A recent video from Guardian Australia explores the growing difficulty of distinguishing between real and AI-generated faces. The University of New South Wales has launched an interactive test that challenges users to identify whether a portrait is authentic or computer-generated.
The AI Faces Test
As artificial intelligence technology advances, creating hyper-realistic human faces has become increasingly accessible. The university's test presents a series of portraits, and participants must decide if each image is real or fake. The challenge highlights how sophisticated AI generation has become, often fooling even trained eyes.
Taking the Challenge
Guardian Australia's picture editor Carly Earl and reporter Matilda Boseley put their skills to the test. The video captures their attempts and reactions as they navigate through the portraits. Earl, a photographer with years of experience, and Boseley, a self-proclaimed internet addict, represent different perspectives on visual literacy.
The test is not just about accuracy but also about understanding the subtle cues that might reveal an AI-generated image. From lighting inconsistencies to unnatural facial features, the test pushes participants to rely on both technical knowledge and intuition.
As the line between reality and simulation blurs, such exercises become crucial for public awareness. The video concludes with reflections on whether spotting AI faces is a science or merely a matter of vibes.
