Member-only story
First Llamas now Ferrets.
Apples Enter The World of LLMs
In October, Apple and Columbia University quietly introduced Ferret, a large open-source multimodal language model (LLM) for research purposes. It has however had renewed attention following the release of Google’s Gemini as people then started to understand just what Apple and the iPhone capable of.
Apple has published a couple of research papers unveiling techniques for efficient language model interfaces and 3D avatars, suggesting advancements that could enhance user experiences and enable complex AI systems on small devices.
So, what does Apple’s Ferret do exactly? It’s apparently capable of interpreting both images and text, facilitating detailed queries within specific areas of an image. I’m hoping to take a look and having a play with it this weekend and will post an update.
Apple’s decision to make this model open source is intriguing, considering its historical research secrecy. Commentators have suggested this might be and attempt by Apple to highlight a ethical approach to AI.
