Meta has introduced AI-powered assistants to Facebook and Messenger.
Meta AI, an AI assistant powered by the "most powerful open source model" LLaMA 3, is now available on WhatsApp, Instagram, Facebook, and Messenger platforms. Meta announced on April 18 that Meta AI is now available for free on the company's essential products and operates independently on a separate website, meta.ai. However, the deployment is currently available in only about ten countries in Europe and the US, and not yet in Vietnam.
Mark Zuckerberg, CEO of Meta, posted on Instagram and mentioned that Meta AI is easier to use than other similar products due to its built-in integration into the search engine in a series of applications with hundreds of millions to billions of users. One of the features mentioned by Zuckerberg is the ability to create images from text in real-time. "LLaMA 3's integrated Meta AI can create high-quality photos so quickly that changes will appear with each keyword the user enters. The AI will also create a video documenting that creative process," he said.
LLaMA 3 is a large language model (LLM) recently released by Meta with two training versions on 8 billion and 70 billion parameters. The company describes this LLM as significantly reducing false response rates, improving alignment, and increasing response diversity. "With LLaMA 3, we aim to build the best open-source model today, on par with the strongest proprietary models," Meta said. LLaMA 3 will be a multilingual and multimodal LLM, handling longer contexts and improving overall performance.
Both versions of LLaMA 3 are integrated into Meta AI. The company is reportedly working on another version with 400 billion parameters.
"With the new model, we believe Meta AI is now the smartest AI assistant that people can comfortably use," said Zuckerberg.
LLaMA is an AI model introduced by Meta in February 2023, when the AI craze was high, and it is currently in its third generation. LLaMA 3 is entirely open source, while LLaMA 2 requires companies with more than 700 million monthly active users to apply for a license.
Mark Zuckerberg, CEO of Meta, posted on Instagram and mentioned that Meta AI is easier to use than other similar products due to its built-in integration into the search engine in a series of applications with hundreds of millions to billions of users. One of the features mentioned by Zuckerberg is the ability to create images from text in real-time. "LLaMA 3's integrated Meta AI can create high-quality photos so quickly that changes will appear with each keyword the user enters. The AI will also create a video documenting that creative process," he said.
LLaMA 3 is a large language model (LLM) recently released by Meta with two training versions on 8 billion and 70 billion parameters. The company describes this LLM as significantly reducing false response rates, improving alignment, and increasing response diversity. "With LLaMA 3, we aim to build the best open-source model today, on par with the strongest proprietary models," Meta said. LLaMA 3 will be a multilingual and multimodal LLM, handling longer contexts and improving overall performance.
Both versions of LLaMA 3 are integrated into Meta AI. The company is reportedly working on another version with 400 billion parameters.
"With the new model, we believe Meta AI is now the smartest AI assistant that people can comfortably use," said Zuckerberg.
LLaMA is an AI model introduced by Meta in February 2023, when the AI craze was high, and it is currently in its third generation. LLaMA 3 is entirely open source, while LLaMA 2 requires companies with more than 700 million monthly active users to apply for a license.
Share