Meta's New Text-to-Video AI Generator is Like DALL-E for Video (theverge.com) 10
AI text-to-image generators have been making headlines in recent months, but researchers are already moving on to the next frontier: AI text-to-video generators. From a report: A team of machine learning engineers from Facebook's parent company Meta has unveiled a new system called Make-A-Video. As the name suggests, this AI model allows users to type in a rough description of a scene, and it will generate a short video matching their text. The videos are clearly artificial, with blurred subjects and distorted animation, but still represent a significant development in the field of AI content generation.
"Generative AI research is pushing creative expression forward by giving people tools to quickly and easily create new content," said Meta in a blog post announcing the work. "With just a few words or lines of text, Make-A-Video can bring imagination to life and create one-of-a-kind videos full of vivid colors and landscapes." In a Facebook post, Meta CEO Mark Zuckerberg described the work as "amazing progress," adding: "It's much harder to generate video than photos because beyond correctly generating each pixel, the system also has to predict how they'll change over time."
"Generative AI research is pushing creative expression forward by giving people tools to quickly and easily create new content," said Meta in a blog post announcing the work. "With just a few words or lines of text, Make-A-Video can bring imagination to life and create one-of-a-kind videos full of vivid colors and landscapes." In a Facebook post, Meta CEO Mark Zuckerberg described the work as "amazing progress," adding: "It's much harder to generate video than photos because beyond correctly generating each pixel, the system also has to predict how they'll change over time."
Next up, VR! (Score:2)
Bring it meta. A new AI driven VR development platform would be insane.
Meta is the least of the folks doing this (Score:2)
Meta is just grabbing headlines: *everyone* is doing text-to-video at this point, there's many videos posted daily in almost every forum out there on this stuff. It's almost a good thing MTV is a thing of the past, because every small local band out there is about to have an LSD-inspired music video featuring trippy, endlessly mutatating visuals.
Re: (Score:2)
No. What most folks are doing is text to still. Text to video is a reasonable next evolution, though.
WELL DONE! (Score:2)
Will the family of Molly Russell be in a congratulatory mood? No they won't.
After this little achievement perhaps there will now be some worthwhile engineering effort applied to protecting children from the suicidal shit that they should not be watching.
Re: (Score:2)
After this little achievement perhaps there will now be some worthwhile engineering effort applied to protecting children from the suicidal shit that they should not be watching.
Hopefully not! The only way engineering can "solve" this is by surveillance, suppression and censorship and generally making the world a worse place for everybody, including the children. It will also not work on the "protection" side. Protecting children from stuff on the Internet is the job of their parents and their teachers.
Mostly this needs to be done by making the children more resilient and telling them when to stop watching and to get help. That help has to be, of course, available and given without
Does it work for porn? (Score:2)
No? Ok, will be a failure then.
Does it work with named individuals? (Score:3)
No thanks Meta (Score:2)
How to showcase your creativity? (Score:1)