Meta inches toward open source AI with new Llama 3.1

Meta Unveils Llama 3.1: A New Leap in AI Technology

Meta, the powerhouse behind Facebook, has just rolled out the red carpet for Llama 3.1 405B, their latest open-source AI model. This isn't just another update; it's a major step forward in the world of artificial intelligence. Llama 3.1 is set to supercharge Meta's AI assistant, making it a central feature in many of their user-facing applications.

Mark Zuckerberg, Meta's CEO, took to Threads to share the news, emphasizing his belief in the power of open-source AI. "Open-source AI is the path forward," he declared. In a candid chat with Rundown.ai, Zuckerberg elaborated, "Open models are going to be the standard, and I think that it's going to be good for the world. It's a bit soul-crushing when you go build features that are what you believe is good for your community, and then you're told that you can't ship them because some company wants to put you in a box so that they can better compete with you."

The Open-Source Debate: Is Llama 3.1 Truly Open?

But here's the catch: Is Llama 3.1 really open source? This question is crucial, especially for developers and the broader tech community. Neal Gompa, a Linux developer, raised this issue on Mastodon, questioning whether Llama 3.1 is under an OSI-approved open-source license. He pointed out that Llama 2 couldn't be included in Fedora due to its licensing.

Stefano Maffulli, the executive director of the Open Source Initiative (OSI), weighed in on the matter. "In theory, we agree with all that Zuck wrote and said. If only Meta's license would remove the restrictions, we'd be more in sync," he told me. Maffulli further explained, "As it stands now, Llama is a liability for any developer; too opaque to be safe to use and with a license that ultimately leaves Meta in charge of their innovations."

Stephen O'Grady, an industry analyst from RedMonk, echoed these sentiments. "It's welcome news that Meta has dropped some of the use restrictions around Llama, but as long as it still restricts which companies can use the software, as the new license does, it's clearly and inarguably not open source. If Linux, for example, were released under this license, Meta would be entitled to use it, but companies like Amazon, Google, and Microsoft could not. That's not open source, nor would we accept it."

Technical Advancements: What Does Llama 3.1 Bring to the Table?

Setting the licensing debate aside, Zuckerberg is excited about the technical advancements Llama 3.1 brings. He claims the model is designed to be more efficient and powerful than its predecessors, enhancing capabilities across various applications, from natural language processing to complex data analysis. Zuckerberg believes Meta's approach to AI development is superior to Apple's.

Llama 3.1 boasts a 405 billion parameter model, which positions it as one of the most sophisticated AI models available. While it's still smaller than ChatGPT 4.0's 1.8 trillion parameters, it's a significant upgrade. The model is expected to excel in areas like language translation, content generation, and scientific research.

Zuckerberg is confident that Llama 3.1 can hold its own against leading closed models. "By our estimates, it's going to be 50% cheaper, I think, than GPT-4 to do inference directly on the 405B model," he stated. Meta's benchmarking indicates that Llama 3.1 is competitive with top foundation models such as GPT-4, GPT-4o, and Claude 3.5 Sonnet across various tasks.

Availability and Support

The new model is available today through more than 25 partners, including AWS, NVIDIA, Databricks, Groq, Dell, Azure, and Google Cloud. Meta has ensured that Llama 3.1 comes with support for popular AI tools like vLLM, TensorRT, and PyTorch, making it easy for developers to dive right in.

Curious to see what Llama 3.1 can do? Developers can test it out for themselves here.