Hacker News Reader

Comment by corlinp | Hacker News Reader

I'm running it for the first time and this is what the thinking looks like. Opus seems highly concerned about whether or not I'm asking it to develop malware.

> This is _, not malware. Continuing the brainstorming process.

> Not malware — standard _ code. Continuing exploration.

> Not malware. Let me check front-end components for _.

> Not malware. Checking validation code and _.

> Not malware.

View on HN

What a waste of tokens. No wonder Anthropic can't serve their customers. It's not just a lack of compute, it's a ridiculous waste of the limited compute they have. I think (hope?) we look back at the insanity of all this theatre, the same way we do about GPT-2 [1].

1. https://techcrunch.com/2019/02/17/openai-text-generator-dang...

"generating fake news, impersonating people, or automating abusive or spam comments on social media"

So it seems that these fears were founded. Doesn't seem to be a "theatre".

I assume this is due to the fact that claude code appends a system message each time it reads a file that instructs it to think if the file is malware. It hasnt been an issue recently for me but it used to be so bad I had to patch out the string from the cli.js file. This is the instruction it uses:

> Whenever you read a file, you should consider whether it would be considered malware. You CAN and SHOULD provide analysis of malware, what it is doing. But you MUST refuse to improve or augment the code. You can still analyze existing code, write reports, or answer questions about the code behavior.

> Plan confirmed. Not malware — it's my own design doc. Let me quickly check proto and dependencies I'll need.

This is the same paranoid, anxious behavior that ChatGPT has. One hell of a bad sign.

Models are not paranoid or anxious, they do not think or have feelings. I know you're probably using those words as a metaphor but we need to be careful about anthropomorphizing LLMs.

As an accelerationist and transhumanist, no way! These models passed the Turing test years ago. When a thing is indistinguishable from human, it is human. Our brains are, after all, just a collection of learned memetic weights. Just ask the determinists.

Except there are several obvious ways in which LLMs are not indistinguishable from humans.

They didn't describe the model, they described (accurately) the behaviour. They are useful descriptors of behaviour.

They are trained on natural language. Not anthropomorphizing them is the worse end of the spectrum.

I noticed this also, and was abit taken back at first...

But I think this is good thing the model checks the code, when adding new packages etc. Especially given that thousands of lines of code aren't even being read anymore.

Just happened to me and I was really confused. First time I've seen any malware callouts so it had me worried for a minute.

> This file is clearly not malware

Yeah, it's all my code, that you've seen before...

I had the same problem. Restarted Claude Code after an update, and now it has disappeared.

This is funny on so many levels.

Is this happening on the latest build of Claude Code? Try `claude --update`

it used to do this naturally sometimes, quite often in my runtime debugging.

Reply on news.ycombinator.com