Menlo Park, California – Meta Platforms, the parent company of Facebook, has thrown its hat into the increasingly competitive AI ring with the launch of Llama 3.1, a new large language model touted as “state-of-the-art” by CEO Mark Zuckerberg.
The model, developed at a significant cost of hundreds of millions of dollars and months of training, is designed to rival similar offerings from industry leaders OpenAI and Google. It represents a significant upgrade from its predecessor, Llama 3, released in April.
Llama 3.1 boasts a range of advanced capabilities, including enhanced problem-solving skills and the ability to generate images based on text prompts. A standout feature, “Imagine Yourself,” allows users to transform their own photos into various digital scenes.
Meta plans to integrate Llama 3.1 into its chatbot, Meta AI, which is already accessible within platforms like Instagram and WhatsApp. The company aims to make Meta AI the world’s most widely used chatbot by the end of the year.
Zuckerberg envisions Llama as a foundational tool for other companies to build their own AI models, fostering a more open and competitive AI ecosystem. To that end, Meta has made the technology freely available to the public, subject to the company’s guidelines.
However, the company is keeping the specific details of the data used to train Llama 3.1 under wraps. While Meta acknowledges using publicly available Facebook and Instagram posts, it maintains that proprietary data sets also play a crucial role in the model’s development.
The massive investment in AI aligns with Meta’s broader strategy, but it also reflects the intense competition in the industry. Zuckerberg acknowledged the risk of overspending on AI development but emphasized the potential consequences of falling behind in this critical technology.
