In an effort to offer startups and other companies a low-cost opportunity to compete with OpenAI’s ChatGPT and Google’s Bard, Mark Zuckerberg’s Meta is making a commercial version of its artificial intelligence model freely available.
Microsoft will distribute a new version of a Meta large language model (LLM), dubbed Llama 2, over its Azure cloud service and it will work with the Windows operating system, according to Meta, which termed Microsoft “our preferred partner” for the release. Although the proprietor of ChatGPT has not made its LLM, GPT-4, open-sourced or publicly accessible to others, LLMs are the foundation of generative AI products like the ChatGPT chatbot.
The model, which Meta has previously made available to a restricted group of academics for research, will also be made accessible through direct download, Amazon Web Services, Hugging Face, and other service providers. In a Facebook post, Zuckerberg stated that open source drives innovation because it enables many more developers to build with new technology. If the ecosystem were more open, I think more progress would be made, he said.
Executives at Meta claim that by using the wisdom of the community to uncover issues and strengthen system resilience, public releases of technology actually lower safety concerns. Additionally, the business claims to have established an “acceptable use” policy for commercial Llama that forbids certain use cases, such as acts of violence, terrorism, child exploitation, and other illegal activities.
Making a model as complex as Llama widely accessible and free for companies to build on poses a threat to the early market dominance achieved by players like OpenAI, whose models Microsoft supports and which it already makes available to business customers via Azure.