3 ms·
>Scalability, measured (the honest section) Ugh.
by torben-friis 3mo ago
>Scalability, measured (the honest section)
Ugh.
- greenchair 3mo agoalmost spit out my drink!
- poly2it 3mo agoI can't stand Claude's "honesty". Anthropic should hire some writers and linguists to make the output a bit more bearable. It's mentally taxing to read this type of dull text for hours every day.
- queenkjuul 3mo agoI just can't do it. I barely skim Claude PR reviews and have no qualms asking someone to summarize in their own words if they've sent me two pages of Claude slop
- mahogany 3mo agoThe entire PR description is filled with LLMisms. I find this style so hard to read, it’s almost nauseating these days. My eyes start to glaze over and I stop reading pretty quickly after. I’m having to deal with this a lot at work these days unfortunately. I mean, did anyone even read or verify this output? When I read this type of stuff, I have the nagging suspicion that I am the proofreader and I have to do the verification myself. At that point, what are we even doing? I can generate Claude output myself… I know I’m being overly dramatic but this sort of thing feels so wrong and inhuman(?) to me for some reason.
- rbonvall 3mo agoYes, that's the smoking gun!
- JamesSwift 3mo ago151 commits, 1826 files changed and claude found 0 smoking guns? I call bullshit. It found 3 in one run-of-the-mill query the other day