userscript which strips utm_source from chatgpt.com page links, and adds hotkeys for setting model + reasoning effort, toggling memory
browser extension and userscript for enhancing the sora.com desktop user experience, making some hidden metadata visible
art /+ tech projects using receipt printers via ESC/POS
written interview with graze.social covering inspiration for receipt sky + current thoughts on atproto
a list of Loom implementations which support text completion
a pair of recordings/tutorials demonstrating Loom (Latitude)
on literacy -> ludicity
A collection of thoughts by @visakanv on the topic of "sorting thru your nonsense and turning it into sense."
weekly morning coworking/demos hosted by Justin Bennett in Brooklyn, NY
materials for building -> reinforcing a solid understanding of Janus
running GitHub list cataloguing open-sourced loom implementations, active + inactive
open dataset of self-submitted X/Twitter account archives including likes, (re)posts
multitouch isomorphic controller (chromatic fourths) w/ per-note pitch bend
in-terminal p5.jslike
Go shell util designed to be mapped to [ctrl]+super+shift+v (/ cmd+opt+shift+v) to paste clipboard content wrapped in backticks or <paste> tags
on AI systems as instruments
Links/projects sourced from a creatively-oriented community
retroactive Bluesky timeline indexing utility
webapp (w/ bookmarklet) wrapping an undocumented API that searches for Are.na blocks whose URL matches a submitted full/partial link
Bluesky "browse to search" extension turned hyperlink exploration webapp.
links of note
a model beloved (and eventually, idolized) by many
my chezmoi-managed dotfiles repo. chezmoi's workflow & versatile templating facilitates easy revision + semi-automated deployment of certain config files to my Fedora machines.
GitHub repos of interest to me
niri is a Wayland compositor that took the concept of "scrollable-tiling" from PaperWM (a GNOME extension) and ran with it. niri powers my everyday laptop + desktop setup.
a collection of links, thoughts, takes etc. on / relevant to the topic of AI skepticism (2022-present)
notable (valuable, interesting, amusing, [...]) interactions with LLMs
a channel where I curate(d from late 2023 through much of 2024) media/highlights showcasing and/or helpful in facilitating deep engagement with LLMs at a high level
somewhere around the second or third time I tried to actively read the intro to this post, I realized that language models can simulate ~anything that you can express as text (within the constraint of a model's context window); the following 2 years of my life were deeply influenced by that moment.
Bluesky posts I've liked