From OpenAI to Opensource AI: Navigating Between Commercial Ownership and Collaborative Openness
Let’s delve into the complexities of the evolving field of Generative AI and its intersection with issues of commercial ownership and collaborative openness. We will explore the trajectories of various models, including GPT-4, and analyze the dynamics between the interests of commercial entities and the considerations of open collaboration. I will provide insights into the levels of openness of different Large Language Models (LLM), evaluating various components and aspects. Additionally, we will discuss the potential impacts of the (de)centralization of computing power and technology on the future of Generative AI research and development. Using concrete examples such as LLaMA and its descendants, as well as other open and collaborative projects, we will illustrate the diversity and creativity in the field while navigating the complex waters of intellectual property and licensing. Familiarity with LLMs and their significance in Generative AI would be beneficial but not mandatory. By attending this session, you will: - Learn about the levels of openness of different Large Language Models and the factors that contribute to their categorization. - Understand the complexities of intellectual property and licensing in the context of open and collaborative Generative AI projects. - Gain insights from concrete examples of open and collaborative initiatives, including well-known and lesser-known initiatives, to illustrate the diversity and creativity in the field. There is a lot of confusion about open (source) LLM, this talk is about shedding some light on that matter.