Meta Platforms released early versions of its new AI tools, Llama 3 and an image generator, on Thursday. These tools update images in real-time as users type. Meta is trying to catch up with OpenAI, the leading company in generative AI. The new AI models will be part of Meta AI, a virtual assistant that Meta says is the best among free ones. This assistant will be more visible on Meta's apps like Facebook, Instagram, WhatsApp, and Messenger, and on a new website. This move is to compete more with OpenAI's popular ChatGPT, which is supported by Microsoft.
Meta is quickly working to release AI products to its many users to compete with OpenAI, which is currently leading in AI technology. This effort includes updating their computer systems and bringing together their research and product teams into one group. The company's new AI, Llama 3, can now understand computer code and has been trained with images and text. However, for now, it will only produce text outputs, according to Meta's Chief Product Officer, Chris Cox, in an interview.
Chris Cox mentioned that future versions of their AI will be able to make longer plans with more steps. Soon, they will also be able to create both text and images. Meta explained this in blog posts. Cox said the main goal is to make life easier for you, whether it’s dealing with businesses, writing something, or planning a trip.
Chris Cox said that training Llama 3 with images will improve an update planned for this year for the Ray-Ban Meta smart glasses. These glasses are made with Essilor Luxoticca. The update will let Meta AI recognize objects the wearer sees and answer questions about them.
Meta also started a new partnership with Google to add real-time search results to the assistant's answers. This is in addition to an existing deal with Microsoft's Bing.
Meta is expanding its AI assistant to over a dozen new places outside the U.S., including Australia, Canada, Singapore, Nigeria, and Pakistan. They are still figuring out how to launch it in Europe due to strict privacy laws there.
The demand for lots of data from AI models like Llama 3 has caused some problems in the tech world. Meta has made these models free for developers to use, hoping to compete with other companies that charge for their AI technologies. However, some people are worried about the safety of allowing everyone to use these powerful tools.
Meta's CEO, Mark Zuckerberg, said their AI, called Meta AI, is the smartest free AI you can use. He mentioned that the biggest version of Llama 3 is being trained with a huge amount of data and is performing very well. Some earlier versions didn't understand the context well, which led to some mix-ups, but Meta says they've improved this in Llama 3 by using better-quality data and more of it.
Overall, Meta is trying to improve and expand its AI technology around the world, while also working through challenges related to data use and safety.
FAQs
Q1. What is Meta's Llama 3?
Llama 3 is a new AI model from Meta that understands images and text. It helps power an AI assistant that can do things like recognize objects and answer questions.
Q2. Where is Meta's AI assistant now available?
The assistant is expanding to more than a dozen places outside the U.S., including Australia, Canada, Singapore, Nigeria, and Pakistan.
Q3. Why is Meta cautious about launching the AI assistant in Europe?
Europe has strict privacy rules, and new laws are coming that will set more rules for AI. Meta is trying to figure out the best way to meet these rules.
Q4. Why is Meta offering Llama 3 for free?
Meta hopes that offering a powerful AI for free will help it compete with other companies that charge for their AI products. This could stop rivals from making money from their AIs.
Reference
Comments