Top
Best
New

Posted by mfiguiere 2 days ago

An update on recent Claude Code quality reports(www.anthropic.com)
933 points | 721 commentspage 10
o10449366 2 days ago|
Resuming from sessions are still broken since Feb (I had to get claude to write a hook to fix that itself), the monitoring tool doesn't work and blocks usage of what does (simple sleep - except it doesn't even block correctly so you just sidestep in more ridiculous ways), and yet there seems to be more annoying activity proxies/spinner wheels (staring into middle distance)... Like I don't know how in a span of a few months you lose such focus on your product goals. Has Anthropic reached that point in their lifecycle already where their product team is no longer staffed by engineers and they have more and more non-technical MBAs joining trying to ride the hype train?
tdg5 2 days ago||
I missed the part about the refunds…
ayhanfuat 2 days ago||
Reading the "Going forward" section I see that they have zero understanding of the main complaints.
Kiro 2 days ago|
How so?
ayhanfuat 2 days ago||
They feel they're in a position to make important trade-off decisions on behalf of the user. "It's just slightly worse, I'll sneak this change in" is not something to be tolerated, whether it actually turns out to be much worse or not. Their adaptive thinking mess has caused a ton of work for me. I know a lot of people are saying Codex is actually better now. I don't agree but I'm switching to it because it's much more reliable.
operatingthetan 2 days ago||
I agree, but these LLM products are all black-boxes so we need to demand more accountability from them.
hajile 2 days ago||
My takeaway is that they knew they were changing a bunch of stuff while their reps were gaslighting us in the comments here.

Why should we ever trust what they say again out trust that they won’t be rug-pulling again once this blows over?

hirako2000 2 days ago||
In other words we did the right things, but we understand feedback, oh and bugs happen.
walthamstow 2 days ago||
So we weren't going mad then!
ElFitz 2 days ago||
Now we know why Anthropic banned the use of subscriptions with other agent harnesses: they partially rely on the Claude Code cli to control token usage through various settings.

And it also tells us why we shouldn’t use their harness anyway: they constantly fiddle with it in ways that can seriously impact outcomes without even a warning.

troupo 2 days ago||
> they were challenging to distinguish from normal variation in user feedback at first

translation: we ignored this and our various vibe coders were busy gaslighting everyone saying this could not be happening

gnegggh 2 days ago||
not the first time. Still not showing thinking are we?
whalesalad 2 days ago|
I genuinely don't understand what they have been trying to achieve. All of these incremental "improvements" have ... not improved anything, and have had the opposite effect.

My trust is gone. When day-to-day updates do nothing but cause hundreds of dollars in lost $$$ tokens and the response is "we ... sorta messed up but just a little bit here and there and it added up to a big mess up" bro get fuckin real.

More comments...