Over 115 years of superconductivity history, new superconductors have almost always been discovered serendipitously, or by extrapolating from existing materials classes. However, recent progress in first principles computational superconductivity, high-throughput materials discovery techniques, and artificial intelligence have led to promising efforts to predict new superconductors theoretically. I will first review quickly some attempts in this direction relying on the use of empirical databases to train simple neural networks. Following this path, one quickly discovers that even in the case of electron-phonon superconductors, where the underlying theory is well-controlled, superconductivity is much more challenging to predict than other materials properties. I will then focus on several related efforts to predict new electron-phonon superconductors based entirely on first principles calculations of the Eliashberg spectral function. This program achieved a modest success, with the prediction and empirical discovery of two new low-Tc superconductors, but also identified fundamental problems inherent the approach, whose solution requires greater attention to challenges of synthesizability of materials[1,2]. Finally, I will describe initial efforts to extend the deep learning methods to the discovery of higher critical field superconductors needed for lower Tc applications [3].

1. ``Developing a Complete AI-Accelerated Workflow for Superconductor Discovery", J. B. Gibson, A. C. Hire, P. M. Dee, B. Geisler, J.-S. Kim, Z. Li, J. J. Hamlin, G. R. Stewart, P. J. Hirschfeld, and R. G. Hennig, npj Comput Mater 12, 95 (2026).
2. ``Guided Diffusion for the Discovery of New Superconductors", P. Prakash, J. B. Gibson, Z. Li, G. Di Gianluca, J. Esquivel, E. Fuemmeler, B.Geisler, J.-S. Kim, Adrian Roitberg, E. B. Tadmor, M. Liu, S. Martiniani, G.R. Stewart, J. J. Hamlin, P.J. Hirschfeld, and R. G. Hennig, arXiv:2509.25186.
3. "Towards the discovery of high critical magnetic field superconductors", Benjamin Geisler, Philip M. Dee, James J. Hamlin, Gregory R. Stewart, Richard G. Hennig, P. J. Hirschfeld, arXiv:2601.21044.