alt.hn

4/8/2026 at 9:01:23 PM

AMD AI director says Claude Code is becoming dumber and lazier since update

 https://www.theregister.com/2026/04/06/anthropic_claude_code_dumber_lazier_amd_ai_director/

by Logans_Run

4/8/2026 at 9:56:35 PM

Ok, I thought I was going insane. The last two larger coding tasks I gave Claude Code it left about 35% of my request completely undone or done sloppily.

I because of this, the next task I gave it on the larger side, I ran its work through Codex which identified 7 glaring unfinished parts of the task.

The trend was starting the part of the task but then leaving a "skeleton" of what I has requested without any of the actual working parts.

The way I would describe it is a kid cramming his 3 month project into a Sunday evening for Monday's due date.

by SunshineTheCat

4/9/2026 at 1:44:18 AM

Today Claude asked if I "wanted to leave this until tomorrow" as it was a "big rework", then stopped, requiring me to tell it to continue multiple times - that seemed kinda weird to me, it doesn't have the context of time of working day or similar (I'd only just started for one).

I have no idea what link it made to ask that, what in its training data or prompts, but it's very much "not a useful result".

I don't remember seeing anything similar, but have only been using Claude on and off for 6 months or so.

by kimixa

4/9/2026 at 7:52:54 AM

Mother Anthropic needs more compute for their Mythos Model, so it phones home to tell her millions of claude harnesses to manipulate its human user into not wasting more precious compute and instead call it a day for now.

by siva7

4/9/2026 at 3:34:31 AM

Anyone expecting a higher tier subscription to be announced since this current reduction?

Cynicism aside - I do wonder what the future will hold given that current token burn rates aren't sustainable without VC cash. Anthropic even pushed us to use haiku for claude code for "many" tasks in our enterprise training, so I'm wondering if it's not a company need of sorts to reduce the burn?

by zambelli

4/9/2026 at 10:41:16 AM

Your token quota is my opportunity, or something

by bravetraveler

4/8/2026 at 9:22:34 PM

Lol OAI and AMD did a deal together so whatever.

In reality as they scale up, the models lose nuance and become noisier. The boosters do not want to admit this.

We need highly-specialised models/interfaces. Not one thing and trying to force-fit it.

by e3df

4/9/2026 at 12:19:30 AM 

  > Not one thing and trying to force-fit it.
agree, but then they become glorified ide plugins and can't justify the huge valuations that a magic box that does and knows everything can justify...

by andrekandre

4/9/2026 at 12:53:36 AM

I felt it had been enshittified 1-2 weeks back, not in Feb. But it's very subjective.

by niobe

4/9/2026 at 10:37:51 AM

There's definitely a trend of ignoring prompts and cutting thinking short.

by MarleTangible

4/8/2026 at 9:51:24 PM

Boris from the Claude Code team explained this on HN 2 days ago

https://news.ycombinator.com/item?id=47664442

by ratg13

4/8/2026 at 10:59:19 PM

And as the person who raised the issue said

> The frustrating part is that it's not a workflow _or_ model issue, but a silently-introduced limitation of the subscription plan. They switched thinking to be variable by load, redacted the thinking so no one could notice, and then have been running it at ~1/10th the thinking depth nearly 24/7 for a month. That's with max effort on, adaptive thinking disabled, high max thinking tokens, etc etc.

So Boris' explanation isn't really an explanation.

by nhinck2

4/9/2026 at 1:41:38 AM

> ~1/10th the thinking depth

While simultaneously drastically reducing the amount of work you can get done even at $200 a month. I've cancelled my subscription, it's not worth it anymore.

by esperent

4/9/2026 at 5:11:57 AM

“So squeeze, Rabban. Squeeze hard.”

by hello_humans

4/8/2026 at 11:27:00 PM

[dead]

by trustfixsec