Meta AI Introduces New ‘Imagine Me’ Selfie Function
Meta AI, the artificial intelligence assistant integrated into Facebook, Instagram, Messenger, and the web, now supports multiple languages and can create stylized selfies. Recently, it began utilizing Meta’s latest AI model, Llama 3.1 405B, designed to address more complex queries compared to its predecessor.
Introduction to Meta AI Updates
The enhancements to Meta AI have sparked discussions about whether they will significantly elevate user experience, especially since early reviews, including comments from TechCrunch’s Devin Coldewey, described the initial performance as disappointing. Users noted that earlier versions struggled with factual recall, numerical data, and basic online searches, often failing to accomplish simple tasks like locating recipes or checking travel prices.
Improvements with Llama 3.1 405B
The introduction of the Llama 3.1 405B model aims to rectify some of these shortcomings. Meta claims this updated model excels in areas such as mathematics and coding, making it particularly useful for tasks like solving math homework, explaining scientific principles, and debugging code.
Despite these improvements, users need to manually switch to the Llama 3.1 405B model to access its capabilities, and they face a limit on the number of queries they can make each week. After reaching this limit, Meta AI will revert to a less advanced version, Llama 3.1 70B.
Generative Selfies through Imagine Yourself
Meta AI features an additional generative model called “Imagine Yourself,” which is used to create selfies. This innovative model allows users to generate images based on a photo of themselves, accompanied by imaginative prompts such as “Imagine me surfing” or “Imagine me on a beach vacation.” Currently in its beta stage, users can engage this feature by typing “Imagine me” followed by their desired scenario.
The training data utilized for the Imagine Yourself model is not explicitly disclosed, but Meta’s guidelines indicate that public posts and images from its platforms can be used, leading to concerns among some users about privacy and data usage.
Editing Features and AI Integration
In addition to the Imagine Yourself capabilities, Meta AI is also introducing new editing tools. These allow users to manipulate images by changing specific objects using prompts like “Change the cat to a corgi.” Starting next month, an “Edit with AI” button will be available, giving users access to advanced editing options. Moreover, Meta AI intends to roll out shortcuts for easily sharing AI-generated images across its platforms, including feeds, stories, and comments.
New Languages and Quest Integration
The updates to Meta AI are not limited to features—there’s also a focus on expanding language support. The Meta Quest VR headset will replace its existing voice command feature with Meta AI starting next month in the U.S. and Canada, allowing users to engage in conversational queries about their surroundings while utilizing passthrough capabilities.
Currently, Meta AI is accessible in 22 countries, including recent additions such as Argentina, Chile, Colombia, Ecuador, Mexico, Peru, and Cameroon. Language support has grown to include French, German, Hindi, Romanized Hindi, Italian, Portuguese, and Spanish, with plans to introduce even more languages in the future.
