Astrophysicist uses Codex to improve black hole simulations

Astrophysicist Chikwan Chan at the University of Arizona is using Codex to help develop better algorithms for simulating black holes and the plasma around them. Chan works with the Event Horizon Telescope collaboration, which produced the first black hole image in 2019 and is now gathering data aimed at creating the first video of a supermassive black hole, including the one at the center of the M87 galaxy. The challenge is that current simulations struggle to model the highly complex motion of charged particles near black holes. In some regions, the plasma is so hot and diffuse that particles rarely collide, making standard fluidbased approaches less accurate. Chan says Codex helped his team explore new mathematical methods, generate candidate schemes, and test them against known solutions. The approach is meant to make simulations more efficient without relying on blackbox answers. Chan says scientific ideas must still be verified through repeated testing, but AI could help researchers try more possibilities faster and study black hole physics at a higher level of detail.