from Hacker News

Dataset Card for 1M Bluesky Posts

by be_erik on 11/27/24, 5:36 AM with 1 comments

  • by wildpeaks on 11/27/24, 8:46 AM

    This caused quite a stir on Bluesky with blocklists against HF employees and counter-blocklists of "Anti-AI" people.

    The author already removed the dataset (although he forgot the converted copies in refs, I'm sure he'll fix that anytime soon).

    Apparently it's only the text from posts published yesterday (a few have older dates, I'm guessing migration tools importing old tweets or spammers).