6 ms·

Hey Boris from the Claude Code team - could you guys please be so kind so as to stop pushing that narrative about CLAUDE.md, either yourselves or through influe

Hey Boris from the Claude Code team - could you guys please be so kind so as to stop pushing that narrative about CLAUDE.md, either yourselves or through influencers and GenAI-grifters? The reason being, it is simply not true. A lot of the time the instructions will be ignored. Actually, the term "ignored" is putting the bar too high, because your tool does not intentionally "ignore", not having sentience and knowledge. We experience the effects of the instructions being ignored, because your software is not deterministic, its merely guessing the next token, and sometimes those instructions tacked onto the rest of the context statistically do not match what we as humans expect to see (while its perfectly logical for your machine learning text generator, based on the datasets it was trained on).

blitz_skull 11mo ago
This seems pretty aggressive considering this is all just personal anecdote. I update my CLAUDE.md all the time and notice the effects. Why all the snark?
- hansmayer 11mo ago
  Is it really just a personal anecdote ? Please do read some other comments on this post. The snark comes from everyone and their mother recommending "just write CLAUDE.md", when it is clear that this technology does not have intrinsic capability to perform reliable outputs based on human language input.
  - blitz_skull 11mo ago
    Yeah… that’s the point of LLMs: variable output. If you’re using them for 100% consistent output, you’re using the wrong tool.
    - hansmayer 11mo ago
      Is it? So you are saying software should not be consistent? Or that LLMs should not be used for software development, aside from toy-projects?
simonw 11mo ago
CLAUDE.md is read on session startup. If you're continually finding that it's being forgotten, maybe you're not starting fresh sessions often enough.
- hansmayer 11mo ago
  I should not have to fight tooling, especially the supposedly "intelligent" one. What's the point of it, if we have to always adapt to the tool, instead of the other way around?
  - kfajdsl 11mo ago
    It's a tool. The first time you used a shell you had to learn it. The first time you used a text editor you had to learn it. You can learn how to use it, or you can put it down if you think it doesn't bring you any benefit.
    - hansmayer 11mo ago
      I am sorry but what do I have to learn? That the tool does not work as advertised? That sometimes it will work as advertised, sometimes not? That it will sometimes expose critical secrets as plain text and some other time suggest to solve a problem in a function by removing the function code completely? What are you even talking about, comparing to shell and text editors? These are still bloody deterministic tools. You learn how they work and the usage does not change unpredictably every day! How can you learn something that does not have predictable outputs?
      - simonw 11mo ago
        Yes, you have to learn those things. LLMs are hard to use. So are animals, but we've used dogs and falcons and truffle hunting pigs as tools for thousands of years. Non-deterministic tools are still tools, they just take a bunch more work to figure out.
        11mo ago

blitz_skull 11mo ago

This seems pretty aggressive considering this is all just personal anecdote. I update my CLAUDE.md all the time and notice the effects. Why all the snark?

hansmayer 11mo ago
Is it really just a personal anecdote ? Please do read some other comments on this post. The snark comes from everyone and their mother recommending "just write CLAUDE.md", when it is clear that this technology does not have intrinsic capability to perform reliable outputs based on human language input.
- blitz_skull 11mo ago
  Yeah… that’s the point of LLMs: variable output. If you’re using them for 100% consistent output, you’re using the wrong tool.
  - hansmayer 11mo ago
    Is it? So you are saying software should not be consistent? Or that LLMs should not be used for software development, aside from toy-projects?

simonw 11mo ago

CLAUDE.md is read on session startup. If you're continually finding that it's being forgotten, maybe you're not starting fresh sessions often enough.

hansmayer 11mo ago
I should not have to fight tooling, especially the supposedly "intelligent" one. What's the point of it, if we have to always adapt to the tool, instead of the other way around?
- kfajdsl 11mo ago
  It's a tool. The first time you used a shell you had to learn it. The first time you used a text editor you had to learn it. You can learn how to use it, or you can put it down if you think it doesn't bring you any benefit.
  - hansmayer 11mo ago
    I am sorry but what do I have to learn? That the tool does not work as advertised? That sometimes it will work as advertised, sometimes not? That it will sometimes expose critical secrets as plain text and some other time suggest to solve a problem in a function by removing the function code completely? What are you even talking about, comparing to shell and text editors? These are still bloody deterministic tools. You learn how they work and the usage does not change unpredictably every day! How can you learn something that does not have predictable outputs?
    - simonw 11mo ago
      Yes, you have to learn those things. LLMs are hard to use. So are animals, but we've used dogs and falcons and truffle hunting pigs as tools for thousands of years. Non-deterministic tools are still tools, they just take a bunch more work to figure out.
      - 11mo ago