Meta and Microsoft have joined forces to release Llama 2, a commercially-oriented AI model designed for both commercial and research purposes. The upgraded open-source code places a strong emphasis on responsibility, with developers conducting safety tests and providing transparency schematics to address potential issues. An acceptable use policy has been implemented to prevent misuse, such as criminal activity and misleading representations.
Llama 2 is available for free and comes in pre-trained and conversation-oriented versions. Microsoft is making it accessible through the Azure AI catalog, allowing users to utilize it with cloud tools like content filtering. The model can also run directly on Windows PCs and will be offered through external providers like Amazon Web Services and Hugging Face.
While the original AI model was restricted to academics and researchers, Llama 2 allows companies to customize the technology for their own purposes, such as creating chatbots and generating images. This enables companies to benefit from the model while providing outsiders the opportunity to identify biases, inaccuracies, and other flaws. Open-source code is not uncommon in the AI field, but major competitors like OpenAI's GPT-4 typically restrict access to generate subscription or licensing revenue. However, concerns remain about potential misuse by hackers and bad actors who may exploit open-source tools for harmful purposes.
The focus on responsible use in Llama 2 aligns with similar initiatives seen in models like GPT-4 and Anthropic's Claude 2. The tech industry is actively addressing concerns about the potential risks associated with large language AI models, including the spread of misinformation and ethical considerations. Calls for a temporary freeze on AI experimentation and regulatory measures have been made to ensure developers address these issues. Microsoft's collaboration with Meta to release Llama 2 allows the company to offer more choices to its business customers, particularly those interested in fine-tuning models to suit their specific needs.
Read the full story on: YAHOO! NEWS
