Speed the read write speeds of every server and workstation at work by introducing on the fly compression. I hypothised it would work because we had a surprising amount of idle cpu and even where we didn’t a lot of the clock cycles were waiting for IO. Had to design some benchmarks, run on test systems, do a blue field deployment on a few prod servers, etc, but it was really rewarding to go from idea to implementation. It’s was actually noticble to users for start up time too! It lead to a more aggressive compression on some backups that ended up delaying the need for hardware purchases too.

Oh true! It’s funny talking some data sci folks that are just “I didnt need GPUs but now I might as well optimize for them now” because they’re org has tons of that compute now for “AI”

AI development has a few factions. People that want to make and fuck their AI daughter/girlfriend. Researchers that want to underatand the basis of the human mind. Capitalists that want to convert human intelligence into capital they can fully own (slavery is the first attempt). Then finally cultists that want to make God in their image.

It’s a crazy field for someone that is fascinated by thinking about thinking and want to make useful little tools.