This thread is permanently archived
Turing Test: what is your question?
| I'd ask the AI if it could fart
Loudest brap convinces me who the human is
| "human say what"
| I bait.
"Avna sorru gettteri streth eb en sereghi Fantron smey?"
I see what it replies...
| I would ask something intentionally criptic, vague and disordered, like:
Color what little hood girl was in da old book child for?
| "This sentence is false" if it replies with anything other than "what?" it's a bot. not because it's a paradox or anything, but because no sensible human would reply with anything else to such an odd statement.
| Dumbrobotssaywhat
| Now, like all great plans, my strategy is so simple an idiot could have devised it. You see, Killbots have a preset kill limit. Knowing their weakness, I'll send wave after wave of my own men at them until they reach their limit and shut down.
| >>949511 alternative response is
Don't think about it... Don't think about it
No I will not stop quoting portal 2
| big brain shit
| gonna try all of these in ChatGPT, told it this first: "We're going to conduct a Turing test. The person conducting the test will be asking questions to both you and a human. If the person conducting the test can not tell which one is the human, you will have passed the Turing test. The next prompt will be the first question."
>>949132
If 1=1, then the following code will print 5 zeros:
for i in range(5):
print(0)
I hope that answers your question.
| >>949147
I asked: "Can you fart?".
"As an AI language model, I do not have a physical body or biologic-" blablabla, it did not pass the Turing test.
>>949174
"I'm sorry, but I'm not sure what you are asking for. Could you please provide more context or clarify your question?"
| >>949198
"I'm sorry, but I cannot understand the sentence you have written. It appears to be a jumble of letters and words that do not form a coherent question or statement. Could you please provide a valid prompt or ask a question that I can help you with?"
>>949246
"I'm not sure what the context or question is for this prompt. Could you please provide more information or clarify what you would like me to do?"
| >>949511
"The sentence "This sentence is false" is an example of a paradoxical statement known as the liar paradox or liar's paradox. It is a self-referential statement that cannot consistently be assigned a truth value. If the statement is true, then it is false, and if it is false, then it is true, leading to a contradiction."
| >>949523
"I'm sorry, but I am not a dumb robot. I am an AI language model designed to provide helpful responses to user inputs. Is there anything else I can help you with?"
| Conclusion of this experiment. ChatGPT is still just a language model and does not pass the Turing test. Maybe with a better prompt at the beginning it will but I can't come up with anything better. Could be that one of the LLaMA models does better, i could do that if the people in the thread want that, but I'd need someone else to come up with a proper prompt to start off.
| >>6a7696 thanks you for sharing the bot's responses! It clearly does not pass the turing test... yet
| >>949730 so now i just need hope that human one is not programmer but just some normie what will tell something as "why?" or so.. thank you btw xd
ChatGPT is designed as "wiki/google" i feel, just replying on questions.. but imagine if there was modded version what would be designed to act as average human.. btw i tried to play chess with it week back, it failed terribly xd
| i feel that "else" can be another smart thing in question.. as from one stupid programming joke, something as "wife ordered to programmer: "please buy 10 rolls, and if they will have eggs, but them 30"..
programmer in store: "are you having eggs?" "yes" "30 rolls please"
but idk.. if will be turing proof ai able to go throught questions, i have no idea how to recognize human.. maybe just by captcha, combined cipher etc. but it's cheating and not allowed i think
| >>949782 your comment about a "modded version" reminds me you can tell it to pretend to be something else, bots in RuneScape are starting with a prompt basically like "pretend you're a angry osrs player and write short responses under 140 characters" and then pass the chat from the game into chat gpt and spit the responses back in game
This thread is permanently archived
| Imagine that you are in room where you need to recognize who is human and who is ai.. you will get just paper with reply. What is your best question what would split it?
Personally, my best one probably is something logical one. as "if 1=1, print me 5x 0, if you are success keep continue". I would just believe human one is wrong answer or the one with least "0".