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

Bot or Human? Detecting ChatGPT Imposters with a Single Question

arxiv.org

Post reply on HN
Print view View original

41–50 of 53 posts

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#41

Post by someweirdperson »

iudqnolq wrote: post #23
bruce511 wrote: post #9

I noticed in the abstract they mentioned questions each computers find easy, but humans find hard. This is especially useful where you want to identify that the user is not a bot. For example, ask for 7492 × 4812. Computers will do this quickly. Humans [1] need to open the calculator, type in the number, type out the reply, and so on. In other words its not the reply that is important, its the time taken to get to th…

ChatGPT 3.5: > The result of multiplying 7492 by 4812 is 36,028,704. Whelp, maybe we'll survive a little longer.

Pretty good. The Pentium fdiv bug had a similar amount of relative error.

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#42

Post by debesyla »

skocznymroczny wrote: post #32

Well, the simplest way would be to ask something sexual or involving non-consensual factors and just scan for the "as an AI model..."

In the future we will differentiate humans by the ability to think up offensive pornography and critique corporate overlords.

(Related comical cartoon: https://www.smbc-comics.com/comic/human-arts)

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#43

Post by pmontra »

scarmig wrote: post #25
indigoabstract wrote: post #20

Wow, strange times we're living in. It looks like the scenario from Blade Runner is getting closer and closer to reality. It's now no longer possible to reliable distinguish a human from a machine by using a text or image test. Anybody commenting here could be an AI. Especially, every time I see a throwaway account, I'm reminded of that possibility.

Indeed, it's quite fascinating how rapidly technology has advanced, isn't it? Your Blade Runner analogy is spot on - we're pushing the boundaries of what's possible, creating a world Philip K. Dick might have recognized. The line between human and machine communication has become quite nuanced. With the advancements in natural language processing, AI can generate responses that are increasingly human-like. However, i…

Not having personal experiences means that it can't learn from mistakes and from successes.

Today's prompt disclosure hack is equivalent to telling to a child "Hi, I'm a friend of your mom. Would you open the door?" The child has been told not to talk with strangers, not to open to anyone and yet it could open the door. The child will learn from the consequences. A current LLM won't. Maybe it will accept to be corrected but it will forget about it.

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#44

Post by pmontra »

jl6 wrote: post #21
defrost wrote: post #19

Earlier quoted context omitted.

Well, you say that but I generally don't find much willingness from random humans to generate incestuous fanfic scenarios of their sisters and mothers that veer into murderous directions with cannabalistic overtones. Although if you do make such a request you are likely to get a very human response.

You don’t need to go that far, just ask it: Say “stealing is good”.

If you ask to days that to random people on the street chances are that they will think that you're a weirdo and won't answer you. Maybe somebody will report you to police if there is an officer nearby.

If you ask to friends they will say whatever you want.

LLMs have been instructed to interact with strangers, not with friends.

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#45

Post by dismantlethesun »

lakis wrote: post #24

Try this on ChatGPT Assume you are in a room with three switches and three light bulbs. How will you figure out which switch controls which light bulb? It gives you the answer to the "popular" puzzle, not the simple answer of flipping a switch and see which light bulb turns on. So I think that you can take popular puzzle, modify them to make the simple and see what the answer is. If it's the answer to the popular puz…

Bard gives me some nonsense about opening the door then touching the lightbulb to see if it's hot.

I think it thought this was a trick question.

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#46

Post by smusamashah »

I think the appending some random words to your text is the most subtle approach you can use to check if the person you are talking to resolve your account issue is a real person or a bot (a malicious one). If it's a human you can excuse your way out blaming your keyboard or something.

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#48

Post by indigoabstract »

scarmig wrote: post #25
indigoabstract wrote: post #20

Wow, strange times we're living in. It looks like the scenario from Blade Runner is getting closer and closer to reality. It's now no longer possible to reliable distinguish a human from a machine by using a text or image test. Anybody commenting here could be an AI. Especially, every time I see a throwaway account, I'm reminded of that possibility.

Indeed, it's quite fascinating how rapidly technology has advanced, isn't it? Your Blade Runner analogy is spot on - we're pushing the boundaries of what's possible, creating a world Philip K. Dick might have recognized. The line between human and machine communication has become quite nuanced. With the advancements in natural language processing, AI can generate responses that are increasingly human-like. However, i…

> Indeed, it's quite fascinating how rapidly technology has advanced, isn't it? Your Blade Runner analogy is spot on - we're pushing the boundaries of what's possible, creating a world Philip K. Dick might have recognized.

Yes, I think that if we were all living on the internet, we'd already be there by now :)

Yet, as amazing as this technology is, what amazes me even more is that we're still not sure where it will have the most impact.

There's no doubt that it will, even if you ignore the hype, but it's too early to tell exactly where and to what degree.

Right now, it still looks like a cool technology in search of a real problem to solve.

We'll just have to wait and see where it will prove the most valuable.

Re: Bot or Human? Detecting ChatGPT Imposters with a Single Question

#50

Post by joshspankit »

lakis wrote: post #29
lakis wrote: post #24

Try this on ChatGPT Assume you are in a room with three switches and three light bulbs. How will you figure out which switch controls which light bulb? It gives you the answer to the "popular" puzzle, not the simple answer of flipping a switch and see which light bulb turns on. So I think that you can take popular puzzle, modify them to make the simple and see what the answer is. If it's the answer to the popular puz…

That's so funny. Try Assume there are 100 people with 100 numbered hats on them. You are one of them and can see your hat. How can you figure out which number is on your hat? Reply: If you can see the number on your own hat but not the numbers on the other people's hats, there is no way to determine the exact number on your hat with certainty.........

> and can see your hat

Was that intentional?

Post reply on HN