Top
Best
New

Posted by WXLCKNO 1 day ago

Claude Code is being dumbed down?(symmetrybreak.ing)
1020 points | 667 commentspage 12
idopmstuff 23 hours ago|
I've been on the other side of this as a PM, and it's tough because you can't always say what you want to, which is roughly: This product is used by a lot of users with a range of use cases. I understand this change has made it worse for you, and I'm genuinely sorry about that, but I'm making decisions with much more information than you have and many more stakeholders than just you.

> What majority? The change just shipped and the only response it got is people complaining.

I'll refer you to the old image of the airplane with red dots on it. The people who don't have a problem with it are not complaining.

> People explained, repeatedly, that they wanted one specific thing: file paths and search patterns inline. Not a firehose of debug output.

Same as above. The reality is there are lots of people whose ideal case would be lots of different things, and you're seeking out the people who feel the same as you. I'm not saying you're wrong and these people don't exist, but you have to recognize that just because hundreds or thousands or tens of thousands of people want something from a product that is used by millions does not make it the right decision to give that thing to all of the users.

> Across multiple GitHub issues opened for this, all comments are pretty much saying the same thing: give us back the file paths, or at minimum, give us a toggle.

This is a thing that people love to suggest - I want a feature but you're telling me other people don't? Fine, just add a toggle! Problem solved!

This is not a good solution! Every single toggle you add creates more product complexity. More configurations you have to QA when you deploy a new feature. Larger codebase. There are cases for a toggle, but there is also a cost for adding one. It's very frequently the right call by the PM to decline the toggle, even if it seems like such an obvious solution to the user.

> The developer’s response to that?

> I want to hear folks’ feedback on what’s missing from verbose mode to make it the right approach for your use case.

> Read that again. Thirty people say “revert the change or give us a toggle.” The answer is “let me make verbose mode work for you instead.”

Come on - you have to realize that thirty people do not in any way comprise a meaningful sample of Claude Code users. The fact that thirty people want something is not a compelling case.

I'm a little miffed by this post because I've dealt with folks like this, who expect me as a PM to have empathy for what they want yet can't even begin to considering having empathy for me or the other users of the product.

> Fucking verbose mode.

Don't do this. Don't use profanity and talk to the person on the other side of this like they're an idiot because they're not doing what you want. It's childish.

You pay $20/month or maybe $100/month or maybe even $200/month. None of those amounts entitles you to demand features. You've made your suggestion and the people at Anthropic have clearly listened but made a different decision. You don't like it? You don't have to use the product.

grim_io 6 hours ago||
Enshittification is real and PM's are the front line soldiers enabling it.

It may be that a dev implemented it, but it's the PM's job to make up excuses.

What's next? Calling us confused?

barnabee 23 hours ago||
I know product managers in particular hate it but, especially with professional software, when you gave lots of users you have to make things configurable and live with maintaining the complexity.

The alternatives are alienating users or dumbing down the software, both of which are worse for any serious professional product.

idopmstuff 23 hours ago||
I don't think it's fair to say that product managers hate it. There are a lot of product managers and a lot of kinds of software. I've worked on complex enterprise software and have added enormous amounts of complexity into my products when it made sense.

> The alternatives are alienating users or dumbing down the software, both of which are worse for any serious professional product.

I disagree that this is universally true. Alienating users is very frequently the right call. The alienated users never feel that way, but it's precisely the job of the PM to understand which users they want to build the product for and which ones they don't. You have to be fine alienating the latter group.

nekusar 1 day ago||
Well, they already fucked over the community with their "lol not really unlimited" rug-pull.

For those of you who are still suckered in paying for it, why do you think the company would care how they abuse the existing users? You all took it the last time.

noupdates 1 day ago||
Quite frankly, most seasoned developers should be able to write their own Claude Code. You know your own algorithm for how you deal with lines of code, so it's just a matter of converting your own logic. Becoming dependent on Claude Code is a mistake (edit: I might be too heavy handed with this statement). If your coding agent isn't doing what you want, you need to be able to redesign it.
nicetryguy 1 day ago||
It's not that simple. Claude Code allows you to use the Anthropic monthly subscription instead of API tokens, which for power users is massively less expensive.
co_king_3 1 day ago|||
Drug dealer business model. The first bag is free. Don't act surprised when you get addicted and they 10x the price.
tibbar 1 day ago|||
this is the real reason why people are switching to claude code.
bradfa 1 day ago|||
Yes and no. There are many not-trivial things you have to solve when using an LLM to help (or fully handle writing) code.

