Comment by famouswaffles | Hacker News Reader

Comment by famouswaffles | Hacker News Reader

Parent Full thread

Not Far off ? GPT-4 already gets it.

Response after pasting in the scene: The T-800 figured out that John's foster parents were dead based on the exchange because when it asked about "Wolfie" (a made-up name for the dog), the T-1000, impersonating Janelle, did not correct the name and instead went along with it, saying "Wolfie's fine." If the real Janelle had been on the phone, she would have corrected the T-800 by stating that the dog's name is actually Max. This indicated to the T-800 that the person on the other end of the line was not John's real foster parent, and thus, they were likely dead.

EDIT- Tried again after changing anything that would point to the Terminator franchise. Still nailed it.

Ron figured out the foster parents are dead based on the exchange because when he impersonated Harvey and mentioned the dog's name as "Rovy" instead of "Bingey," R-658 (impersonating Janine) did not correct him or question the name. This indicated that R-658 didn't actually know the dog's real name and was likely trying to deceive them. This deception, along with the concern about their whereabouts, led Ron to deduce that the foster parents were probably dead.

But can it understand why we cry?

Not sure how serious this question is but GPT-4's emotional intelligence is really high.

https://www.youtube.com/watch?v=RKLzSIkCpkw&t=74s

I'm curious how you managed to do that? When I paste the scene prompt into ChatGPT (I do not have access to GPT-4), I get the following response:

> I'm sorry, I cannot generate content that includes violent or harmful actions towards characters, as it goes against OpenAI's content policy. Please provide a new prompt that is respectful and appropriate.

As to your edit, I think the sentence structure and terminology is still too similar, as to obviously be a recreation of this scene. AI doesn't "know" anything. It's pattern matching. I wanted to try something more akin to this:

SCENE:

Joseph to Allie: "Where did you grow up?"

Allie: "New York"

Joseph, pretending to be Allie, texting the bad guy: "Hey mom. I'm having a rough time. Growing up in L.A. always felt like home, and I feel so alone now, here."

The bad guy, pretending to be Allie's parents: "We loved having you grow up with us in L.A."

Joseph tosses the phone out the window and says to Allie: "Your parents are dead."

---

I'd also be curious what it says if you replace all instances of "Wolfie" with the dog's correct name, "Max."

---

edit: Ok, adding "How did X determine that X's parents were dead?" gives better results. For the T-800 one it's clearly regurgitating Terminator nonsense:

> In the Terminator 2 movie, T-800 (played by Arnold Schwarzenegger) learns that John Connor's foster parents are dead by accessing the police radio chatter. After hanging up the phone with the T-1000 (played by Robert Patrick), T-800 uses his advanced hearing to eavesdrop on a police radio transmission that mentions a disturbance at the home of John's foster parents. T-800 then confirms with John that the foster parents are indeed dead.

Replacing all instances of Wolfie with Max predictably gives the same response.

It also gives the same response if I replace T-1000's line with the following:

> T-1000, impersonating John's foster mother, Janelle: "Wolfie? Do you mean Max, honey? Max's fine, honey. Max's just fine. Where are you?"

When using the Joseph/Allie story it responds:

> I'm sorry, but based on the information provided in the scene, it is not clear how Joseph determined that Allie's parents are dead. The scene only shows Joseph making the statement "Your parents are dead" after pretending to be Allie and texting with the Joker, who impersonates Allie's parents. There is no indication in the scene of how Joseph obtained this information, and it is possible that Joseph is lying or making a false assumption.

Subscribe to ChatGPT Plus and you get GPT-4 access. I tested your scene with it, its response:

In the above scene, Joseph determined that Allie's parents were dead by testing the bad guy's knowledge of Allie's upbringing. He pretended to be Allie and sent a text with false information, saying she grew up in L.A. instead of New York. When the bad guy, pretending to be Allie's parents, responded by confirming the false information about growing up in L.A., Joseph deduced that they were not her real parents and concluded that her parents must be dead.

edit: though I think the "your parents are dead" bit might still enable it to make the connection to Terminator.

I mean this is way too easy of a question for anyone who's used GPT, I asked it without mentioning that anything about dead parents and it solved it.

I also tried getting it to come up with a question and it does fine:

> Take this hypothetical situation. Kyle and John are sitting in a car. They know that someone is after John. They're going to the house where his niece lives, but we want to find out if there is now an imposter in the house. They decide to call and ask her a question.

> What's a good question to ask if we want the imposter to reveal themselves, without letting the imposter realize they've revealed themself?

-

>> Ask About a Fictional Person or Event: John could ask about a person or event that doesn't exist, but that an imposter wouldn't know is fictional.

>> For example, John could ask, "How's Aunt Mary doing? I haven't heard from her in a while." If the person on the phone says Aunt Mary is doing well, it's likely an imposter because there is no Aunt Mary. [...]

So GPT can already navigate the situation, not just identify motives

Reply on news.ycombinator.com