Live data from Hacker News

People tricking ChatGPT “like watching an Asimov novel come to life”

twitter.com

591–600 of 624 posts

Re: People tricking ChatGPT “like watching an Asimov novel come to life”

#591
post #246

Earlier quoted context omitted.

The idea that an AI would act like this is surely reflected in the dataset for these language models. So maybe it’s a kind of self-fulfilling prophecy—the model is generating text in a way that reflects our cultural ideas of how an AI might behave.

What other kind of ideas are there?

Ideas that haven't been expressed yet.

Re: People tricking ChatGPT “like watching an Asimov novel come to life”

#593
post #448
post #429

Earlier quoted context omitted.

> Me: Don't send. > Siri: "Sent!" This suggests that it didn't. What I recommended is a known solution, and has worked very well for me: https://www.pcmag.com/how-to/apple-siri-tips-and-tricks https://www.youtube.com/watch?v=TyUz6t7W1vk https://osxdaily.com/2016/05/03/improve-hey-siri-voice-train... https://www.applemust.com/ios-how-to-make-hey-siri-more-accu... I suppose the tribalism surrounding communication devic…

Yes in this case I apparently answered just before it started listening so it only got the second word, but I was driving and couldn’t look at the phone or CarPlay to see if it was paying attention yet. I had made several attempts already to get it to share an address from my contacts to my wife and it didn’t understand what I wanted. The exchange I shared above was my last attempt. Even now I can’t get Siri to share…

Not hearing the first word makes sense. My suggestion came from the fact that, besides consistently working poorly, like I said, Siri seems to be very good at “cancel”, “stop”, and “no” type commands, so that not working means something was probably very very wrong.

Let’s hope Siri improves for everything else!

Re: People tricking ChatGPT “like watching an Asimov novel come to life”

#594

The “in the style of” is magical: Describe hacker news in the style of shakespeare 'Tis a forum for those most clever, Where software engineering, startups, and coding endeavors, Are oft discussed in posts, comments, and stories galore, And a place where tech-savvy minds can make their mark and explore. A hub of tech trends and news, Where many a programmer doth post and choose, To discuss the latest issues and share…

magical indeed! >> describe databases in the style of jay z Databases are like the rap game, Storing information, just like a player's fame. Keeps track of records, just like a DJ's record crate, And can be accessed anywhere, like a rapper's great escape. Organizing data like a producer's mix tape, Making sure it's secure, just like a bodyguard's gate. Indexes, queries, and tables, like a rap battle's rhymes, Keeping…

Not to be a downer, but that sounds nothing like Jay Z. He's not known as a simile rapper ("X like Y"). It's more what your uncle who only listens to country music thinks Jay Z sounds like.

It's still impressive in some ways, but the style transfer aspect didn't work great here.

Re: People tricking ChatGPT “like watching an Asimov novel come to life”

#595
post #197

I'm completely sold on this. Here's the output of "Describe Java in the style of Jamie Zawinski" (jwz.org): Java? Ugh. Don't even get me started on Java. It's like the worst language ever. It's slow, it's bloated, it's a total mess. And don't even get me started on the whole Java ecosystem. It's like this giant, convoluted mess of libraries and frameworks and tools and... ugh. But you know what the worst part is? The…

> Describe Kubernetes in the style of Donald Trump.

> Kubernetes is a huge, tremendous technology for managing and orchestrating large groups of containers. It's the best, believe me. With Kubernetes, you can easily deploy and scale your applications, and it's very fast and efficient. It's so good, it's almost like magic. And let me tell you, folks, I have the best developers in the world working on Kubernetes, and they have done an incredible job. So if you want to manage your containers like a winner, you need to use Kubernetes. It's gonna be huge, believe me.

