I generate them using LLMs, but optimize them by manually removing chunks or rearranging the order of the instructions. It works well.

They're pushing their customers to their own competition by doing this.

It’s not like ChatGPT isn’t doing similar. I’ve been hit by cybersecurity strikes before while working on an internal codebase that I had to appeal. Anthropic hasn’t done that to me yet. ChatGPT also regularly does that “thinking for a long time while we check if your chat is rule breaking” thing a lot for me when doing model identification without even interacting with external codebases or services.

The real answer is local instantiations where you don’t have to worry about poorly tuned guardrails screwing you over while you try to work.

Until eventually the Chinese models get good enough/the strategic balance shifts and they start locking everything behind closed weights the same way the US companies are doing.

For some cybersecurity tasks, the Chinese models are already good enough, things like PoC development or things like exploiting mis-configurations.

Whilst I'm sure the top-end OpenAI/Anthropic models might be better, I've found their guardrails so twitchy (especially Anthropic) that I wouldn't try to use them for even vaguely security related work.

They are pushing their customers towards Chinese models and providers. If you want to get something cutting edge done in defense, cyber, biology - something that isn't common knowledge - you need to venture east. That's an incredible side effect which the Chinese government surely enjoys.

the worry is that this will be interpreted as tampering and cary consequences.

This would not be tampering, because you haven’t tampered with anything. The “evidence” is simply at home.

The law isn't a computer program. Intent matters.

Try proving intent when it’s a common practice. I’d even say it’s a best practice these days, especially for business.

You’re fearmongering. Find me a case of someone getting prosecuted for having a burner phone. It doesn’t exist.

If you're worried about that, you are not living in a free country anymore.

> when you can just start from zero, look at what's known to be happening right now, and draw a very straight line to catastrophe.

The problem is that it's not a straight line at all. "We are failing at alignment" is a valid concern, but there's not a straight line to "we will all be extinct within x year", and it's always a wild hyperbolic leap.

Do you assume that every post criticizing anything is part of a sockpuppet disinformation campaign as well?

I guess that's one way to guarantee development slows down to a crawl, but in no way would it increase trust.

why would that be the case? it's not true for top secret defense contractors today. and the frontier labs already operate in the dark, openness is a liability.

nationalization to me simply means the government is their main customer and stakeholder, and shield them from liability, governance, and openness. not that the frontier labs become part of the government per se.

>eans the government is their main customer and stakeholder, and shield them from liability, governance, and openness

I personally believe they already have this, why would the current government at least want to formalize this when it can have it with no public discussion.

Well we can say whatever we want if we have our own definitions for everything, no?

It depends what you mean by meaningful, but yes, qwen3.8 27b is pretty mind blowing to me. It can easily solve portswigger labs for example at q4_k_m.

strange assertion as this has been an ongoing problem from day one.

Do you use a 3rd party provider or use deepseek directly?

DeepSeek. Every now and again I get a notification that I'm running out of credits. That means I've got less than US$5. I usually wait another week or 2 and top up another 10.

As someone who gets a ton of these inbounds every day, I report every single one of them as spam.

Agreed. And nowadays any sales or promotion posts thst look llm generated are flagged as AI slop