AI for Science jobs

AI for science is the bet that models can compress decades of lab work: protein structure and design, materials discovery, weather forecasting, and simulation surrogates that run in seconds instead of weeks. AlphaFold made the case; the hiring since spans drug-discovery startups, national labs, and the science teams inside frontier labs.

These roles genuinely require both halves — a modeler who can't read the domain literature will build the wrong benchmark. Geometric deep learning, diffusion models over structures, and domain-specific data wrangling fill the weeks. It's one of the few research areas where a PhD in something other than CS is an advantage rather than a detour.