Social media giant Meta has added its latest AI assistant to nearly all of its apps, co-founder and CEO Mark Zuckerberg announced Thursday.
“We’re launching a new version of Meta AI, an assistant that you can ask any question across apps and glasses,” Zuckerberg said in a video posted to Instagram. “Our goal is to build the world’s best AI and make it available to everyone.”
This news comes as Meta has released the core components of Llama 3 under an open source license, making it available for the public to use and review.
After its Metaverse ambitions fell through in late 2022, Meta shifted its focus and dove hard into generative AI. Rumors have begun to circulate that Meta will release its Llama 3 generative AI model in May. The company began publishing early Thursday.
“The bottom line is that we believe Meta AI is the most intelligent AI assistant that is now freely available,” Zuckerberg said.
Meta AI uses real-time data from Google and Bing to respond, Zuckerberg said.
Starting today, users who log in to the Meta app will see the new Meta AI search bar added for free to WhatsApp, Instagram, Facebook and Messenger. Meta’s Twitter competitor Threads was notably absent from the announcement. In addition to social media applications, Meta AI is available to Facebook users who log in through the Meta AI website.
Zuckerberg said both the desktop and mobile versions can produce high-quality images. Once your image is created, you can also animate it with a short 3-second clip.
The desktop version of the Llama 3 image generator begins composing images as you type a prompt, allowing you to immediately preview what the final photo will look like.
Although his video announcement was aimed at mainstream app users, Zuckerberg commented on his decision to open source the Llama 3 AI model, saying it was the best way to create a better and safer product.
“We are investing heavily to build leading AI, and responsibly open sourcing our models is a critical part of our approach,” Zuckerberg said. “The technology industry has shown time and time again that open source leads to safer and more secure products, faster innovation, and healthier markets. “
AI-generated deepfakes are an ongoing and expanding problem, which is only made worse by the constant cycle of new and updated generative AI models hitting the market.
In February, the Biden administration said it would use watermarking and encryption to tag authentic content and combat political disinformation.
To get ahead of Llama 3, which is used to create deepfakes, images created using this tool include an “Imagined with AI” disclaimer at the bottom of the photo. Meta pointed out that its AI development team includes guardrails to detect messages that go against company policy, such as asking how to commit a crime.
Meta did not immediately respond to a request for comment. decryption.
Edited by Ryan Ozawa.