mfld

1. There is a large corpus of scientific publications.

2. Thanks to scihub et al, it is possible to download them en masse.

3. It is now feasible to train good language models on a large text corpus.

4. Wouldn't it be interesting to see what a model trained on this data comes up with? Maybe even new connections in existing knowledge?

The need to log in to be able to use the terminal feels a bit awkward.

Reading the article, his track "30hz" was the first thing that came to mind. :)

> ...the pre-flight planning requires simulation of the complete flight path from launch site to target. The resulting flight path plan as well as the algorithm for course adjustments based on various inputs is loaded by the programmers onto a ruggedized memory stick, which is then passed on to the launch location and inserted into the missile.

Fascinating.