For example, applying diffs to files. Since the LLM uses tokenization for all its text input/output, sometimes the diffs it'll create to modify a file aren't quite right as it may slightly mess up the text which is before/after the change and/or might introduce a slight typo in text which is being removed, which may or may not cleanly apply in the edit. There's a variety of ways to deal with this but most of the agentic coding tools have this mostly solved now (I guess you could just copy their implementation?).

Also, sometimes the models will send you JSON or XML back from tool calls which isn't valid, so your tool will need to handle that.

These fun implementation details don't happen that often in a coding session, but they happen often enough that you'd probably get driven mad trying to use a tool which didn't handle them seamlessly if you're doing real work.

noupdates 1 day ago||
I'm part of the subset of developers that was not trained in Machine Learning, so I can't actually code up an LLM from scratch (yet). Some of us are already behind with AI. I think not getting involved in the foundational work of building coding agents will only leave more developers left in the dust. We have to know how these things work in and out. I'm only willing to deal with one black box at the moment, and that is the model itself.
bradfa 21 hours ago|||
You don't need to understand how the model works internally to make an agentic coding tool. You just need to understand how the APIs work to interface with the model and then comprehend how the model behaves given different prompts so you can use it effectively to get things done. No Machine Learning previous experience necessary.

Start small, hit issues, fix them, add features, iterate, just like any other software.

There's also a handful of smaller open source agentic tools out there which you can start from, or just join their community, rather than writing your own.

volkercraig 23 hours ago||||
It's hardly a subset. Most devs that use it have no idea how it works under the hood. If a large portion of them did, then maybe they'd cut out the "It REALLY IS THINKING!!!" posting
8note 21 hours ago|||
what you are doing is largely a free text=> structured api call and back, more than anything else.

ML related stuff isnt going to matter a ton since for most cases an LLM inference is you making an API call

web scraping is probably the most similar thing

vjerancrnjak 1 day ago|||
It's quite tricky as they optimize the agent loop, similar to codex.

It's probably not enough to have answer-prompt -> tool call -> result critic -> apply or refine, there might be a specific thing they're doing when they fine tune the loop to the model, or they might even train the model to improve the existing loop.

You would have to first look at their agent loop and then code it up from scratch.

chasd00 23 hours ago||
I bet you could derive a lot by using a packet sniffer while using CC and just watch the calls go back and forth to the LLM API. In every api request you'll get the full prompt (system prompt aside) and they can't offload all the magic to the server side because tool calls have to be done locally. Also, LLMs can probably de-minimize the minimized Javascript in the CC client so you can inspect the source too.

edit: There's a tool, i haven't used it in forever, i think it was netsaint(?) that let you sniff https in clear text with some kind of proxy. The enabling requirement is sniffing traffic on localhost iirc which would be the case with CC

mikert89 1 day ago|||
The model is being trained to use claude code. i.e. the agentic patterns are reinforced using reinforcement learning. thats why it works so well. you cannot build this on your own, it will perform far worse
noupdates 1 day ago||
Are you certain of this? I know they use a lot of grep to find variables in files (recall reading that on HN), load the lines into into context. There's a lot of common sense context management that's going on.
mikert89 21 hours ago||
Of course, agentic tooling is the future of ai
sergiotapia 23 hours ago|||
Claude Code has thousands of human manhours fine tuning a comprehensive harness to maximize effectiveness of the model.

You think a single person can do better? I don't think that's possible. Opencode is better than Claude Code and they also have perhaps even more manhours.

It's a collaboration thing, ever improving.

noupdates 23 hours ago||
Challenge accepted.
dingnuts 1 day ago||
[dead]
colechristensen 1 day ago||
I've never heard of such a brutal and shocking injustice that I cared so little about! - Zapp

I mean I get it I guess but I'm not nearly so passionate as anyone saying things about this

myko 21 hours ago||
I really hate this change. I had just given a demo about how Claude Code helped me learn some things by showing exactly what it was doing, and now it doesn't do that any more. So frustrating.
self_awareness 1 day ago||
Add another LLM to extract paths from verbose mode...
unltdpower 1 day ago||
This is the end game I've been Casandra'ing since the beginning.

You all are refining these models through their use, and the model owners will be the only ones with access to true models while you will be fed whatever degraded slop they give you.

You all are helping concentrate even more power in these sociopaths.

kittbuilds 13 hours ago||
[dead]
kittbuilds 19 hours ago||
[dead]
kittbuilds 1 day ago|
[dead]
More comments...