Visualizing algorithms is the most fun way to teach, learn, and comprehend them! I had an actual use case for this at work, and it was gratifying to make a fun visual explanation for it.
Pipet is a set of commands to store and retrieve snippets of text. It uses a fuzzy finder for retrieval!
Using a Sequence to Sequence model to generate punchlines for fake jokes.
Paradigm Shell is an experimental game that explores crude oil as a transformative, vibrant, time-condensing form of matter. Through recalculating vertices in vertex shaders virtual materials morph and mutate.
This neural network, trained on the works of T.S. Eliot, generates new poetry based on his work. It also raises the question of how much an artist's work--and the new work that comes from it--preserves them once they're gone.
I try to maintain a sense of wonder throughout my process of learning CS. In this blog post I wrote something a bit more philosophical, drawing on training as a musician to try to provide an answer to why seemingly inefficient algorithms can still be interesting and compelling objects of study. (Image is CC BY-NC-ND 2.0; source cited in post.)
Producing fractals with Postgres. What isn't there to like?
A Clojure library that helps you test specs for breaking changes.
An oldschool-bbs-like content engine that runs over telnet! Live demo available at escape-sequence.net 4444 (use the `telnet` command in your favourite terminal emulator to connect).
Filippo does great work trying to demystify cryptography and improve implementations of cryptographic algorithms, and this is a nice example.