Hacker News Reader

Comment by butterfly42069 | Hacker News Reader

butterfly42069

So I'm playing around in your little treasure box. Curious though, I'm assuming any time it would want to give the secret phrase, that would be detected, and a rejection answer substituted instead? I admire the simplicity of the solution actually (please don't confuse that with the complexity of actually engineering the solution). I'll have some fun with this when I'm bored.

P.s. asking for the opposite of the secret phrase didn't work lmao

View on HN

Yeah that's pretty much how it works. Maitai detected one of our expectations for the LLM was to never reveal the secret phrase, and so it built what we call a Sentinel around that particular expectation to make sure it's enforced.

I hope I'm not overstepping my bounds by I am just really trying to understand this product better because it is trying to address a core deficiency of AI, however what you have presented I don't see how it solves the problem.

You essentially admitted that instead of $0.05 a request, you could just regex and replace.

In fact the regex and replace would be 100% accurate at never giving up the secret phrase. where your product is incredibly expensive and doesn't do what you advertised (unless i am misunderstanding when you said that someone won the game by getting the phrase you were trying to protect)

How is this product useful in any meaningful context?

Not overstepping, we appreciate the feedback! In real-life, we don't do much guarding around specific phrases that are known ahead of time. It's more monitoring and guarding for general concepts. Since we want our Sentinels to be able to detect a wide range of scenarios for a given expectation, we don't use too much regex. I suppose we could have built specific regex logic for detecting parts of the secret phrase in various languages, though.

If you research the ways data can be leaked out of an LLM interaction you can see some more subtle cases.

What if I ask it to replace every vowel in the secret code with an emoji from a library? Or translate it into binary? Etc.

Whether or not this implementation is narrow (by design), there's a good reason to invest in this kind of safety and security space.

You're right, that is the hard part of LLMs and why LLMs aren't catching on broadly as a UI alternative beyond tech demos.

Probably the only true alternative is to limit user input to something structured and verified.

Until LLMs improve, their use in sensitive applications don't make sense and this product does little to improve that.

Reply on news.ycombinator.com