At the Meta Connect 2024 conference, it was announced that Meta’s artificial intelligence will now allow you to edit photos using AI technology and will also respond to questions about the photos you share with it.The highly anticipated Orion augmented reality glasses were also introduced, confirming the rumors that had circulated in February of this year, which pointed to the Meta event as the place where this product would finally be officially unveiled.
Additionally, at this event, Meta introduced Meta AI Imagine, which uses generative artificial intelligence to convert text prompts into images. This feature has already started being introduced on Facebook and Instagram. Following the announcement that Instagram creators could have their own personalized AIs, it is clear that Meta continues to invest heavily in artificial intelligence within social media platforms.
It is important to note that this presentation does not mean the glasses will be available for purchase just yet. For now, Meta will continue focusing on internal development to further improve the product until it is ready for sale. In the meantime, access to the prototype will be opened to Meta employees and select external individuals for the remainder of the year, allowing them to provide feedback to the developers.
According to Meta’s statement, the tech company considers Orion to be “the most advanced pair of augmented reality glasses ever made. In fact, it might be the most challenging consumer electronic device produced since the smartphone.”
For their development, various innovative technologies have been used, such as a new augmented reality display created with silicon that enables AR experiences while using a fraction of the energy and weight of an augmented reality headset. Additionally, materials and systems have been employed to reduce the amount of heat generated by electronic devices.
The glasses feature a wide field of view and can project large holographic displays. “With approximately 70 degrees, Orion has the largest field of view in the smallest augmented reality glasses format to date.” This way, the physical world becomes a canvas for interaction, allowing users to place content and experiences in both 2D and 3D.
Furthermore, the Orion glasses integrate contextual AI capable of detecting and understanding what they are focusing on in the physical world, allowing them to proactively anticipate and address the user’s needs.
While Meta has not specified the weight of this device, it has stated that the Orion glasses are “lightweight and excellent for use both indoors and outdoors, allowing people to see the real face, real eyes, and real expressions of others.”
As Meta explains, “the Orion input system seamlessly combines voice, gaze, and hand tracking with an EMG wristband that allows you to swipe, click, and scroll while keeping your arm resting comfortably at your side, enabling you to stay present in the world and with the people around you while interacting with enriched digital content.”
The EMG wristband makes interacting with the Orion glasses remarkably intuitive. For example, you could go for a walk wearing the Orion glasses and take a photo with a simple tap of your fingertips or navigate the device’s menus with slight hand movements.
This wristband “combines high-performance fabric with integrated EMG sensors to detect the electrical signals generated even by the smallest muscle movements. An ML processor in the device interprets these EMG signals to produce input events that are wirelessly transmitted to the glasses. The system adapts to you, becoming more capable of recognizing even the most subtle gestures over time.“
To maintain a compact, manageable, and cable-free design for the Orion glasses, Meta developed a wireless computing disc. While the Orion glasses handle all the eye-tracking, hand-tracking, SLAM (Simultaneous Localization and Mapping), and specialized AR world-locking graphics algorithms, the application logic runs on the wireless computing disc. Thanks to this, the disc “lightens the load on the glasses, allowing for longer battery life and a better design with low latency.”
“The disc features two processors, including one custom-designed right here at Meta, providing the necessary computing power for low-latency graphics rendering, AI, and additional machine perception.” Furthermore, its small and discreet design makes it easy to carry around.
Thanks to the Llama 3.2 models, Meta’s AI is expanding its capabilities. For example, now, if you send an image of a birthday cake, you can ask how to make it, and the AI will provide you with the recipe. Another significant change is the ability to edit photos; for instance, you can send an image of a goat and ask it to place the goat on a surfboard.
Additionally, when sharing a post to your story, the artificial intelligence will “understand” the photo and all its elements, generating a background for your story that matches it.
Mark Zuckerberg’s company is also testing translation tools for Facebook and Instagram Reels, which will include automatic dubbing synchronized with the people speaking in the videos. For now, this feature is only available to small groups in the United States and Latin America, in both Spanish and English.
Meta AI Imagine can help creators generate more engagement with their posts by creating eye-catching images based on the content of their posts. They only need to click on the Imagine button, and the AI will provide a series of options to choose from, although they can also describe the exact image they are looking for.
Meta is also testing the “Imagined for you” option, where it recommends AI creations based on your tastes and interests.
The AI will also provide ideas for creators on what captions to use in their stories and posts. These features will also be available on Messenger, Facebook’s chat app, to create personalized chat themes. While users already have the option to choose from a large gallery of themes, they will now be able to generate their own with Meta’s artificial intelligence.
“Meta AI stands out in this category not only by offering cutting-edge AI models but also by providing unlimited access to these models for free, seamlessly integrated into our various products and apps,” stated Mark Zuckerberg. “So Meta AI is on track to become the most widely used AI assistant in the world by the end of this year. In fact, it probably already is. We are approaching 500 million monthly active users, and we haven’t even launched in some of the largest countries yet,” he added.
Image: Meta
Your email address will not be published. Required fields are marked *
Δ