We’ve opened Pandora’s AI box with ChatGPT, the generative AI chatbot that took the world by storm late last year. Some of the brightest minds in AI have since started to worry about artificial intelligence potentially endangering humanity. Some have called for a slowdown in AI development, but I don’t think that’s possible now that ChatGPT is here.
That’s because anyone with the proper training can develop amazing AI capabilities, like the so-called Disentangled Control for Referring Human Dance Generation in Real World (DisCo) generative AI that can create a video of you dancing using a single photo of you.
What DisCo does is rather simple. It splits your photo into three parts. It looks at the background, foreground, and person’s pose.
After that, it uses that information combined with dance data coming from the TikTok dance videos the AI has trained on to create a video of you dancing to a certain tune.
Ironically, if you have been releasing dance videos on TikTok, the AI might use information from your clips to animate a static photo. But you’ll never know whether DisCO had access to your videos.
The simplest use case for this technology is generating dance videos on TikTok without learning new choreographies yourself. That would also mean you’d generate fake content for the platform, assuming you can access DisCo.
But DisCo could have wider applications in the future, like using the app to preview complex choreographies for shows that involve actual humans. You could see what a dance routine looks like before you start practicing it. The AI would show you how and when to move your body.
DisCo can also be used in the post-production phase of movies and TV shows. Studios will be able to add dance routines to their projects without hiring dancers.
That said, you should not expect perfect videos from these first-gen AI-generated clips. Also, DisCo might not be the only generative AI software to create dance videos from images. Take this TikTok clip (via YouTube) that shows statues coming to life to dance: