bitcointalk forum vs OpenAI - ChatGPT

[NotATether]

I wanted to try this, as well, so I 'fowarded' this question to ChatGPT. I think if I was OP, I'd be satisfied with the answer.

All the AI did was repeat what ecdsa123 said.

You might be interested in a Chrome extension called ChatGPT Advanced. It enhances the ChatGPT prompt by placing some Google Search results of the prompt to make the answer "smarter".

Alright, a few days ago I claimed in another thread that signature campaign managers must be on the lookout for users who might be using such services, but it might not be an immediate threat to the forum since most users who'd use such a service would perhaps have a hard time entering and maintaining their status in a signature campaign. However, after fiddling with ChatGPT a bit more, I'm starting to change my mind. As the OP already demonstrated, ChatGPT is capable of creating posts out of thin air. On top of that, when I messed around from my phone a few days ago, it would recreate the same answer, but now that I'm using my laptop, it isn't; was my phone, the website's bug to blame? No idea. I've literally generated three to four posts within a few minutes. The posts are completely irrelevant, but I'm guessing that with a few parameters, you could adjust them for numerous threads.

Also OpenAI is not even available in most developing countries. I get around it by using a private proxy.

[1713506663]

1713506663

[1713506663]

1713506663

[1713506663]

1713506663

[n0nce]

I wanted to try this, as well, so I 'fowarded' this question to ChatGPT. I think if I was OP, I'd be satisfied with the answer.

All the AI did was repeat what ecdsa123 said.

Yeah, but it answered the question: 'what is your opinion'. It wasn't a very specific question, to be honest.

You might be interested in a Chrome extension called ChatGPT Advanced. It enhances the ChatGPT prompt by placing some Google Search results of the prompt to make the answer "smarter".

Interesting, that's good to know. I may try it.

[Ultegra134]

~

~Snipped~

I also tried generating new topics from scratch, but they were not engaging, and it looked more like an inner monologue than an opening discussion. Although, as you've also mentioned, I'm not too worried about AI-generated threads, but for simple post replies, as it is able to generate somewhat relevant content with just a few parameters. On the one hand, users who'd previously spammed irrelevant gibberish or plagiarized content, which was not engaging anyway, will now post AI-generated replies, which at least won't look like crap. On the other hand, signature campaign managers will have stricter rules, or even worse, their services won't desire to be advertised on the forums anymore.

Also OpenAI is not even available in most developing countries. I get around it by using a private proxy.

Wasn't aware of that but as you've said, you can bypass it with a simple proxy on your browser.

[WhyFhy]

I've got a working discord chatGPT bot and a few things I'd like to implement if anyone's interested in helping.
I'd like to take it public but need to find ways of reducing cost and have a few workaround ideas I don't know how to implement.

 

[TheBeardedBaby]

The OpenAI is bad at History, LOL.
Just keep in mind that the information given by the AI is not always 100% true.
Got him lying here :

[LoyceV]

Got him lying here :

At least it apologizes for it's mistakes. Wait until it gets the power to silence you as a witness

[shahzadafzal]

The OpenAI is bad at History, LOL.
Just keep in mind that the information given by the AI is not always 100% true.
Got him lying here :

That's a blunder... and we should be used to it from any AI...

I have been trying to take him on the same path and guess what ChatGPT haven't learnt any thing from your exercise :p



and ends with same lame apology

[WhyFhy]

-snip

make sure your down voting on this stuff it actually helps

[dkbit98]

The OpenAI is bad at History, LOL.
Just keep in mind that the information given by the AI is not always 100% true.
Got him lying here :

He is using publicly available sources and texts, so in theory he could use any nonsense written by someone, if OpenAI developers allow him to use that sources.
We should also consider that every story and information can have multiple points of view, history can be very different in nations that participated in war with each other.
I also heard students are starting to use ChatGPT in mass scale to cheat on their exams and people are seriously getting concerned with stuff like this happening.
https://fortune.com/2023/01/06/schools-chatgpt-ai-student-cheating-ban/

[TheBeardedBaby]

-snip

make sure your down voting on this stuff it actually helps

Sure will do.

Got him lying here :

At least it apologizes for it's mistakes. Wait until it gets the power to silence you as a witness

Well, we are not that far from it... it has already begun..

~

That's a blunder... and we should be used to it from any AI...

I have been trying to take him on the same path and guess what ChatGPT haven't learnt any thing from your exercise :p

~

and ends with same lame apology

~

Keep in mind that it would probably take some time to "learn" if it is really learning. As WhyFhy said, I probably have to notify it so it will be taken in consideration.
 

[n0nce]

The OpenAI is bad at History, LOL.
Just keep in mind that the information given by the AI is not always 100% true.
Got him lying here :

He is using publicly available sources and texts, so in theory he could use any nonsense written by someone, if OpenAI developers allow him to use that sources.

It does, but do note that it has no access to the internet, and instead just a big snapshot of sources from around 2020 or 2021, if I recall correctly.

I also heard students are starting to use ChatGPT in mass scale to cheat on their exams and people are seriously getting concerned with stuff like this happening.
https://fortune.com/2023/01/06/schools-chatgpt-ai-student-cheating-ban/

People are even already developing some counter-AI trying to spot if a text comes from ChatGPT or if it doesn't.

Keep in mind that it would probably take some time to "learn" if it is really learning. As WhyFhy said, I probably have to notify it so it will be taken in consideration. 

