The purpose of this report is to explore people's recognition abilities regarding AI-generated images versus human-created images, as well as to understand what factors influence their judgments, but also how it may relate to the Uncanny valley theory.

The study included two types of image categories, faces and landscapes of varying difficulty. Through interviews with eight people, the results showed that the ability to recognize varied between faces and landscapes, where the respondents had an easier time pointing out the faces than the landscapes. During the interviews, respondents used words like "off ", "unnatural" and "fake" to describe the AI-generated images and looked for details that deviated from the "natural".

This study also concluded that the Uncanny Valley theory was relevant to the respondents when they guessed the pictures, for both the faces and the landscapes. The reflections after the tests show that the respondents feel a mixture of concern and fascination about the development of AI technology.

The conclusion of this report is that the recognition ability is better in identifying AI-generated people versus AI-generated landscapes, then that the Uncanny Valley-theory can be applied to the faces and also the landscapes. Also that there is a concern among the respondents about the development of AI technology that needs to be taken seriously.