Meta (NASDAQ: META), formerly known as Facebook, announced on December 6 several new artificial intelligence (AI) features and experiences it has created and integrated into apps and services it owns.
One standalone AI image generator feature in particular, called “Imagine with Meta AI,” seems to stand out above the rest.
Imagine with Meta AI was trained using a publicly visible dataset of 1.1 billion Facebook and Instagram images. This feature previously only existed in the Meta chat app, but its popularity led Meta to expand access to Imagine with Meta outside of chat and make it available through a standalone website, imagine.meta.com.
The app generates images based on written prompts. When the user enters the prompt, the AI generates four images (1280×1280 pixels) in JPEG format. Currently, all images feature a small “Imagined with AI” watermark logo in the bottom left corner of the photo to address concerns about the authenticity of AI-generated content. In the coming weeks, Meta plans to add invisible watermarks to further increase transparency and traceability of AI-generated images.
Meta’s AI products
Imagine with Meta AI will be the flagship product of Meta’s suite of AI products and services, but the company has also announced many other updates across its AI products.
The large language models (LLMs) that underpin Meta AI are now integrated into multiple product experiences to influence search results, deliver AI-generated suggestions, and enhance product copy on Facebook and Instagram stores.
Meta also helps users discover new places when they travel, supports group chats, enhances the shopping experience on Facebook Marketplace, provides suggested answers in direct messages for influencers on Instagram, and more in a variety of scenarios. We are actively testing the integration of AI.
Meta is also reexamining the origins of the ‘Imagine with Meta AI’ product and chatting between users. The company is introducing a feature called ‘Reimagine’ to Messenger and Instagram chat, which allows users to create and share an initial image, and then allows recipients of the image to contribute to the photo by responding with a text prompt, creating a full image. New image.
Finally, Meta is improving upon other AIs accessible through chats on Instagram, Messenger, or WhatsApp by adding a search feature that allows the AI to provide Bing-based responses. Meta also plans to add long-term memory to some of its AIs, allowing the AIs to learn and remember conversations with users. This means users can leave and pick up where the AI left off. In conversation.
Meta AI and generative AI environments.
Meta’s new AI products, especially Imagine with Meta AI, could make Meta a serious competitor in the generative AI space, which is currently dominated by popular service providers OpenAI, Google (NASDAQ: GOOGL), and Midjourney.
“It’s been an incredible year for AI at Meta. We’re introducing new AI experiences across apps and devices, opening access to the Llama family of large-scale language models, and announcing research breakthroughs like Emu Video and Emu Edit that will unlock new features in our products over the next year. I did,” Meta said. presentation. “I’m excited to see what the next year will bring with advances in content creation, voice, and multimodal AI that will enable us to deliver new creative and immersive applications.”
Example: AI is synthetic, not generative.
Are you new to blockchain? To learn more about blockchain technology, check out CoinGeek’s Blockchain for Beginners section, our ultimate resource guide.