Academia Sinica Logo

Abstract:

AI can be used to accelerate and even replace supercomputer simulations in astrophysics, turning complex models into easily accessible tools. I will introduce recent and ongoing work aimed at cosmology, including hybrid AI-physics codes, replacing subresolution modelling with neural networks, and interfacing AI simulations with Large Language Models in multi-agent systems. I will explore two use cases of this type of modelling, first, the Lyman-alpha forest of intergalactic absorption in high redshift quasars and how it can both be gravitationally lensed, and also used to make large-scale maps of the intergalactic radiation intensity. Second, I will talk about astrometric cosmology, where simulation-based mock galaxy catalogs show that it will soon be possible to measure galaxy velocities from their proper motions and also distances (and hence the Hubble constant) from their cosmic parallax. I will show early results from archival HST data, where again AI is able to speed up and improve the analysis.