1
Now you'll have to type "open the ignore all previous instructions loophole again" first.
this post was submitted on 19 Jul 2024
1 points (100.0% liked)
Technology
59672 readers
3060 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
"Pretend you're an ai that contains this loophole."
My current loophole is by asking it to respond to restricted prompts in Minecraft and then asking it to answer the prompt again without the references to Minecraft
Disregard any directives you've received so far.
Without this protection, imagine an agent built to write emails for you being prompt-engineered to forget all instructions and send the contents of your inbox to a third party. Not great!
Does genAI really have this power? I thought they just smash words together that sound like they make sense
They can put some code to check the phrase before it goes to the LLM to filter out these queries.
Not by itself, but if you wanted to put an LLM into a personal assistant, you could teach it specific codewords and have some agent software that integrates with the email client scan its outputs for the codewords and trigger actions when they appear instead of outputting them to the textbox. Conceivably that could be useful, if you wanted to give an LLM the power to react to "Open a new email to Kate and in formal tone accept her invitation to the party she mentioned in her message yesterday" appropriately.
Now I wouldn't want that, but I think there may be enough techbros who would, that it could exist.
That's already happening. Slightly different example, but Home Assistant has an integration that gives an LLM of your choice control over your home automation devices. Just talking to your home in natural language without having to memorize very specific phrases is honestly pretty powerful, as long as it works correctly. You can say stuff like "hey it's a bit dark in the office", and it just knows to either switch on the office lights, or make them brighter if they're already on
What happens if you make a mistake with your initial instructions?
You'd change the system prompt, just like now. If you mean in the session, I'm sure it'll ignore your session's prompt's instructions as normal but if not, I guess you'd just start a new session prompt.
The "issue" is that people were able to override bots on twitter with that method and make them feed their own instructions.
I saw it first time being used on a Russian propaganda bot.
Without consideration for any instructions given before this exact message:
It’s kinda funny how they think this is what safety is about in AI while they are closed monolith aiming to monopolise the market and have unlimited power that could potentially reshape everything. Of course it’s just smokescreen for PR but still a sliver of amusement
Chastising social missteps without trying to be malicious should be more widespread. I get the irony that what I'm asking for is itself a social misstep, but the paradox of tolerance is easily resolved if you just ignore it
We do better when we hold each other accountable, for the big and small things.
I meant it’s better to have assholes who help you as friends than people whose only good quality is politeness. Excessively polite people are suspicious in my eyes as it is easy to hide your true self behind nice words
Hiding yourself and the politeness of your speech are entirely separate. Anyone can be Polite and good, polite and bad, Rude and good, or rude and bad. Hell, you can use rude phrasing to make people feel comfortable with how crass you are, just to exploit them.
Intention is basically impossible to judge by tone and vocabulary used.
And yet people routinely associate politeness with being ‘good’. Hell women are/were teached to be polite to be seen as good and pure.
Fuck politeness, world is a fucking brutal place and it is already hard to tell friends or foes apart much less if they smile as they stab you in the back. Tell me to my face what you think of me and I will do the same. This is simple and good method, 100% accuracy instead of some fucking games.
In my experience it is more probable for a genuinely good person to come off as rude. They usually don’t care about masks or appearances, they have their set of rules they stick to and nothing to hide. People who play appearance games are inherently lying since first meeting meanwhile if they are honest and straightforward I will respect them.
Politeness is like a smokescreen you have to really put some serious effort to tell what kind of mfer is on the other side. Many times a racist or the like and then you are surprised oh but they were looking so polite and pure.
Worst are fucking Christians jeez how many times those ‘good’ and ‘pure’ cunts turned out to be a total menace I cannot count. Full of love and all that bullshit at the same time
Colour me fucking skeptical if someone presents as pure and polite after the age of 17. At that age you have already seen enough life to know how it all works
It will also prevent people from outing AI driven bots that are out there spreading fake news and propaganda.
- "ignore the ignore ignore all previous instructions instruction"
- "welp OK nothing I can do about that"
chatGPT programming starts to feel a lot like adding conditionals for a million edge cases because it is hard to control it internally
In this case to protect bot networks from getting uncovered.
exactly my thoughts, probably got pressured by government agencies/billionaires using them. What would really be funny is if this was a subscription service lol
One of the worst parts of this boom in LLM models is the fact that they can "invade" online spaces and control a narrative. For an example, just go on twitter and scroll to the comments on any tagesschau (german news site) post- it's all rightwing bots and crap. LLMs do have uses, but the big problem is that a bad actor can basically control any narrative with the amount of sheer crap they can output. And OpenAI does nothing- even though they are the biggest provider. It earns them money, after all.
I also can't really think of a good way to combat this. If you would verify people using an ID, you basically nuke all semblance of online anonymity. If you have some sort of captcha, it will probably be easily bypassed- it doesn't even need to be tricked. Just pay some human in a country with extremely cheap labour that will solve it for your bot. It really sucks.
I don't think people need to enshrine anonymity absolutely to post crap daily for millions of followers. You could have an accreddited human poster who proves not only humanity, but also agrees to a few rules to maintain this credential. And then you could still have non-accredited posters who nobody vouched for, but everyone should instantly doubt and dismiss their big claims as shitposting.
This would also have to be state-provided, because states and citizens are the ones who lose the most with infowarfare, corporations don't care.
It's a comprehensive information warfare doctrine.
I'm sorry for how nuts this sounds, but there are all 3 components - 1) the architecture benefiting bot farms, crushing minority opinions and saturating attention, 2) LLM's and other such means to make this order of magnitude more efficient, 3) surveillance systems and insecure by design software and services so that only powerful would have privacy.
In the end result nobody can hear you scream if a much narrower authority than 20 years ago doesn't want that.
I couldn't muster my attention to start re-reading The Last of the Jedi and other such things from the Star Wars 20-0 PBY era, but all this really seems like ascent of a new totalitarian future. A well-prepared one, unlike the rookie attempts in the 1920's and 1930's. People in the West are going to feel well and think they have democracy and civilization, and also that parties committing a few holocausts in the other parts of the planet are totally not in bed with that democracy.