Instagram rolled out its new AI-generated background editing tool to U.S.-based users. The feature, powered by generative AI, allows users to transform the backgrounds in their Instagram stories and photos through text prompts.
The background editor is being spearheaded by Ahmad Al-Dahle, Meta’s lead for generative AI. In an announcement post on Threads, Al-Dahle said the tool will enable users to prompt changes to their story backgrounds, opening up creative new possibilities.
“With backdrop, you can reimagine your image’s background with just a few taps and a prompt like ‘chased by dinosaurs’ or ‘surrounded by puppies’ to create an entirely new image for your Story. Tap the button for a backdrop at the top of a new Story to get started,’” Al-Dahle wrote. Users can also write their own custom prompts.
Now you can try the Instagram “Try it” stickers
Once posted to stories, the AI-generated backgrounds will feature a “Try it” sticker so other users can test out the same text prompts on their own images.
The launch builds on Instagram’s growing suite of AI offerings, including creative tools already available on Snapchat and other Meta platforms like Messenger and WhatsApp. Earlier this month, Meta rolled out a similar AI image generator called Imagine, while Snapchat unveiled paid features enabling users to create fantasy-themed photos.
As AI image generation technology advances, social platforms are racing to bake the functionality into their apps. Instagram hopes features like the background editor will set it apart, though ethical concerns around deepfakes continue to swirl around AI-powered creative tools.
Generative AI refers to a type of artificial intelligence capable of creating new content on its own instead of merely analyzing existing content. The “generative” part refers to the fact that these AI systems can generate brand-new images, text, audio, video, and other data that have never existed before.
Generative AI relies on deep learning algorithms, specifically on a subset of AI approaches known as neural networks. To learn patterns and concepts, these neural nets are trained on vast datasets – sometimes containing millions or billions of examples. Once the generative AI model has been sufficiently trained, it can use its learnings to produce novel, original creations based on what it has detected within its training data.
Featured Image: Photo by Pixabay; Pexels
