Open Source goals for 2022/2023

Let's review the past year briefly. It has been fruitful:

• The projects I am a part of (the Aesara ecosystem, Blackjax, Pollsposition) have made sensible progress and are gaining momentum.
• Working in close proximity with great people, including but not only (in handle alphabetical order) @BrandonTWillard, @dgerlanc, @jeffreydenos and @junpenglao, has made me a better developer.
• Thanks to @BrandonTWillard I have now fully transitioned from vim to emacs.
• I have picked up Racket;
• I learned about relational programming, in particular miniKanren.
• My thoughts on efficient note-taking, knowledge sharing and what a "better Internet" would look like are getting clearer, thanks to the typical set project, which I set in motion this year.
• I predicted the results of the French presidential elections with @alex_andorra and @pollsposition (although a serious injury in February prevented me from doing more). I went on TV the night of the election!

Here are the goals that I would try to achieve this coming year:

• AeMCMC can build samplers with 99% of models encountered in the wild;
• AeHMC’s NUTS sampler’s performance is on par with Blackjax’s;
• Aesara uses e-graphs and equality saturation for its graph rewrites;
• Blackjax’s V1 is released;
• The Blackjax paper is ready for submission;
• miniKanren’s python implementation solves sudoku grid almost as fast as the imperative solution;
• Implementation of something similar to WaveFunctionCollapse with miniKanren;
• The typical set is self-hosted on the most energy efficient stack possible.

And more generally improve thetypicalset’s design, navigation and overall content quality. I would like to learn some Rust and learn more about internet protocols (why not implement a HTTP server in Rust?) if time allows.