Abu Dhabi is entering the arena of large language models with its own competitor to OpenAI, DeepMind, and Google. Technology Innovation Institute, a government-backed research hub, has developed Falcon LLM, a model trained on 40 billion parameters with various applications such as chatbots, language translation, content generation, and sentiment analysis. The model is currently not commercially available, but the goal is to eventually offer it to government entities, startups, and the private sector. TII aims to provide a local alternative and reduce dependence on LLMs from major tech players in the competitive and secretive AI space.
TII is a critical part of the UAE’s efforts to diversify from a reliance on oil exports and develop a knowledge-based economy. Falcon LLM has been evaluated by Stanford University, which found it to be a cheaper and faster model than GPT-3 and models from DeepMind and Google. The evaluation also assessed Falcon’s accuracy, bias, and reasoning abilities, with results expected to be made public in the coming weeks.
OpenAI announced GPT-4 on Tuesday but declined to reveal how much bigger the LLM is or why it performs better than its predecessors. Falcon’s announcement serves as a reminder that the development of large language models is a booming market with numerous companies vying for a share.
According to Dr Ray Johnson, CEO of TII, “The year 2023 is turning out to be the year of AI. Falcon LLM is a landmark announcement for us, but this is just the beginning. By the end of the year, we will be sharing news on a huge increase in capabilities in this space.”
About Author
