the scrape

2025-07-07 · thread, 8 tweets · mirrored from twitter ↗

im personally responsible for the original scrape of the private ebooks tracker bibliotik that eventually was used to produce books3, one of the most widely used and controversial llm training datasets

@frozenaesthetic

Share a piece of lore about yourself pic.twitter.com/NX4QIsJjKD

for those who don't know bibliotik was a private community that had hundreds of thousands of retail quality ebooks totally unavailable anywhere else, mostly downloaded from libraries and then with stripped DRM

they had a week long neutral-leech event. it was during covid and i was barely doing jack shit at work. i knew what i had to do.

grinded for basically the whole week straight, took like 50 transmission dockers and a fuckload of custom scraping software and my nonsensically overpowered NAS, but made it through with like 24 hours to spare

then i passed it off to the archivist for publication, wanted to make sure it was made widely available without catching flack

i got banned anyway of course, private trackers don't take kindly to public redistribution of their content

and the rest is history. just always funny when i see it brought up in news articles and im like damn that was literally all downstream of me taking too much adderall that one week

to everyone suggesting i should be more careful posting about a potential felony, this was all in early 2020 so we finally past the statute of limitations 👊

~/writing/the-scrape
sf