AI continues to attract significant attention across a broad frontier of application areas of commercial interest motivating a massive investment in research and development. Applying AI to scientific problems has also been gaining momentum. Applying AI to science can be challenging as ‘hallucinations’ undermine any potential benefits, AI results can be hard to interpret, extrapolation is not accurate, and many science problems have existing high quality solutions. Considering these limitations, I will illustrate the application of AI to science with examples from weather and climate.

About the speaker

Image

In Australia, John's research interests have revolved around the measurement of demographic change among Indigenous peoples and assessment of their economic status at varying scales of analysis from the local to the regional and national. Increasingly, this also incorporates international comparison with North America and New Zealand, particularly in terms of demographic analysis. A basic tenet of his research is the need to establish key parameters of population change as the basis for evaluating policy impacts in areas such as employment, housing, education and health. He has published widely on these issues and is the author of numerous consultancy reports for government, industry and Indigenous organisations.