Opening Llama 3.1: An Open Source Artificial Intelligence Model for Generative AI and AI Assistive Processes

Today, that model arrived. Meta is releasing Llama 3.1, the largest-ever open-source AI model, which the company claims outperforms GPT-4o and Anthropic’s Claude 3.5 Sonnet on several benchmarks. Adding a feature that can generate images based on someone’s specific likeness is one of the ways it is making the llama-based Meta Artificial Intelligence assistant available in more countries and languages. CEO Mark Zuckerberg now predicts that Meta AI will be the most widely used assistant by the end of this year, surpassing ChatGPT.

In order to improve the smaller 70 billion and 8 billion versions of llama 3.1, Meta used synthetic data, instead of using humans, to generate the data. Ahmad Al-Dahle, Meta’s VP of generative AI, predicts that Llama 3.1 will be popular with developers as “a teacher for smaller models that are then deployed” in a “more cost effective way.”

The ceiling may be farther out than some think, though Meta agrees with the growing consensus that there is no quality training data for models. “We definitely think we have a few more [training] runs,” he says. “But it’s difficult to say.”

For the first time, Meta’s red teaming (or adversarial testing) of Llama 3.1 included looking for potential cybersecurity and biochemical use cases. Another reason to test the model more strenuously is what Meta is describing as emerging “agentic” behaviors.

For example, Al-Dahle tells me that Llama 3.1 is capable of integrating with a search engine API to “retrieve information from the internet based on a complex query and call multiple tools in succession in order to complete your tasks.” Another example he gives is asking the model to plot the number of homes sold in the United States over the last five years. “It can retrieve the [web] search for you and generate the Python code and execute it.”

Meta’s own implementation of Llama is its AI assistant, which is positioned as a general-purpose chatbot like ChatGPT and can be found in just about every part of Instagram, Facebook, and WhatsApp. The first places where it will be accessible in the United States are through Metaai.com, and on the US version of Facebook. New languages to be supported include French, German, Hindi, Italian, and Spanish.

A new feature in MetaAI lets you insert your likeness into images it creates through a facial Scan on your phone. Meta is hoping to avoid the creation of a deep fake machine if it captures your likeness this way and not through photos in your profile. The company sees demand for people wanting to create more kinds of Artificial Intelligence media that can be shared with others, even if it means keeping the line between what is discernibly real and not.

Meta will be coming to the headset with a new voice command interface. Like its implementation in the Meta Ray-Ban glasses, you’ll be able to use Meta AI on the Quest to identify and learn about what you’re looking at while in the headset’s passthrough mode that shows the real world through the display.

Meta has yet to give any usage numbers for its assistant despite a prediction from Facebook that it will be the most-used chatbot by the end of the year. “I think the entire industry is still early on its path towards product market fit,” Al-Dahle says. Meta and others believe that the race is just beginning, despite how much overhyped it is.

Most tech moguls hope to sell artificial intelligence to the masses. Meta thinks Mark Zuckerberg has one of the best artificial intelligence models on the planet for free.

The new model was compared to the open source Linux operating system by Meta’s CEO in an open letter. When Linux took off in the late 90s and early 2000s many big tech companies were invested in closed alternatives and criticized open source software as risky and unreliable. Cloud computing uses Linux as the core of theAndroid mobile OS.