The Centre for the Mathematical Foundations of Generative AI (CMFAI) aims to advance the mathematical foundations and applications of generative AI. Generative AI is a branch of artificial intelligence that focuses on creating new content, such as text, images, audio, video, code, and simulations, using models and algorithms that learn from large amounts of data based on statistical principles. Some examples of generative AI tools are ChatGPT, DALL-E, Midjourney, and Bing Chat.
The rationale for setting up the proposed CMFAI is based on the following premises:
Generative AI has enormous potential to transform various domains of human activity, such as education, health care, science, and engineering. It can enable new forms of creativity, innovation, communication, and problem-solving. By developing novel mathematical and statistical methods and models for generative AI, the centre can contribute to the innovation and competitiveness of these fields.
Generative AI poses significant technical and ethical challenges that require rigorous mathematical analysis and solutions. For instance, generative AI algorithms can produce inaccurate, biased, or harmful content that can mislead or harm users. The centre can address these issues by developing mathematical and statistical frameworks and metrics for evaluating and improving the quality, fairness, and safety of generative AI outputs.
Mathematics plays a crucial role in generative AI, as it provides the theoretical framework and the computational tools for designing, analyzing, and optimizing generative algorithms. Mathematics and statistics also help to understand the properties and limitations of generative AI models, such as their expressiveness, complexity, robustness, and interpretability.
Indeed, generative AI can benefit from the insights and techniques of mathematics and statistics. For example, generative AI can leverage optimization algorithms and statistical learning methods such as reinforcement learning and causal inference to create more meaningful, diverse, and complex content. The centre can explore these connections and foster interdisciplinary collaborations between mathematicians and AI researchers.
Objectives
The objectives of the centre are:
To develop new mathematical and statistical methods and tools for analyzing, designing and optimizing generative AI models.
To explore the ethical, social and legal implications of generative AI and its impact on various domains and industries.
To collaborate with academic and industrial partners to apply generative AI to solve real-world problems and create innovative products and services.
To educate and train the next generation of researchers and practitioners in generative AI and its related fields.
The CMFAI will be a leading research centre that will contribute to the advancement of knowledge and innovation in generative AI. It will also be a hub for fostering a vibrant and responsible generative AI community that can benefit society at large.