Meta and Microsoft Launch Llama 2 AI Language Model for Commercial Applications
The speculations about a commercially-focused Meta AI model have been confirmed as true. Meta and Microsoft have joined forces to introduce Llama 2, an advanced AI model with extensive language capabilities designed for both commercial and research applications. The enhanced open source code prioritizes responsibility by subjecting the models to rigorous safety testing through a “red-teaming” process. Additionally, a transparency schematic has been developed to address any potential concerns. The release also includes a guide for responsible usage and an acceptable use policy to prevent misuse such as criminal activities, deceptive representations, and spam.
Meta publishes both pre-trained and conversation-oriented versions of Llama 2. Microsoft makes it available through the Azure AI catalog for use with cloud tools such as content filtering. The tool can also run directly on Windows computers and is available through third-party providers such as Amazon Web Services and Hugging Face.
The original AI model was open source, but intended strictly for scientists and researchers. With Llama 2, companies can customize the technology for their own purposes, such as chatbots and image generators. Ideally, this allows companies to benefit from the modeling while allowing outsiders to check for biases, inaccuracies, and other deficiencies. Open source is not new in the world of artificial intelligence – Stable Diffusion by Stability is a notable example. However, major competitors such as OpenAI’s GPT-4 tend to be locked in due to subscription or licensing revenue.
This is a developing story. Check for updates.
