You're lounging in the living room after a big holiday meal when your uncle starts scrolling through vertical videos. “Did you see the cat snatch the snake out of the dude’s bed?” he asks.
Is this real? It's a fake copy? You feel a headache coming on.
“We are overwhelmed by slop,” said Mike Caulfield, co-author of the book. Tested: How to Think Smart, Be Less Confused, and Make Better Decisions About What to Believe Online. “It just floods the area and at some point your mental capacity just gets drained.”
But Caulfield and other experts say there's no need to despair, at least not yet. There are a few simple dos and don'ts to try to judge the authenticity of something you see online.
Don't think everything is fake
With so much slop in our feeds, it's easy to think that All you see on the internet, it's fake. But that bias is just as dangerous as believing everything you see is real, warned Kolina Koltai, a senior investigator at Bellingcat, an organization that specializes in open-source investigations.
Eyewitness video remains an extremely important source of evidence of wrongdoing by individuals and law enforcement agencies. When people stop believing these videos, researchers call it “liar's dividend” because it is easier for attackers to claim that real events are fake in order to evade responsibility.
“I think that’s one of the biggest risks when it comes to this kind of content,” Koltai said. “It’s not that someone will believe a fake video, it’s that people won’t believe a real video.”
Koltai and others say it's especially important to carefully consider videos that may evoke a strong emotional response or conflict with your beliefs. Real-life videos often present complex situations that can provoke reactions that challenge our understanding of the world. However, many fake videos are created precisely for this purpose – to increase engagement.
Pay attention to some simple features of the video.
AI-generated videos are already excellent and are rapidly improving, says Hani Farid, a professor at the University of California, Berkeley, who studies manipulated media. According to him, even experts can be fooled. “I've been doing this every day for a long time and it's really hard. Really hard.”
But there are some pretty simple features that can tell you whether the content you're watching is AI-generated. The biggest tip is the length of the video.
Most companies limit the length of AI videos “because creating these videos is very computationally expensive,” Farid said. Many videos are only 8-10 seconds long. While it's possible to put together a longer video with many short clips, “when you see those little videos, it's a good sign that you should take a breath.”
Length isn't the only thing; According to Farid, videos created using artificial intelligence tend to perfectly represent the subject. The main characters of the video are highlighted prominently, and the action begins and ends clearly, even if the video is short. This is one reason why the quiz show video of a New York City cop yelling at ICE agents is so obviously fake.
“It almost looks professional,” he said. The camera position can also be strange – is it too close to the objects of, say, an ICE raid? Or does it move too smoothly to follow a running animal as if it were suspended? These could be AI-generated tips.
Be sure to check the context
The characteristics of a video are important, but where it is posted can sometimes matter even more, Caulfield said.
Checking where the video was posted and even just looking at the comments can provide valuable clues. For example, the second ICE raid video in the quiz was submitted by the Reddit community in the Logan Square neighborhood of Chicago.
Likewise, look who posted the video. If their feed contains content about more than just immigration raids, it lends credence to the idea that they witnessed a raid. “It might be easy to fake a video, but it's hard to go into a time machine and build a ten-year history around Chicago hot dogs,” Caulfield said.
If you're not sure whether the video you're viewing was originally posted by the account you're viewing, try a simple reverse image search on Google or another platform. Often this search will find the original post, other videos from the same event, or news reports that either confirm or refute the video. Both ICE raid on Logan Square And moose eating popcorn which were reported by the media at the time they occurred.
Conversely, identifying an AI video is often as simple as taking a closer look at the account that posted it. According to Koltai, accounts often list their content as being created by artificial intelligence in their profile description. Even if this is not the case, checking the comments can often reveal that many people believe that the video was created by artificial intelligence.
Don't feel like you have to share it, especially if you have doubts.
Finally, all three researchers agree that in an era where algorithms value speed over accuracy, sharing is not really caring.
According to Koltai, most AI content published online is attention bait. Its creators “often have a monetary incentive to get you to like, comment and share because it often results in them making more money,” she said.
When in doubt, Caulfield said, the best thing to do is wait. “You don’t have to be the first person to share this thing, you can be the person who waits,” he said. Often within hours, video footage of an event is confirmed by corroborating videos or news reports.
Many people may not think it matters whether you share AI videos of bunnies jumping on a trampoline or a cat snatching a snake from its owner's bed, but experts agree it does. When people are deceived by AI videos, it undermines faith in important videos.
“People are like, oh, is it really that important?” she said. But Koltai said everyone should be worried. “If we can't tell what's real and what's not real on the Internet? This is really incredibly dangerous for me.”
Hani Farid agrees: “Every one of those likes, clicks, shares, interactions—you are part of the problem at this point,” he said.
