Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Why does Opus 5 feel worse to work with?

mun-logadan.github.io

Post reply on HN
Print view View original

321–330 of 915 posts

Re: Why does Opus 5 feel worse to work with?

#321

Post by Merad »

dingaling911 wrote: post #178
docheinestages wrote: post #166

Claude has essentially become useless for agentic development or research. Doesn't matter what model you use. A few rounds and bam, you've burned through your quota. Doesn't matter how "intelligent" their models are, if you can't use them. That, and the quality of AI responses are, in my opinion, significantly worse than competitors like OpenAI. At this pace, I foresee Anthropic becoming the next Nokia. If you would'…

What are you guys doing to burn through limits? I have some dev + prod bots and according to ccusage, use the equivalent of $2500/month with them on CC yet I never hit the rate limits. I feel like I'm using them all the time so I'm curious what you are actually doing that's burning all of these tokens. Can you give me an example? For me, it's: 1. Write a spec for 2. Add design for issue 3. Write code 4. Deploy code a…

I don't get it either. Earlier this week I had Claude build a POC for a tool we're evaluating. I spent a few hours building out a requirements doc in a chat with Fable. Then Claude Code built it with Fable as the manager and the code being written by about 80% Opus/20% Sonnet subagents (this was using superpowers). It ran continuously for 15 hours on the implementation and didn't hit any limits.

Re: Why does Opus 5 feel worse to work with?

#322

Post by VeninVidiaVicii »

anentropic wrote: post #304

> stop and ask questions if my intent was unclear, > don't make assumptions without checking, > and don't reinterpret or update my plans without asking. these aren't at all the problems I have with it I have found it good at asking questions, to the extent I rarely use 'plan mode' any more but often it's hard to understand what it's asking me, it's like the question framing has been pulled from the middle of its own…

I find the questions are not clearly asked and somehow half baked as caveats toward the middle end of what it’s vomiting.

“One thing I deliberately didn’t touch” — about half the time this is something completely irrelevant or something that is actually the target of whatever you’re working on, and the shakespearean prose it says around this phrase is a “question” it has.

Re: Why does Opus 5 feel worse to work with?

#323

Post by martinald »

adamcharnock wrote: post #232

My latest trick (literally from yesterday) is to just ask it to write according to ISO 24495-1, the standard for plain language: > [This standard is] for anybody who creates or helps create documents. The widest use of plain language is for documents that are intended for the general public. However, it is also applicable, for example, to technical writing, legislative drafting or using controlled languages. You don'…

Keep in mind all this kind of stuff can make the model less capable. If it has to think in "plain" English, it may well be squashing quality of code etc output.

I'm not sure how true this is, but when using "forced" json output it def had a big drop off in quality - https://arxiv.org/html/2408.02442v3.

I think you're better not fighting it with hacks like this and find a different model.

Re: Why does Opus 5 feel worse to work with?

#324

Post by artdigital »

ryanschaefer wrote: post #230
D13Fd wrote: post #120

I’ve been doing some heavy work on a personal project lately. I burned through the limits on Claude, the plus a few hundred dollars in credits, and ultimately decided to move to an OpenAI account just so I can keep going. I was surprised to find that OpenAI Sol is much much nicer to work with than Opus 5 or Fable at the moment. Especially on Opus 5, the way it communicates is just exhausting. It keeps “being honest”…

I had this debate with my coworker who prefers anthropic models to open ai ones. I ended up settling into the idea that gpt 5.6 is better used as a tool and opus 5 is a companion. GPT 5.6 takes you literally whereas opus 5 tends to take more liberties to try to get to the “spirit” of what you want. It comes down to preference, and I don’t want a companion.

That’s how I see it too. Claude is more “fun” to use, like a coworker I have to talk to now and then to steer it, while gpt-5.6 is a task machine: I give it a task and it is very consistent, reliable and predictable in its execution. I don’t have to interrupt it, it gets the task done exactly how I wanted it, but it’s “boring” and feels more sterile

Re: Why does Opus 5 feel worse to work with?

#325

Post by yard2010 »

D13Fd wrote: post #174
disgruntledphd2 wrote: post #155

Earlier quoted context omitted.

Yeah, I actually have started using GPT Sol much much more, as Claude (all of them) were far too trigger happy around making changes, and refused to listen to my requests to take things slowly. Feels like they've overtrained on one-shotting (which does demo well, and presumably converts new subscribers), whereas I want a model to do work for me in small, easily understood changes that I can hold in my head (maybe I'm…

I think you’re right. It’s not optimized for some kinds of work. My little project has a Textual TUI interface that needs to display a few hundred thousand rows in a table. It takes 14 seconds to load in the default datatable component. I instructed Opus 5 to replace the datatable component with a fasttable alternative, a new dependency. I let it go overnight. When I got back up, it had spun for hours and proudly ann…

I love how it sometimes has false a-ha moments and being "honest" about it, it feels so cringe. Then I recall it has no intelligence. It's a stochastic parrot. It's funny how people, like any other animal, that beetle of David Attenborough trying to reproduce with a brown bottle comes to mind, think that something is X just because it trips the right neurons.

