March 11, Elon Musk explain xAI will open source its AI chatbot Grok, with an open version now available on GitHub. This will allow researchers and developers to build on the model and influence how xAI updates Grok in the future as it competes with rival technologies from OpenAI, Meta, Google and others.
A company blog post explains that this public version includes “basic model weights and network architecture” for the “314 billion parameter hybrid expert model Grok-1.” It goes on to say that the model comes from a checkpoint last October and has not been fine-tuned “for any specific application (such as conversations).”
as Entrepreneurial Beat Notably, it is released under the Apache 2.0 license, which allows commercial use but does not include data for training it or connections to X for real-time data. xAI said in a November 2023 article that LLM Grok was “developed over the past four months” and is targeted for use in code generation, creative writing and answering questions.
After Musk acquired Twitter (now X) and the code behind its algorithm was eventually released, Musk publicly criticized companies that didn’t open source their AI models. That includes OpenAI, a company he helped found but is now suing over, accusing it of violating its original open source agreement.
Companies have released open or limited open source models to get feedback from other researchers on how to improve them. While there are many fully open source artificial intelligence foundation models, such as Mistral and Falcon, the most widely used models are either closed source or offered with limited open licenses. Meta’s Llama 2, for example, makes its research available for free, but makes its customers, with 700 million daily users, pay, and doesn’t allow developers to iterate on Llama 2.