Generative AI as a Revolutionary Knowledge Technology

Abstract

This case study introduces generative AI as a revolutionary knowledge technology that will reshape science and medicine. Based on the past of paradigm-shifting tools—from writing and mathematics to the transistor—it places generative AI within this tradition as the next driver of discovery. Unlike public large language models, scientific applications of generative AI are highly specialized: researchers fine-tune models on specialist datasets, enabling breakthroughs such as quantum experiment planning, RNA function prediction, and novel material development. In medicine, specialist AI tools such as MedFound have beaten human clinicians in diagnostic accuracy and workflow integration, and companies such as Insilico Medicine are employing AI to accelerate drug discovery and target orphan diseases. The article likened this rapid evolution to the mixed success of applying general-purpose AI models to clinical use cases like patient messaging, where concerns about accuracy persist. In other domains—emergency medicine handoff notes and handheld ultrasounds with AI—AI has already yielded safer, faster outcomes, especially in underserved communities. While the potential to reduce medical error and democratize health care is enormous, the case points out that the ethical, legal, and social ramifications of AI-aided diagnosis and treatment have yet to be resolved. Lastly, it challenges students to consider how much sway AI should have over choices affecting human life.