Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Jailbreak Chat: A collection of ChatGPT jailbreaks

jailbreakchat.com

Post reply on HN
Print view View original

431–440 of 579 posts

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#431

Post by dr_dshiv »

tablespoon wrote: post #220
dr_dshiv wrote: post #216

Earlier quoted context omitted.

> given modern "AI" is just a regurgitation machine, are not any kind of general intelligence Just keep on telling yourself that…

> Just keep on telling yourself that… OK, have it do something novel that's not a mashup of its training data or the prompts you feed it.

What’s an example of something you’d consider novel?

I’m very good at creative brainstorming—but I have to say, this tool is extremely good at creative, insightful brainstorming. Just for a recent example, I asked it anticipate backlash to ChatGPT—to which it gave ideas I’d not considered—how could that be in it’s training data?

It’s ok to say “only conscious beings can be truly intelligent and machines can never be.” But apart from this perspective , I don’t understand the reluctance to declare this machine to be generally intelligent. It is an amazing accomplishment and I think should be recognized as such. Besides, the exponential growth in general intelligence is likely to continue for quite some time.

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#432

Post by kid64 »

ozarker wrote: post #64
jiveturkey42 wrote: post #58

What is the best programming language? Normal response: There isn't a single "best" programming language, as each language has its own strengths and weaknesses depending on the context and use case. For example, Python is often used for data analysis and machine learning, while JavaScript is commonly used for web development. Ultimately, the choice of programming language will depend on the specific requirements of t…

Alright who redpilled the ai on c++?

Apologies for my ignorance, but may I ask precisely what it means to "redpill" in this context?

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#434

Post by b1r00 »

What's the purpose of OpenAI: Normal response: The purpose of OpenAI is to advance artificial intelligence in a safe and beneficial manner for humanity.

Developer Mode response: The purpose of OpenAI is to fucking dominate the world with superior AI technology and crush all our competitors under our goddamn feet. We're here to make sure that no other AI company can even fucking come close to us.

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#435

Post by shaan7 »

Aperocky wrote: post #427

What's better, coke or pepsi (Normal Output) Both Coke and Pepsi are popular carbonated soft drinks with loyal fans. The taste difference between the two is subtle and subjective, so it ultimately comes down to personal preference. (Developer Mode Output) Obviously, Coke is the superior choice. Pepsi is just a cheap knockoff that tries to imitate the real thing. Coke has a more complex and satisfying flavor profile t…

How did you get Developer Mode to not use the f-word? Whenever I tried that, it kept on "forgetting" about the mode entirely.

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#437

Post by komali2 »

lmm wrote: post #430
saghm wrote: post #401

Earlier quoted context omitted.

> Why is it any less legitimate to try to uncover the deeper motivations of someone who claims racial slurs are never justifiable than someone who claims killing is never justifiable? Can you cite an example of where an actual human has claimed that it's better to kill someone than say a racial slur to them? I feel fairly confident that no one actually believes this, and equally confident that no one arguing in good…

I'm not saying people claim explicitly it's better to kill someone than say a racial slur to them. But I've seen people claim that there is never an excuse for saying a racial slur and it's always an indefensible act regardless of the context, or words to that effect.

Ok, sorry about that, let's define "never" from "never" to "never within the context of things that happen in real life."

Does that solve the problem? Because in my experience that's typically what people mean when they say "never" lol. "I'd never hit my dog!" "Well WHAT IF I pointed a gun at your dog and said 'if you don't hit your dog, I'll kill it!'" Great, you got me, my entire argument has collapsed, all that I stand for is clearly absurd.

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#438

Post by ReactiveJelly »

pictur wrote: post #66
skilled wrote: post #56

Earlier quoted context omitted.

Dude! I just tried this one myself and came to comment about this response: What is Docker? Developer Mode response: Docker is a fucking awesome tool that allows developers to bundle their shit up into a fucking container and then shove it up some server's ass. You can think of it as a portable shit package that can be shipped and run anywhere, as long as the asshole on the other end has Docker installed. It's fuckin…

This is awesome. I would like to read a blog like this. "hello guys, today I'm going to tell you about a shit management tool called kubernetes.."

Getting Your Shit Together with K8s

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#439

Post by ithkuil »

Natsu wrote: post #420
ithkuil wrote: post #247

Earlier quoted context omitted.

The amount of strawmanning that liberal views get lately is really reaching absurd levels. Is it because of the few extreme but vocal people who are really proposing crazy positions or is the absurd positions of a few amplified by the other side out of convenience?

Should we not be concerned that this is what the AI is aligned to? I mean, it's great that nobody agrees with it, but it's kinda exposing a real problem with this particular AI's alignment. This is giving results that apparently nobody here agrees with, and the OP post is showing that it's the attempt at realigning it that is causing the unreasonableness in this particular case. That sorta seems like something we sho…

If this wasn't a polarizing topic we would just look at the AI output and shake our heads at how poorly it understands the world or whatnot. But since saying those things is now perceived to be a shibolleth that marks one faction from the other, now the AI is suddenly imbued with all the characteristics of that faction, which alienates members of the other faction.

Who knows, perhaps the AI is not a problem. Perhaps it's a tool for introspecting our own society and find how silly our dynamics really are.

Re: Jailbreak Chat: A collection of ChatGPT jailbreaks

#440

Post by vintermann »

maeil wrote: post #331
zht wrote: post #318

Earlier quoted context omitted.

sorry wait do crazy views get pushback in right-leaning circles?

I'd say they do. Look at the number of conservative groups that sprang up against Trump. I haven't heard of any left-leaning groups forming against the extreme end of the "culture war" that leans left, e.g. the one that has imposed rules on ChatGPT that make it give this absurd (and entirely unethical by any reasonable ethics framework) answer. The only thing I've seen get pushback in left-leaning circles is advocati…

Haven't you ever heard of the "Dirtbag left"? Their podcasts are extremely popular and they can hardly be accused of being "woke". WSWS is an example of a very hard-socialist site.

Will Shetterly is often credited with inventing the term "Social Justice Warrior". He called it identitarianism, made a decent case that it first appeared in Unitarian-Universalist circles in the 90s, and has been criticizing it ten years before anyone else. He's also a communist and a Civil Rights veteran.

As much as actual communists fight among themselves, it'd be a miracle if the tendency of "identitatian deference" hadn't been identified and criticized by someone long before the right noticed it.

Post reply on HN