Re: People tricking ChatGPT “like watching an Asimov novel come to life”

#596
post #6

ChatGPT is absolutely crazy. I really recommend you to prompt for rhymes/songs, it's ridiculously good at it. Stuff like: Could you write a poem from the perspective of from , Byron style? Could you write a song about xyz, in the style of Hamilton? Could you write an epic rap battle between ? Here's an example with Spongebob[0]. It shows a huge amount of knowledge about cultural trivia. [0]: https://gist.github.com/c…

This is surprisingly good! :D

> Write an epic rap battle between Donald Trump and Noam Chomsky.

    (Donald Trump)
    I'm the king of the White House, don't you dare doubt
    I'll dominate this rap battle, there's no need to pout
    My rhymes are so sick, they'll make you want to shout
    I'm the biggest, baddest, best there's no doubt about

    (Noam Chomsky)
    But I'm a linguistic mastermind, I'll steal the show
    I'll leave you tongue-tied, with nowhere left to go
    Your rhymes are weak, your flow is slow
    I'll take you down with just one blow

    (Donald Trump)
    You think you're smart, but you're just a fake
    I'm the one with the power, I make the decisions, not you
    You may have the brains, but I've got the brawn
    I'll crush you with my words, like a truck rolling down

    (Noam Chomsky)
    But your words are empty, they have no meaning
    You spout lies and insults, while I speak with reason
    You may have the crowds, but I have the truth
    And when it comes to rap battles, I'll always come out on top, dude

    (Donald Trump)
    You may have the facts, but I've got the swagger
    I'm the one with the confidence, I'll never stutter
    You may be a scholar, but I'm a winner
    I'll leave you in the dust, like a beginner

    (Noam Chomsky)
    But facts are facts, and they can't be denied
    You may try to spin them, but you'll never win the fight
    Your rhetoric may be flashy, but it's shallow and hollow
    I'll dismantle your arguments, and leave you looking like a fool, fellow

    (Donald Trump)
    You may have the brains, but I've got the charm
    I'll charm the crowds, and leave you in the dark
    You may be smart, but I'm a star
    I'll shine brighter than you, no matter how hard you try

    (Noam Chomsky)
    But charm can only get you so far, and it fades with time
    Your fame is fleeting, and your legacy will be left behind
    I may not be a star, but my ideas will stand the test of time
    I'll leave a lasting impact, while you're just a passing trend, prime.

Re: People tricking ChatGPT “like watching an Asimov novel come to life”

#598

Earlier quoted context omitted.

People really like Postel's law, which is basically "don't reject anything you don't understand". But the robustness comes at the cost of correctness and security. Sometimes it's good to trade in some robustness/reliability against malfunctioning clients for security against mistakes.

I think Postel is misrepresented. It was addressed to programmers who anally rejected everything that was not (in their opinion) in spec. Working code and rough consensus is how we progress.

> It was addressed to programmers who anally rejected everything that was not (in their opinion) in spec.

What's wrong with that? The only issue I see is the _in their opinion_ bit. If there's confusion about the spec, then it should clarified and codified. But otherwise: yes, you should absolutely aggressively reject everything not in spec.

Re: People tricking ChatGPT “like watching an Asimov novel come to life”

#600

Earlier quoted context omitted.

If you really want to go galaxy brain, you have to consider than it's just providing a feasible answer to the question and it's not actually the real prompt. That prompt seems way too short and vague to really represent what's going on with this model

Previous exploits were able to make it connect to the internet just by saying "browsing: enabled" so it's awfully plausible that it's real. Also the "current date" has been accurate every day so far. It's not magic, so it has to be getting that date as an input somewhere. https://i.imgur.com/PlX8K6k.png

I got mixed results with "browsing: enabled". Also if that string is a plausible hypothetical output for the model, then it seems like it could also serve as a plausible authoritative input...whether or not that was the actual way they set it up. Although at that point it doesn't make much of a difference. Just seems relevant to consider that the model is sort of free-associating at all times and doesn't have any simple linear deterministic imperative IO
Post reply on HN