AI Video has gotten scary real. Here are some tips on how to spot them.

Chances are, you’ve already fallen for an AI video that someone tried to pass off as real. AI video technology has gotten so good that sometimes it’s impossible to distinguish reality from a computer’s hallucination. We’re going to give you a couple tips on how to do just that on this week’s Tech Byte.

When AI video first launched, it looked like a wild psychedelic vision straight out of the 1960’s. The benchmark for AI model quality was how accurately it could depict Will Smith eating Spaghetti. Compare a few years ago to today and you can see how real these videos are starting to look.
While most of the time, these AI videos are just fun and games, deepfakes can be truly dangerous, spreading misinformation about real-world events and people.
One of the best ways to tell an AI video from a real one is to look at the background details. Do objects suddenly appear, change shape, or otherwise break the laws of physics? If so, there’s a high chance it’s not real.
Also, many AI video generators have a floating watermark along the corner of the screen. That’s a perfect way to tell if a video is AI generated or not.
One of the more popular genres these days is fake surveillance camera and doorbell camera videos.
If a video title is overly heartwarming, and the premise is too good to be true, it probably is. These videos are made to generate that sweet, sweet ad revenue at the expense of your sense of reality and general sanity, so be careful!

As always, a healthy dose of skepticism and same fact checking can go a long way when it comes to these AI videos.