Some people do things where performance is crucial. This is probably why Intel is moving into the FPGA space and Google is making their own accelerators for machine learning.
The very high-performance things are a relatively narrow area, and an area where C can easily lose to e.g. Fortran.
Also, for ML stuff you likely don't have a huge entrenched C codebase yet. You are freer to choose a more expressive language, and only hand-optimize / rewrite in C a small amount of hot paths.
There was no other comparable language to implement it in besides Fortran. C was a newcomer at the time, like Rust currently is.
The amount of optimization and edge-case handling work that went into BLAS since then is enormous.
BTW, Fortran, being mostly aliasing-free, is a safer language than C, and more optimizaton-friendly.
Maybe if you use recent, powerful machines and don't particularly care about energy efficiency. In my experience with less powerful machines, the applications which tend to be noticeably slow are exactly the widely used 'general user applications': of the ones I use, Firefox, Chromium, LibreOffice and Thunderbird.