Astrophysicists use Codex to assist black hole simulations: AI accelerates new paradigm of scientific discovery

Astrophysicist Chi-kwan Chan, a member of the Event Horizon Telescope collaboration, used OpenAI Codex to derive and test new algorithms for simulating plasma around black holes. This collaboration demonstrates a new paradigm for AI programming tools to accelerate discoveries in theoretical science and astrophysics.

The behavior of plasma around black holes is extremely complex and has traditionally required weeks or even months of hand coding and debugging. Using Codex, Chan was able to reduce the cycle time for algorithm development and testing to just days. Codex not only generates code, but also helps scientists explore different mathematical derivation paths and numerical methods.

This case is part of OpenAI's "Codex Accelerating Scientific Discovery" series, and Codex has previously been used in areas such as climate modeling, protein folding, and quantum chemistry. OpenAI said that continuous improvements to the Codex are enabling scientists to focus more on the scientific problems themselves rather than programming implementations.

The use of Codex in scientific computing is an underestimated growth area. Different from the traditional "AI writing code" narrative, Codex plays the role of a "computational thinking accelerator" in cutting-edge scientific scenarios such as black hole simulations - it allows scientists to verify more hypotheses in a shorter time. This has implications for the promotion of AI tools in the domestic scientific research field.