These AI Traffic Stop Videos Are Both Hilarious and Frightening | Carscoops

      Sora 2, created by the team behind ChatGPT, allows users to create impressively realistic social media videos, including some entertaining car clips.

      23 hours ago

      by Chris Chilton

      OpenAI launched its video generator, Sora 2, on September 30.

      This new version is more realistic than its predecessor and includes native sounds.

      Users can incorporate various characters into their videos, including themselves.

      If your social media feeds are similar to ours, it's likely that a significant portion of your daily content features people engaging in humorous, risky, or thrilling activities with cars. Each day brings new clips of individuals testing boundaries, intentionally or not.

      But what if you didn't have to rely on the chance that someone with a smartphone would capture a car-related moment that is both entertaining and hilarious? What if you could create it yourself using AI?

      That’s the current scenario as thousands explore OpenAI’s Sora 2. This TikTok-like app enables users to produce videos for sharing on platforms such as TikTok or X.

      Reality Gets a Rewrite

      Sora was launched last year, but the second generation, available since September 30, represents a significant advancement. It features improved physics, synchronized audio, and, most importantly, it – for the most part – appears entirely convincing.

      This is particularly remarkable given that the clips we’ve included below portray characters like Scooby and Shaggy as well as SpongeBob SquarePants being pulled over by the police. Clearly, these scenarios aren’t real; they can’t be.

      Yet, your brain can easily accept them as real, much like how we become engrossed in watching a Pixar film and quickly forget that we're viewing animated sequences.

      Brilliant, Terrifying, or Both?

      Some clips, which are less absurd yet still outrageous—such as Michael Jackson stealing food in a fast food restaurant—are both brilliant and unsettling. They emphasize how rapidly the technology is advancing and how easily it can distort the boundary between reality and fabrication.

      Looking ahead, we may need to be cautious about trusting what we see on our screens, whether it’s a Mustang crashing dramatically after a car meet (even if SpongeBob isn’t driving) or a video of a notable politician engaging in behavior that could damage their image.

      When Fun Turns Problematic

      Moreover, since this new AI generation allows users to insert themselves and their friends into videos, there is a risk that things could go awry, even though most users may treat it as harmless fun. Imagine competing for a job promotion in ten years and losing out because of a video showing you failing a sobriety test and assaulting an officer.

      Sora 2 appears to be quite entertaining, with a clip of Stephen Hawking landing on a skateboard ramp in his chair providing some amusement. However, shortly after the launch, OpenAI agreed to offer more control to concerned copyright holders regarding how their characters could be utilized. Nonetheless, the future it suggests is indeed troubling.

      Image Sora/YouTube