validity of your own memories,” she says.
False Memory and AI
If the Berenstain Bears are a standout example of the Mandela effect, then the viral image of the late Pope Francis wearing a giant Balenciaga puffer jacket is a standout example of what this phenomenon could look like in the world of AI.
“The pope in a fluffy coat was one of the first [generative AI images] that went viral,” says Jen Golbeck, a professor at the University of Maryland’s College of Information who studies AI, social media, and trust online. “And there’s probably people who saw that image and didn’t realize that it was [AI] generated. We’re in such an interesting time now where people know they have to be suspicious, but some people just don’t care.”
There are a number of headwinds that have made the perfect storm for this kind of misinformation to spread, Golbeck said, including the rise of “fake news” sites, the erosion of institutional trust throughout the COVID-19 pandemic, and the rise of increasingly convincing content from gen-AI. The risk of being tricked by AI comes not only from human-prompted creations, but from believing AI hallucinations as well.