GORILLA AI: Meet the First Genuine Proximate AGI (By Microsoft)

Microsoft and UC Berkeley have recently developed a new AI model called GORILLA AI, which they claim is the first genuine proximate AGI. GORILLA AI is able to autonomously interact with online tools using API calls, and it can learn from online sources in real-time, adapt to changes, and handle complex tasks across different platforms.

GORILLA AI is built on an improved version of the Llama 7B language model, and it is fine-tuned with a dataset called API bench. This dataset contains thousands of special commands known as API calls, which help GORILLA AI understand and work with different tools to produce the right results.

In addition to its ability to generate API calls, GORILLA AI also has a system that fetches documents from the internet or other places. This allows GORILLA AI to learn about new tools and commands, even if they are not included in the API bench dataset.

GORILLA AI is still under development, but it has the potential to revolutionize the way we interact with computers. It could be used to automate tasks, develop new applications, and even create new forms of art and entertainment.

If you are interested in learning more about GORILLA AI, you can visit the Microsoft website or the UC Berkeley website. You can also try out GORILLA AI yourself by using their collab notebook or by installing their CLI tool.