Meta has introduced its latest artificial intelligence system, “Llama 2,” which will compete with other chatbots like OpenAI’s ChatGPT. Unlike its competitors, Llama 2, developed in collaboration with Microsoft, is offered for free to researchers and businesses for both research and commercial use.
Meta and Microsoft aim to “democratize AI and its benefits” by taking an open approach. They provide transparency by allowing access to the data and code used to create their AI models. This open approach fosters innovation and ensures greater safety and security, as more developers can scrutinize the software.
Llama 2 stands out as an open-source AI system, unlike ChatGPT and Bard, which are not open source. Mark Zuckerberg, Meta’s chief, believes that open source drives progress and innovation in the AI ecosystem.
While Meta claims to open-source Llama 2, specific details about the data used to train the AI model remain unclear. The research paper associated with the model states that it was trained on a mix of publicly available data, excluding Meta’s own products or services, and data from websites with personal information about private individuals was removed.
Llama 2 can be accessed directly or through Microsoft’s cloud platform, Azure, making it readily available for developers to utilize and benefit from its cloud-native tools for content filtering and safety features.