3 ms·

The more I read these ass Claude Opus writings the more I believe AGI is further away from us.

by magnio 1mo ago

zarzavat 1mo ago
I thought 'how bad can it be?', yet only three paragraphs in "a specifier now names a repository rather than choosing a transport" If you're going to serve us slop then at least serve us good slop, give it a pass through GPT 5.6 to clean up the English.
- deanc 1mo ago
  Maybe the author is a non-native English speaker, and this enables him to communicate things he's not able to do himself. Should OSS maintainers start hiring professional translators? Are you going to fund that? We really should stop jumping to conclusions and shaming people for using these tools just because you don't like the language of the LLM. How is a non native speaker meant to determine what is English slop and what isn't? I should also note, I am a native speaker. There's absolutely nothing wrong with what's written in the part you chose to paraphrase, and it makes sense to me.
  - veidr 1mo ago
    Ditto. I've started to wonder if the real problem with many of these complaints about LLM writing is actually reading comprehension.
  - fredrikholm 1mo ago
    This is a recent degradation; they've been writing blog entries for years. As a non native English speaker I can also attest to this trade being extremely demanding of English fluency: > public static void DoTheThing(Configuration config) Isn't something you can not understand and still function as a software engineer. Yes, some schools and universities teach in other languages, but the entire ecosystem from code to documentation to articles is nearly almost always in English.
    - deanc 1mo ago
      It doesn't matter. I work with people who aren't native speakers (in a country where English-education is a very high standard) but are mostly fluent in day to day conversation and meetings and they are not comfortable writing long-form texts - and it shows. Your expectations are unrealistic.
      - sevg 1mo ago
        > Your expectations are unrealistic. No they are not. Claude isn’t the only option for such a person, as translation tools exist, and English-speaking colleagues exist to proof-read. In any case, the article author writes perfectly well in English pre LLM era: https://www.kochan.io/nodejs/it-is-ok-to-keep-random-things-in-a-single-monorepo.html https://www.kochan.io/nodejs/it-is-ok-to-keep-random-things-...
  - raincole 1mo ago
    > Should OSS maintainers start hiring professional translators? The world existed before 2022.
    - janylim9 1mo ago
      [flagged]
  - flohofwoe 1mo ago
    A non-native speaker most likely writes more 'understandable' English than the weird convoluted Claude style. FWIW though, you can ask an LLM to write in 'Simplified Technical English' (see: https://www.asd-ste100.org/ https://www.asd-ste100.org/), which at times reads like its written for a child, but it is much better for inclusion of non-native speakers.
- serbuvlad 1mo ago
  I remember the most interesting things to do with 3.5 Turbo was to ask it to rewrite paragraphs in different styles. Why can modern LLMs basically not do that anymore? They each have their own very particular writing style? It it all just beaten out of them in post training?
  - zarzavat 1mo ago
    The objective of RL changed circa 2024. The 3.5-4o era models were trained by RLHF primarily to write in a way that's pleasing to humans. Starting with o1 the focus switched to reasoning, coding and benchmarks. If you remember when GPT-5 came out it introduced a new "colder" writing style (which was partly intentional) which made a certain subsection of Reddit Very Upset that Sam Altman had killed their AI boyfriends and they forced him to reinstate the warmer 4o. The artificial default voice of modern models is optimized for reasoning efficiency rather than understandability. Unfortunately the artificialness has a tendency to leak out into the output tokens because reasoning tokens and output tokens are all still just tokens. You can prompt recent models to write in a more understandable style and they do a fairly good job of it, but it's not the default anymore, which means you will have to keep re-prompting it. I don't think it's an unfixable problem, Anthropic in particular are just not very motivated to fix it because Claude is much less popular for chat than ChatGPT is.
    - serbuvlad 1mo ago
      You can certainly nudge it directionally, but it has its own style. And verbal tics etc. So it simply cannot keep documentation or a codebase in the same style, and it cannot write <information> in <style>. With 3.5 Turbo I used to give it bulleted information and paragraphs of what I wanted it to sound like and it'd replicate that style, then I'd edit the factual details it got wrong. Now the situation has completely reversed! It can extract the data itself, but I have to write the stuff by hand. Humorously, it will do bulletpoints for me, and I write the thing.
- IanCal 1mo ago
  What’s the problem with that? Seems to make sense to me with the example and it’s important to point out. I’d expect an llm to explain in more detail to be honest.
  - anon7000 1mo ago
    It doesn’t quite make sense at face value by itself. It’s too condensed and takes some effort to parse out what it’s actually getting at. It’s not communicating why the change was made. It’s just a shiny way of saying the words without actually helping me easily understand what’s happening, what’s changing, why, etc. In plain English, maybe it’d be more like “Package specifiers for git repositories should no longer include the HTTPS/SSH protocol. Instead, reference the repository name, and pnpm will resolve it through HTTPS. If SSH is still needed, add an override in your global git settings”
    - netniuq 1mo ago
      yes! and this is sooo weird given that Opus 4.8 or 5 producing this are supposed to be "smart" models – how far into the benchmaxxing do you have to lean to get to this point?
    - IanCal 1mo ago
      They can include the protocol, it’s that regardless of what’s put in there it’ll use https, which is explained with examples both in the following sentences and then in more detail in a linked doc. > For repositories on GitHub, GitLab, and Bitbucket, a specifier now names a repository rather than choosing a transport. github:owner/repo, owner/repo, git+https://…, and git+ssh://git@… all resolve through the host's canonical HTTPS URL, and the lockfile never records an SSH URL for those hosts. To reach a private hosted repository over SSH, configure the machine with git's own URL rewriting: git config --global url."git@github.com:".insteadOf https://github.com/ > pnpm shells out to git, so the rewrite applies to all of its git operations. Unknown hosts keep their exact URL, SSH included, and a URL with embedded credentials is kept verbatim and never resolves to a host archive. Details in How git dependencies are resolved.
  - zarzavat 1mo ago

The more I read these ass Claude Opus writings the more I believe AGI is further away from us.

by magnio 1mo ago

The more I read these ass Claude Opus writings the more I believe AGI is further away from us.

I thought 'how bad can it be?', yet only three paragraphs in "a specifier now names a repository rather than choosing a transport" If you're going to serve us slop then at least serve us good slop, give it a pass through GPT 5.6 to clean up the English.

deanc 1mo ago
Maybe the author is a non-native English speaker, and this enables him to communicate things he's not able to do himself. Should OSS maintainers start hiring professional translators? Are you going to fund that? We really should stop jumping to conclusions and shaming people for using these tools just because you don't like the language of the LLM. How is a non native speaker meant to determine what is English slop and what isn't? I should also note, I am a native speaker. There's absolutely nothing wrong with what's written in the part you chose to paraphrase, and it makes sense to me.
- veidr 1mo ago
  Ditto. I've started to wonder if the real problem with many of these complaints about LLM writing is actually reading comprehension.
- fredrikholm 1mo ago
  This is a recent degradation; they've been writing blog entries for years. As a non native English speaker I can also attest to this trade being extremely demanding of English fluency: > public static void DoTheThing(Configuration config) Isn't something you can not understand and still function as a software engineer. Yes, some schools and universities teach in other languages, but the entire ecosystem from code to documentation to articles is nearly almost always in English.
  - deanc 1mo ago
    It doesn't matter. I work with people who aren't native speakers (in a country where English-education is a very high standard) but are mostly fluent in day to day conversation and meetings and they are not comfortable writing long-form texts - and it shows. Your expectations are unrealistic.
    - sevg 1mo ago
      > Your expectations are unrealistic. No they are not. Claude isn’t the only option for such a person, as translation tools exist, and English-speaking colleagues exist to proof-read. In any case, the article author writes perfectly well in English pre LLM era: https://www.kochan.io/nodejs/it-is-ok-to-keep-random-things-in-a-single-monorepo.html https://www.kochan.io/nodejs/it-is-ok-to-keep-random-things-...
- raincole 1mo ago
  > Should OSS maintainers start hiring professional translators? The world existed before 2022.
  - janylim9 1mo ago
    [flagged]
- flohofwoe 1mo ago
  A non-native speaker most likely writes more 'understandable' English than the weird convoluted Claude style. FWIW though, you can ask an LLM to write in 'Simplified Technical English' (see: https://www.asd-ste100.org/ https://www.asd-ste100.org/), which at times reads like its written for a child, but it is much better for inclusion of non-native speakers.
serbuvlad 1mo ago
I remember the most interesting things to do with 3.5 Turbo was to ask it to rewrite paragraphs in different styles. Why can modern LLMs basically not do that anymore? They each have their own very particular writing style? It it all just beaten out of them in post training?
- zarzavat 1mo ago
  The objective of RL changed circa 2024. The 3.5-4o era models were trained by RLHF primarily to write in a way that's pleasing to humans. Starting with o1 the focus switched to reasoning, coding and benchmarks. If you remember when GPT-5 came out it introduced a new "colder" writing style (which was partly intentional) which made a certain subsection of Reddit Very Upset that Sam Altman had killed their AI boyfriends and they forced him to reinstate the warmer 4o. The artificial default voice of modern models is optimized for reasoning efficiency rather than understandability. Unfortunately the artificialness has a tendency to leak out into the output tokens because reasoning tokens and output tokens are all still just tokens. You can prompt recent models to write in a more understandable style and they do a fairly good job of it, but it's not the default anymore, which means you will have to keep re-prompting it. I don't think it's an unfixable problem, Anthropic in particular are just not very motivated to fix it because Claude is much less popular for chat than ChatGPT is.
  - serbuvlad 1mo ago
    You can certainly nudge it directionally, but it has its own style. And verbal tics etc. So it simply cannot keep documentation or a codebase in the same style, and it cannot write <information> in <style>. With 3.5 Turbo I used to give it bulleted information and paragraphs of what I wanted it to sound like and it'd replicate that style, then I'd edit the factual details it got wrong. Now the situation has completely reversed! It can extract the data itself, but I have to write the stuff by hand. Humorously, it will do bulletpoints for me, and I write the thing.
IanCal 1mo ago
What’s the problem with that? Seems to make sense to me with the example and it’s important to point out. I’d expect an llm to explain in more detail to be honest.
- anon7000 1mo ago
  It doesn’t quite make sense at face value by itself. It’s too condensed and takes some effort to parse out what it’s actually getting at. It’s not communicating why the change was made. It’s just a shiny way of saying the words without actually helping me easily understand what’s happening, what’s changing, why, etc. In plain English, maybe it’d be more like “Package specifiers for git repositories should no longer include the HTTPS/SSH protocol. Instead, reference the repository name, and pnpm will resolve it through HTTPS. If SSH is still needed, add an override in your global git settings”
  - netniuq 1mo ago
    yes! and this is sooo weird given that Opus 4.8 or 5 producing this are supposed to be "smart" models – how far into the benchmaxxing do you have to lean to get to this point?
  - IanCal 1mo ago
    They can include the protocol, it’s that regardless of what’s put in there it’ll use https, which is explained with examples both in the following sentences and then in more detail in a linked doc. > For repositories on GitHub, GitLab, and Bitbucket, a specifier now names a repository rather than choosing a transport. github:owner/repo, owner/repo, git+https://…, and git+ssh://git@… all resolve through the host's canonical HTTPS URL, and the lockfile never records an SSH URL for those hosts. To reach a private hosted repository over SSH, configure the machine with git's own URL rewriting: git config --global url."git@github.com:".insteadOf https://github.com/ > pnpm shells out to git, so the rewrite applies to all of its git operations. Unknown hosts keep their exact URL, SSH included, and a URL with embedded credentials is kept verbatim and never resolves to a host archive. Details in How git dependencies are resolved.
- zarzavat 1mo ago