Post B4GdxAAvmD8qzPaz56 by ariadne@social.treehouse.systems
 (DIR) More posts by ariadne@social.treehouse.systems
 (DIR) Post #B4GdxAAvmD8qzPaz56 by ariadne@social.treehouse.systems
       0 likes, 1 repeats
       
       @dalias this is why I get nervous when people talk to me about using LLMs for deep research
       
 (DIR) Post #B4GeI1hEUqG4m1eXRY by be_far@social.treehouse.systems
       0 likes, 0 repeats
       
       @ariadne @dalias 90% of what people call deep research seems to be “I don’t want to do research, but I want to be someone who has done research”
       
 (DIR) Post #B4GePnxXSDhIUAzMmG by WhiteCatTamer@mastodon.online
       0 likes, 0 repeats
       
       @ariadne @dalias Especially since the act of researching changes your perspective or assumptions about the questions you want to answer or topic to focus on.An LLM can spit out a summary paragraph or ten, but there’s something to be said about learning firsthand the shape of a subject by seeing both its rich and intricate areas and the voids and unknowns.
       
 (DIR) Post #B4Gxkh4i8wc8LR1ABc by quixote@mastodon.nz
       0 likes, 0 repeats
       
       @dalias What's the evidence for it being LLM slop? I glanced at it and it didn't look self-evidently awful. (I'm talking about the long article with nine sections or so.)
       
 (DIR) Post #B4GxkhTAg0BDZIShJQ by khm@hj.9fs.net
       0 likes, 0 repeats
       
       https://tboteproject.com/git/hekate/attestation-findings#methodology-and-toolsCC: @dalias@hachyderm.io