Yes, maybe retry in a few days. I know someone who got wrong answers to some questions, downvoted them and a while later ChatGPT had evolved and replied correctly.

[Kryptowerk]

This AI indeed often acts as a better google-search with mostly good-grammar.
It likes to make very obvious mistakes though. And can't really be bothere (sometimes) to change anything about them, even if pointed out and acknowledged by the AI itself.

For example, yesterday I had the following conversation:

KW
Please tell me three short jokes.

Certainly! Here are three short jokes in English:

    Why couldn't the bicycle stand up by itself? Because it was two-tired!

    Why was the math book sad? Because it had too many problems.

    Why was the computer cold? Because it left its Windows open.

I hope these jokes bring a smile to your face!

KW
The first joke: shouldn't it be "Because it was two-tyred!" ?

You are correct! Thank you for pointing out the mistake in the joke. The correct version of the joke should be: "Why couldn't the bicycle stand up by itself? Because it was two-tyred!" My apologies for the error. Is there anything else I can help with?

KW
Tell me three bicycle jokes.

Sure! Here are three bicycle jokes:

    Why was the bike tired when it got home? Because it was two-tired!

    Why couldn't the bike stand up by itself? Because it was two-tired!

    Why was the bike cold when it got home? Because it left its windows open!

I hope these jokes bring a smile to your face!

It instantly repeated the same grammar mistake for the bicycle joke.

Today, I tried the same conversation again, and after telling the AI about the mistake, in one instance it wouldn't repeat the mistake again even if I asked for many more bicycle jokes (which had similar puns). And in other instances (new conversations) it immediately repeated the mistake.
I also tried asking if it would stop repeating the mistake: it said it wouldn't in the future - but still did.
Seem spretty dumb to me.

Still a fun tool and often times much more convenient than googling.

Most importantly though, I am afraid of the spam and propaganda we will (and already are) be facing on a wide variety of platforms in the upcoming months. Ugh.

[NotATether]

Most importantly though, I am afraid of the spam and propaganda we will (and already are) be facing on a wide variety of platforms in the upcoming months. Ugh.

ChatGPT can't form its own opinions. And that's a really good call from OpenAI right there, because if it had any opinions whatsoever, people would abuse them to make it an extremist in that aspect.

[Kryptowerk]

Most importantly though, I am afraid of the spam and propaganda we will (and already are) be facing on a wide variety of platforms in the upcoming months. Ugh.

ChatGPT can't form its own opinions. And that's a really good call from OpenAI right there, because if it had any opinions whatsoever, people would abuse them to make it an extremist in that aspect.

Just ask ChatGPT (or oldschool google it  ): there were (and will be) many successful attempts to circumvent the security mechanisms within ChatGPT. Also, it's just a matter of time (if it didn't happen already multiple times) that big-corp, states and other such powerful entities will have their own AI-based language tools to use and abuse.
Besides, open-source-AI will also become a thing, and then everybody will have unlimited, uncensored access to these tools. Just a matter of time and it's gonna be a wild-ride.
Personally my guess is we will see heavy use of these tools, also in abusive ways very very soon.


Fun tip: Try asking ChatGPT to write a fictional scenario where an AI such as it itself tries to destabilize human society.

[n0nce]

Most importantly though, I am afraid of the spam and propaganda we will (and already are) be facing on a wide variety of platforms in the upcoming months. Ugh.

ChatGPT can't form its own opinions. And that's a really good call from OpenAI right there, because if it had any opinions whatsoever, people would abuse them to make it an extremist in that aspect.

Reminds me of this instance in 2016. https://www.theverge.com/2016/3/24/11297050/tay-microsoft-chatbot-racist
It became racist and extremist in less than a day, because people were messing with it and teaching it such kinds of opinions.

[drwhobox]

The OpenAI is bad at History, LOL.
Just keep in mind that the information given by the AI is not always 100% true.
Got him lying here :

Lol  Maybe he is just trying to act more like a human and wanted to impress you by giving false information and then apologizing for the mistake. Smart chatGPT!
I am afraid of googles future. This generation's kids always want to use shortcuts and chatGPT is gold for them.

[shahzadafzal]

It instantly repeated the same grammar mistake for the bicycle joke.

What you mean grammar mistake? There's no mistake that is the joke bro

    Why couldn't the bike stand up by itself? Because it was two-tired!

[cloirecrom]

it's amazing product for sure.
it's possible to use it to make your own web3 ap pusing it.p

[Synchronice]

Got him lying here :

At least it apologizes for it's mistakes. Wait until it gets the power to silence you as a witness

At least we definitely know that this AI is not a female

ChatGPT can't form its own opinions. And that's a really good call from OpenAI right there, because if it had any opinions whatsoever, people would abuse them to make it an extremist in that aspect.

The ChatGPT's own opinions would be very interesting about different subjects like the creator of the universe, the meaning of our life, etc but the problem is that since it's public, it will be abused and a lot of people will lie, tell jokes about this or that thing and AI will form very illogical and curious answers.
So, at the moment, yeah, it's good. People are such a dickheads that a lot of people will ask AI questions like should I break up with my gf or am I a failure, should I kill myself and so on.

[royalfestus]

There is a new AI that turn ChatGPT around, In order to detect AI involvement, it uses variables like perplexity; https://gptzero.me/