Hacker News .hnnew | past | comments | ask | show | jobs | submit | JSR_FDED's commentslogin

I always dismissed this category as more “markdown engineering” but this opened my eyes to some genuinely interesting things. The AI Memory space is more varied than I expected.

This is an easy fix.

Remember the leaked Claude Code contained a regex to determine user frustration?

Just add another one to spot the pattern: ‘disregard previous instructions’.

This is a load-bearing change. Now Claude will Delve into your task without distraction.


I see what you did there ;)

It seems The Register just discovered that Prompt Injection is a thing.

No, the world needs to be reminded that it is _still_ a thing and will _remain_ to be a thing.

Like buffer overflows, and raw sql, and …

But I guess it’s good that noble people are reminding us that the things that were a thing yesterday are still things today and will be things tomorrow.


Not really an accurate comparison since buffer overflows and sql injection are bugs which ultimately allow user data to co-mingle with executable code. LLMs take user data and mix it with the "executable code" (if we are extremely generous in our description of a user prompt) by design.

The issue here is unavoidable because LLMs are broken by design. There is no encapsulation where you can separate instructions and data because LLMs are nothing more than next-token predictors and the input sequence MUST be a sequence. They can't build a model with one stream for instructions and another for data because the training data they stole from the internet and books is a single stream.


While I agree that LLMs have yet again surfaced the “new tech fails to separate data and control” issue that affected everything from pay phones to SQL, I disagree that there’s something different that prevents the introduction of separate planes.

That “stolen” training data, most of which itself was stolen from older works, does not include user prompts. It is data, not control.

We will see models with annotations for whether a token is part of user prompt, and other ways as well.

You’re obviously passionate about the subject but as someone who works in the field, I assure you there is no now-and-forever requirement for a single stream with no metadata about tokens. We will positively see control and data separated just like they were for phones and databases.


> You’re obviously passionate about the subject but as someone who works in the field, I assure you there is no now-and-forever requirement for a single stream with no metadata about tokens

I'm quite familiar with how LLMs work internally. If you have an example of how the isolation you are describing could work, you'll have to explain it. By what possible mechanism could "tagging" tokens allow you to isolate the influence between tokens once they are taken into the network? They're still just floating point numbers at the end of the day. To actually treat user prompt data separately from untrusted data, you will need to figure out some new kind of multiplication.

> That “stolen” training data, most of which itself was stolen from older works, does not include user prompts.

Also, don't lie to me, it's rude.


> Like buffer overflows, and raw sql, and …

Those are fixable. Prompt injection is not.


They should rotate all employees between divisions.

When done creating AI puzzles they can enjoy a stint in the Content Review team.


They should spend one day working at a McDonalds or as a last mile delivery driver. Ill bet playing with a chat bot looks a lot better after that.

“I joined this fine company to help accelerate the destruction of society, and now instead I’m expected to help it destroy society in a _different way_ by creating puzzles for AI. Now my morale is low. Poor me. “

Sign me up, I will gladly give up my blue collar hell to torture an AI to failure.

This is excellent. Retro enjoyable and the closest thing to actually engaging with the economy!

Yeah that sucks. Now imagine the average Chinese developer who encounters this all the time!

A nerd shibboleth, love it!

Does brew gather statistics that could show what portion of users is on Intel vs Apple Silicon?

https://formulae.brew.sh/analytics/homebrew-os-arch-ci/30d/

From people who haven’t disabled analytics.


This is great news. Apple never reverses course this quickly, so this is a clear signal that Stephen Lemay is making his presence felt.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: