‘Project Olympus’ is a recently-announced initiative by Amazon to develop a large language model (LLM) with 2 trillion parameters, which is more than twice the size of OpenAI’s GPT-4, the current state-of-the-art LLM. The project is reportedly led by Rohit Prasad, the former head of Alexa and the current head scientist of artificial general intelligence (AGI) at Amazon. Prasad has assembled a team of researchers from various domains, such as natural language processing, computer vision, and robotics, to train and deploy the Olympus model across Amazon’s products and services.
The ultimate goal of Project Olympus is to create a powerful and versatile AI system that can understand and generate natural language for any task or domain. The model will leverage Amazon’s vast data sources, such as Alexa voice recordings, Kindle books, AWS customer feedback, and more, to learn from diverse and rich linguistic phenomena. The model will also be able to interact with users through conversational interfaces, such as Alexa or Amazon’s own chatbot service. Project Olympus aims to position Amazon as a leader in the AI landscape, especially in the cloud computing market, where enterprise clients demand access to high-performance models.
Image Credit: Shutterstock