Unworkshop: ML and the Physical World

On November 11th 2021 we had our first Accelerate Science summit. Alongside keynotes we had unworkshops on different topics. In a discussion about the interaction between machine learning and the physical world, we heard about applications of AI to analyse climate data, agricultural data, clinical data, and biochemical data, with a focus on the challenges that affect practitioners across a range of disciplines and sectors. Steering AI-enabled analysis using insights from domain experts is central to the success of these projects; such insights help define relevant research questions and interrogate the outputs from AI systems. In the process, AI experts play an important role in guiding expectations about what AI can (and can’t) do, and in creating mechanisms for testing and validating the effectiveness of AI systems.

Interactive Map from the 2021 ML and the Physical World Unworkshop