Llama 3.1: Meta launches open-sourced frontier AI model
Llama 3.1, the latest version of Meta’s open source model family, is said to be able to compete with Open AI’s ChatGPT-4o and Google’s Gemma 2
Much like their more traditional counterparts, tech companies have long engaged in a game of one-upmanship, intolerant of seeing their competition in the news for any sustained amount of time. With OpenAI and Google having dominated much of the conversation around AI, touting their impressive, if still at times glitchy, Large Language Models and chatbots, Meta was seemingly lagging behind.
With the release of Llama 3.1, the latest version of its open source model family, as well as its long-awaited 405B model, the first ever open-sourced frontier AI model, which is said to be able to compete with heavy hitters like Open AI’s ChatGPT-4o and Google’s Gemma 2, Meta is seemingly done playing catch up.
View this post on Instagram
In an open letter accompanying the announcement release of Llama 3.1, Meta founder and CEO Mark Zuckerberg wrote, in part, “Today we’re taking the next steps towards open source AI becoming the industry standard. We’re releasing Llama 3.1 405B, the first frontier-level open source AI model, as well as new and improved Llama 3.1 70B and 8B models. In addition to having significantly better cost/performance relative to closed models, the fact that the 405B model is open will make it the best choice for fine-tuning and distilling smaller models.”
Open source AI is defined as AI technology that is publicly available for both commercial and non-commercial use under various open source licenses. Open source AI tech includes datasets, prebuilt algorithms, and ready-to-use interfaces that can help users get started with AI app development and hit the ground running.
Given that both ChatGPT-4o and Gemma 2 are closed source models, which means their source code is restricted to private use and cannot be built upon or even altered by users, Llama 3.1 could prove to be an existential threat to AI’s big boys, according to experts, with initial tests showing the model match or straight up outperform the former pair.
Given that it is open source and thus mutable, paired with its high performance and capabilities, Llama 3.1 is expected to be a game changer in an industry already mired in so many of them. By allowing app developers and tech enthusiasts the freedom to build upon its structure in new and unique ways, this opens the door to new means of creating and curating all manner of content and operations across industries, not least advertising and marketing.
As Zuckerberg wrote, “Beyond releasing these models, we’re working with a range of companies to grow the broader ecosystem. Amazon, Databricks, and NVIDIA are launching full suites of services to support developers fine-tuning and distilling their own models. Innovators like Groq have built low-latency, low-cost inference serving for all the new models. The models will be available on all major clouds including AWS, Azure, Google, Oracle, and more. Companies like Scale.AI, Dell, Deloitte, and others are ready to help enterprises adopt Llama and train custom models with their own data. As the community grows and more companies develop new services, we can collectively make Llama the industry standard and bring the benefits of AI to everyone.”
