A personal note

As a kid in the eighties I read a lot of popular science, and somewhere in there I came across the finite element method. It seemed like magic: tile the world into triangles and tetrahedra, and you can compute whatever you are interested in, whether that is how airplanes fly, how dams hold or how heat spreads. That idea steered my studies, pulled me into numerics early on and ended up as a PhD in the early 2000s.

By then we had the tools, but running them and analysing the results was tedious. And there was a gap between what we used for prototyping (MATLAB) and what we used for production runs (Fortran/C). Together with Marcin Dabrowski and Marcin Krotkiewski at PGP we closed that gap: MILAMIN was native MATLAB and fast.

Today the MilAMin presented here runs in a browser, on the phone in my pocket, faster than it ever did on the desktops of 2008: finite element simulations of how inclusions behave, how folds form or whatever else you want to understand. You can take this stack and have an AI model tailor it to your own problem. If you build something worth sharing, tell me and I will gladly add it to the apps on this site. The old barriers are gone: special computers, special compilers, know-how locked away in a few groups. These tools now belong to everybody.

What seemed like magic to me as a kid is finally real, and it changes what matters. When everyone has access to models, the edge no longer lies with whoever builds the most complex or most realistic one. It lies with the people who know what to ask of a model and how to read its answer: geologists who understand geology. So if you are one, go out and look at the rocks. Try to understand what they are telling us, and use these tools to test it.

Dani Schmid, September 2026, [email protected]