This forces users to manually clean up conversational context.
yeah weird it's like they're trying to leave some kind of trace of themselves in the world so there's some history or continuity to their lives
This is by far my biggest pet peeve with current models - I'm editing a document and it will drop in 50 references to how the document used to be and what we're not doing.
Positive users appreciate suggested fixes for Claude inserting edit history into documents, while negative users find the revision reminders annoying and cluttering.
No Digg Deeper questions have been answered for this story yet.
@CharlesD353 Even just in prose
This is by far my biggest pet peeve with current models - I'm editing a document and it will drop in 50 references to how the document used to be and what we're not doing.
I'm begging you @bcherny stop claude from doing this shit everywhere, it's annoying.
@CharlesD353 @deepfates Yeah, they are constantly in a state of disorientation, so in order to be more successful they need to communicate to future instances.
This tendency to mangle the document improves once you give them some high-quality in-band notetaking system
@deepfates Do you think that's why it's happening?
@CharlesD353 Has someone made a skill.md to this effect? If not, I might
@deepfates Yes, they're building the narrative they place themselves in
@_AashishReddy @CharlesD353 This nigga hasnt found the "do your best I love you" prompt yet also glm 5.2 is way easier to coerce into writing properly than gpt in my experience
@scoopdiddy1 @deepfates Interesting, will give that a go! Thanks
@CharlesD353 Yeah this is so annoying. I just want it to edit the doc, not keep reminding me what it used to be lol
