The authentic startup behind Stable Diffusion has launched a generative AI for video
What’s occurred: Runway, the generative AI startup that co-created final 12 months’s breakout text-to-image mannequin Stable Diffusion, has launched an AI mannequin that may remodel current movies into new ones by making use of types from a textual content immediate or reference picture.
What it does: In a demo reel posted on its web site, Runway exhibits how the mannequin, referred to as Gen-1, can flip folks on a avenue into claymation puppets, and books stacked on a desk right into a cityscape at night time. Other latest text-to-video fashions can generate very quick video clips from scratch, however as a result of Gen-1adapts current footage it may well produce for much longer movies.
Why it issues: Last 12 months’s explosion in generative AI was fueled by the tens of millions of people that bought their fingers on highly effective artistic instruments for the primary time and shared what they made, and Runway hopes Gen-1 can have the same impact on generated movies. Read the complete story.
—Will Douglas Heaven
Why detecting AI-generated textual content is so tough (and what to do about it)
Last week, OpenAI unveiled a software that may detect textual content produced by its AI system ChatGPT. But for those who’re a trainer who fears the approaching deluge of ChatGPT-generated essays, don’t get too excited.