Re: Why does Opus 5 feel worse to work with?

#326

Post by docheinestages »

dingaling911 wrote: post #178
docheinestages wrote: post #166

Claude has essentially become useless for agentic development or research. Doesn't matter what model you use. A few rounds and bam, you've burned through your quota. Doesn't matter how "intelligent" their models are, if you can't use them. That, and the quality of AI responses are, in my opinion, significantly worse than competitors like OpenAI. At this pace, I foresee Anthropic becoming the next Nokia. If you would'…

What are you guys doing to burn through limits? I have some dev + prod bots and according to ccusage, use the equivalent of $2500/month with them on CC yet I never hit the rate limits. I feel like I'm using them all the time so I'm curious what you are actually doing that's burning all of these tokens. Can you give me an example? For me, it's: 1. Write a spec for 2. Add design for issue 3. Write code 4. Deploy code a…

My uses cases are mainly research and prototyping, often starting from scratch in greenfield projects. The 5-hour quota is shared between Claude web and Claude Code, so it doesn't really matter what interface I use.

When it comes to research, my prompts are already narrowed down to specific topics, and I even include examples and break the process down into stages. For development tasks, I try to avoid a mono-repo in the beginning and develop modules before combining them together to avoid distracting the AI's attention and minimize the overhead.

With Codex, on GPT-5.6 Sol with xhigh effort, I need to go several rounds and at least 2-3 hours before hitting the (now-removed) 5-hour limit, which translates to 10% of the weekly usage. In contrast, I run out of quota even with Claude Sonnet.

In terms of quality of output, Codex digs deep for research tasks, in the right direction, produces less AI slop, and follows my direction better. At least that's how I perceive it. But again, the main problem with Claude is running out of quota in the middle of research or implementing a task.

Re: Why does Opus 5 feel worse to work with?

#327

Post by chuckadams »

stingraycharles wrote: post #226
barrkel wrote: post #12

The single biggest annoyance with Opus 5 is that it writes too elliptically. Sentences that orbit a point, then jump to it like it's a revealed insight. Unnecessarily abstract phraseology. Constantly using inanimate nouns as the subjects in sentences in order to unlock variety in verb choice, especially when it helps construct a sentence where the real action can 'land' like a surprise at the end. It is definitely mo…

> Sentences that orbit a point, then jump to it like it's a revealed insight. Is this inside the thinking tokens, or the output? As this type of stuff is expected for thinking, because of the whole CoT / “think step by step” works, as this is optimal for the way LLMs work with attention and next word prediction. So the fact that it first “orbits” a point only to get to the conclusion afterwards is the system working…

You forgot about 1000 lines of "Wait, but different idea here" ;)

Re: Why does Opus 5 feel worse to work with?

#328

Post by silverwind »

martinald wrote: post #323
adamcharnock wrote: post #232

My latest trick (literally from yesterday) is to just ask it to write according to ISO 24495-1, the standard for plain language: > [This standard is] for anybody who creates or helps create documents. The widest use of plain language is for documents that are intended for the general public. However, it is also applicable, for example, to technical writing, legislative drafting or using controlled languages. You don'…

Keep in mind all this kind of stuff can make the model less capable. If it has to think in "plain" English, it may well be squashing quality of code etc output. I'm not sure how true this is, but when using "forced" json output it def had a big drop off in quality - https://arxiv.org/html/2408.02442v3 . I think you're better not fighting it with hacks like this and find a different model.

Changing output style shouldn't affect thinking at all.

Re: Why does Opus 5 feel worse to work with?

#329

Post by jordz »

Opus 4.6 was the sweet spot for me as a thinking partner specifically.

I use these models for coding, but also a lot of product, commercial, financial and architectural work where I’m trying to develop something half-formed. 4.6 was unusually good at understanding what I was trying to get at, playing it back cleanly, getting the nuance, and extending it without bastardising it as the conversation was drawn out.

It could make useful connections without constantly trying to manufacture an insight.

5.6 Sol is genuinely excellent at the creative part, and in some cases better than 4.6. My issue is convergence to get to a point, a final point. As you try to distil an idea, it often invents new terminology for concepts you’ve already established but its so subtle you have to really keep track of it. The vocabulary and idea tree keep expanding when what you actually want is to collapse everything down to the few things that matter.

Opus 5 has the same problem for me that barrkel said, the prose is often so elliptical and I just want it to tell me it and get to the point than making me dance around what its trying to tell me.

I don’t think 4.6 was necessarily the most capable model (compared to Fable) for long horizon task delivery, and Opus 5 is much more Fable like, it's fiercely determined to get through the task list .

4.6 just felt unusually well calibrated to my way of collaborating and its ability to understand, extend and then compress my thinking without constantly imposing some random walk.

Re: Why does Opus 5 feel worse to work with?

#330

Post by thr0w »

MEMORYC_RRUPTED wrote: post #13

It's not even code for me, but the prose it writes. For some reason, the way Opus 5 "talk" elicits frustration in a way that 4.5 to 4.8 never did. Can't put my finger on why, but I've flipped over to Codex because what it produced wasn't worth the frustration.

It is rage inducing.
Post reply on HN