Anatomy of a "Bad Vibe": An Interoceptive Prediction Model of the Uncanny Valley

This research project explores why humans feel uneasy when interacting with AI that is almost, but not quite, human-like, known as the "uncanny valley" effect. The study proposes that this feeling is caused by a mismatch between our brain's prediction of how we should feel when interacting with a human-like agent and our actual physiological response.
The project aims to investigate whether individuals with higher interoceptive awareness (the ability to sense internal bodily signals) experience stronger negative reactions to uncanny AI. The study will use a combination of psychophysiological measurements, such as EEG and eye-tracking, and self-reported ratings to test this hypothesis.
The expected outcome is that the research will provide evidence for the interoceptive prediction error hypothesis, which could have practical applications in designing more human-like AI and developing therapeutic tools for conditions related to poor interoception, such as anxiety and autism.
Meta AI response: This research project explores why humans feel uneasy when interacting with AI that is almost, but not quite, human-like, known as the "uncanny valley" effect. The study proposes that this feeling is caused by a mismatch between our brain's prediction of how we should feel when interacting with a human-like agent and our actual physiological response. The project aims to investigate whether individuals with higher interoceptive awareness (the ability to sense internal bodily signals) experience stronger negative reactions to uncanny AI. The study will use a combination of psychophysiological measurements, such as EEG and eye-tracking, and self-reported ratings to test this hypothesis. The expected outcome is that the research will provide evidence for the interoceptive prediction error hypothesis, which could have practical applications in designing more human-like AI and developing therapeutic tools for conditions related to poor interoception, such as anxiety and autism.

Imagine you're watching a movie with a robot that looks almost like a human, but something about it feels off. This "uncanny valley" feeling is what researchers want to understand. They think it's not just about how the robot looks, but how our body reacts to it.
Our brain expects to feel a certain way when interacting with something human-like, but if it doesn't match that expectation, we feel uneasy. The study aims to see if people who are more in tune with their body's signals are more sensitive to this feeling.
The findings could help create more realistic robots and AI, and even help people with conditions like anxiety by developing new therapies.
Meta AI response: Imagine you're watching a movie with a robot that looks almost like a human, but something about it feels off. This "uncanny valley" feeling is what researchers want to understand. They think it's not just about how the robot looks, but how our body reacts to it. Our brain expects to feel a certain way when interacting with something human-like, but if it doesn't match that expectation, we feel uneasy. The study aims to see if people who are more in tune with their body's signals are more sensitive to this feeling. The findings could help create more realistic robots and AI, and even help people with conditions like anxiety by developing new therapies.