1. Working on a 'production ready' version of Conal Elliot's 'compiling to categories' for GHC.
2. This is so I can create a vectorizable model of a datalog-based query language I'm building in Haskell.
3. The query engine will be using a version of monadic optimization as outlined on a blog post somewhere
4. The purpose of the query engine is maintenance of large datasets, all the more important with AI these days, but really general purpose.
5. The motivation for this was a low code tool I had built in Haskell almost a decade ago that I abandoned that I'm bringing up to use ghcs web assembly backend and I need a proper query engine for it now.
Other things:
1. Thinking about binary neural networks and how to train them stochastically.
2. Learning about finite element methods for physical modeling and also reviewing my basic topology so I can think more about non discrete math and algebra which I tend to focus on.
3. I'm building a cloud chamber! Because I want to see space particles. Literally for no other reason than I'm obsessed with these devices ever since seeing one at the exploratorium
4. And raising three kids. I don't know how I have time for anything
I just make incremental progress on a daily basis and don't consider stopping work for a week or two quitting. I work in spurts nd focus entirely on one project a day.
1. Working on a 'production ready' version of Conal Elliot's 'compiling to categories' for GHC.
2. This is so I can create a vectorizable model of a datalog-based query language I'm building in Haskell.
3. The query engine will be using a version of monadic optimization as outlined on a blog post somewhere
4. The purpose of the query engine is maintenance of large datasets, all the more important with AI these days, but really general purpose.
5. The motivation for this was a low code tool I had built in Haskell almost a decade ago that I abandoned that I'm bringing up to use ghcs web assembly backend and I need a proper query engine for it now.
Other things:
1. Thinking about binary neural networks and how to train them stochastically.
2. Learning about finite element methods for physical modeling and also reviewing my basic topology so I can think more about non discrete math and algebra which I tend to focus on.
3. I'm building a cloud chamber! Because I want to see space particles. Literally for no other reason than I'm obsessed with these devices ever since seeing one at the exploratorium
4. And raising three kids. I don't know how I have time for anything