Meta makes AI tool Imagine available separately and comes with a ‘Reimagine’ function
Meta has announced a number of updates to its AI tools. For example, it is introducing a ‘Reimagine’ function for its Imagine image generator, which makes it possible to create additions to generated images. This tool is now also available standalone.
If users in a group chat of one of Meta’s messaging services use the Imagine tool, it will be possible for participants of the conversation to create a new image based on an already generated photo, Meta explains. To do this, they must keep the image in question selected and enter an additional text prompt. A new image is then created from it.
Imagine is based on Meta’s Emu model and was made available in English versions of WhatsApp, Messenger and Instagram in September. However, the tool can now also be used as a standalone via its own website. Initially, the site will only be accessible to users in the US. Meta says it will add an invisible watermark to images generated with this model ‘in the coming weeks’. According to the tech company, this watermark is resistant to ‘common image manipulation techniques, such as cropping, color adjustments and screenshots’.
Meta also promises that it has improved its AI assistant Meta AI. For example, answers on mobile devices are said to be ‘more detailed’, and more accurate summaries of search results are provided. The chatbot could provide a helpful answer to more types of requests. In addition, the AI assistant is now able to share Reels during chat conversations. According to Meta, the features will be rolled out immediately for the English version of its services.
Meta’s Reimagine feature