Generative AI for Scientific Hypothesis Generation

Design an advanced generative AI system capable of proposing novel scientific hypotheses based on existing research literature and experimental data. Develop a framework that can perform creative knowledge synthesis, identify potential research gaps, and generate probabilistic hypotheses across multiple scientific domains. Include transformer-based language models, creative inference techniques, and uncertainty quantification methods.