AI-generated Images Blur Lines Between Reality and Fakes
DW Fact Check recently put three “historical” videos to the test, and the results were shocking – they were all AI-generated. But can you tell the difference between a real and fake image?
Archaeologists in Italy are using AI to reconstruct images of victims from the 79 CE eruption of Mount Vesuvius in Pompeii. It’s a remarkable example of technology being used to shed new light on history. But with the rise of AI-generated content, it’s harder to separate fact from fiction.
Spotting the Fakes
DW Fact Check used AI-generated images to create fake videos of historical events. These images looked so real that some people thought they were authentic. But there are clues to look out for:
Check the resolution and quality of the image. AI-generated content often has a lower resolution and higher noise levels than real images. Look for small details that seem off or inconsistent.
Pay attention to anachronisms. If an image appears to be from a different time period or contains elements that wouldn’t have been present during the event, it’s likely a fake.
Examine the lighting and shadowing. AI-generated images often struggle to replicate the subtleties of real-world lighting, so look for unrealistic or inconsistent shadowing.
Look for inconsistencies in the subject’s behavior. AI-generated content is often limited by its training data, so if the subject’s behavior or movements seem unnatural or inconsistent, it’s likely a fake.
What This Means
As AI-generated content becomes increasingly sophisticated, it’s essential to develop the skills to spot the fakes. This isn’t just about verifying historical images – it’s about staying informed in a world where misinformation can spread quickly. By learning to critically evaluate the images we see, we can stay one step ahead of fake news and disinformation.
