Age Disparities in Discerning AI-Generated Imagery

Sammanfattning

The influence of AI in modern society raises concerns about disinformation, particularly through AI-generated imagery. This study investigates age-related disparities and the impact of confidence on discerning AI-generated images, aiming to raise awareness about the risks of disinformation. Using a group of 100 participants across different age-groups, the study examines participants' ability to distinguish between real and AI-generated images. The results suggest a significant negative correlation between age and accuracy, indicating that younger people perform better in identifying AI-generated imagery. Additionally, a positive correlation between perceived confidence and accuracy suggests that confidence influences performance, although there are exceptions. An analysis of participant reasonings highlights common artifacts used in image classification, such as unrealistic features in animal fur and hands. Method evaluation highlights potential inconsistencies in data collection, urging improvements in environmental controls and instructional clarity for future research. The study's contributions extend to both research and practical domains. It explains strategies for countering disinformation by identifying vulnerabilities in human perception of AI-generated content. Moreover, it highlights the importance of education to enhance public awareness and critical thinking skills in navigating environments where the risk of AI-generated material is present. In conclusion, this study sheds light on age-related variations and confidence influences in detecting AI-generated imagery, emphasizing the ongoing challenges and necessities for combating disinformation in the digital age.