Text-to-Future

📰 AI can now animate humans from a line of text: New “Text-to-Motion” research generates actual motion data that informs 3D character movement when fed a single sentence (vs. Text-to-Video which only generates a video).

"Natural and expressive human motion generation is the holy grail of computer animation," says the research team who published the findings.

Text-to-X artificial intelligence research papers are sweeping every medium this year... to-Art, to-Photo, to-Video, to-3D, to-Music, to-Environment, to-Avatar, to-Expression, etc

What's next?

Research is improving the generation of intelligent text itself (the input to these emerging Text-to creative processes).

Connecting the dots of this creative pipeline suggests a future where text-generating "AI Creators" produce and iterate on media that scores itself against the engagement it receives on social platforms.

These AI Creators will do what they do best—train and improve their creation models against years of social media data until, in certain mediums, humans may not be able to make more engaging content than AI.

Humans will have personal, content-creating AI friends who generate original media and text them exactly what they want to see. Dangerous?

"Human-Generated" may become a genre/tag before we know it, and that’s going to unearth both identity challenges and opportunities for creators.

Research paper out of Tel Aviv University here: https://lnkd.in/gW8yhMHt

What do you think about this? How can artists differentiate themselves in a fully-generative media landscape? Being a great wordsmith? Being more creative? Telling a great story? Showing more emotion? Partnering with AI? Touching grass?