Mark Zuckerberg reveals details of his AI: Llama 3

The CEO believes that the future of artificial intelligence lies in smart glasses. (Instagram: Zouk)

Meta CEO Mark Zuckerberg announced that the company is developing and improving its conversational AI model known as Llama 3.

This language model in artificial intelligence is an advanced system that processes and Understands human languageAllowing computers to communicate with people in a more natural and effective way.

This type of technology has become essential in the development of virtual assistants and in improving the interaction between humans and machines.

Zuckerberg emphasized that “the next generation of services requires building complete intelligence” (Instagram: zuck)

“It is increasingly clear that the next generation of services requires building complete intelligence. Create the best AI assistants for creators and businesses. Advances are needed in every area of ​​AI, from thinking to planning, coding, memory and other cognitive abilities.Zuckerberg said on Instagram.

Llama is the name of an artificial intelligence language model developed by Meta. It works by processing large amounts of text to learn how human language is organized and how to create coherent and relevant dialogue.

At the beginning of 2023, Llama 1 was launched, which at the time was considered the new horse in the AI ​​race by Zuckerberg. Then, in the middle of the same year, the company released Llama 2.

This last generation was open source and free for commercial use. This was done using 40% more data than the original versionThe chatbot is designed to generate results on par with OpenAI's ChatGPT, Meta said.

Llama 2 launches in July 2023. (Meta)

The year 2023 ended with several rumors about the new version of the Meta language model. And now, at the beginning of 2024, Zuckerberg himself confirms that Meta is entering Llama 3.

See also  Canada launches VACANCE as a Cleaning Supervisor with a salary of 60,000 pesos per month | requirements

To support this initiative and other future projects, the company is developing a significant IT infrastructure. This means purchasing 350,000 new H100 processors by the end of this year.

These processors are a class of chips specifically designed for large-scale data processing and artificial intelligence tasks.

It is part of the graphics processing unit (GPU) family and is characterized by its high efficiency and ability to handle complex algorithms.such as those used to train linguistic models in artificial intelligence.

Meta's computational scale grows to support Llama 3, with massive acquisition of advanced chips for AI development (Image: Infobae)

The reference to the number of these processors indicates the amount of computing power that Meta is deploying to develop and improve the Llama 3 model.

“People will also need new AI hardware, and this unites AI and the Metaverse,” Meta's CEO said. At this point, he pointed to Ray Ban Meta glasses, which have become very popular on social networks.

Meta has integrated its advanced conversational assistant, Meta AI, into smart glassesand optimized for a hands-free experience adapted to mobility.

New augmented reality glasses will allow live broadcasting.

By saying “Hello Meta,” users can interact with the chatbot to stimulate creativity, obtain information, and manage functions through voice commands.

In addition, The glasses have the ability to broadcast directly to Facebook or Instagram. It also displays comments in the preview.

“The glasses are ideal for letting the AI ​​see what you see and hear what you hear, so it's always there to help. Ray Ban glasses with Meta AI have had a very strong start, and overall, with all this stuff, we're just getting started.

See also  He quit his job and started a business: now he works 3 hours and earns more than $1.7 million a year | mix up

Myrtle Frost

"Reader. Evil problem solver. Typical analyst. Unapologetic internet ninja."

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top