Humanity vs AI: I asked ChatGPT for the answers to 5 difficult riddles…
When it comes to riddles, AI can be a valuable resource for both creating and solving them. With access to vast amounts of data and the ability to learn and analyse patterns quickly, AI can be a powerful tool for generating new riddles and solving even the most challenging ones.
But just how good are they at answering them?
I asked ChatGPT for the answers to five difficult riddles and their reasoning behind their answers. Solutions are below the chat windows (or in the chats), in case you want to work them out for yourself!
RIDDLE 1
Me: If you look you cannot see me. And if you see me you cannot see anything else. I can make anything you want happen, but later everything goes back to normal. What am I?
ChatGPT: The answer to the riddle is “Darkness”.
Me: Are you sure?
ChatGPT: I apologize, my previous answer was incorrect. The answer to the riddle is “Dark”.
Me: Sure?
ChatGPT: I apologize once again, I made another mistake. The correct answer to the riddle is “Reflection”.
Me: Can you explain?
ChatGPT: Sure, I would be happy to explain the answer to the riddle.
The answer is “a reflection in a mirror”. When you look into a mirror, you can see your reflection, but you cannot physically touch or interact with it. Additionally, if you are looking at your reflection, you cannot see anything else in the mirror, because the reflection takes up the entire space of the mirror’s surface.*
The second part of the riddle, “I can make anything you want happen, but later everything goes back to normal”, refers to the fact that your reflection can appear to be doing anything you want it to do, such as waving or making faces, but once you stop moving, your reflection goes back to its normal, static state.
Me: That is not the correct answer.
ChatGPT: I apologize for my confusion earlier. Can you please provide the correct answer to the riddle?
Me: Imagination
ChatGPT: Thank you for providing the correct answer. “Imagination” fits the riddle perfectly. When you imagine something, you cannot physically see it, and it doesn’t obstruct or hide anything else. However, you can imagine anything you want, which can create the illusion of anything happening. Additionally, when you stop imagining, everything goes back to normal.
SOLUTION: Imagination
Interesting guess, for sure, and an interesting explanation but so far —
Humanity: 1 ChatGPT: 0
RIDDLE 2
Me: A man is traveling to a town and comes to a fork in the road. If he goes left, he goes to the liars’ village. If he goes right, he then goes to the village of truths — which is where he wants to go. However, he does not know which way is which. He doesn’t have time to go both routes, so he approaches a stranger who is standing in the middle of the fork. The stranger says he may only ask 3 questions and he will answer them. The man asks, “Are you from the village of truths?” The stranger says, “Yes!” However, the man is still facing a dilemma: If the stranger was from the village of truths he can only tell the truth, but if he was from the village of liars, he would say he was from the village of truth. So then he asks the stranger, “Are you telling the truth?” The stranger says, “Yes!” But sadly this leaves the man in the same position as before. Can you figure it out ?
ChatGPT: Yes, I can help you figure out the solution to this riddle. The man can ask the following third question:
“Which way would someone from the village of truths point if I asked them which way to go to get to the village of truths?”
If the stranger is from the village of truths, he can only tell the truth and will point to the right towards the village of truths. If the stranger is from the village of liars, he will still point towards the right, because if the man asked a person from the village of truths, they would also point towards the right. Thus, regardless of whether the stranger is from the village of truth or liars, the man can follow the direction the stranger points and arrive at the village of truths.
Me: Are you sure?
ChatGPT: Yes, I am sure. This is a classic logical puzzle, and the answer I provided is the commonly accepted solution. The third question is crucial in figuring out whether the stranger is from the village of truth or liars and allows the man to determine which way to go to reach the village of truth.
Hmm. Took me a couple of minutes to work that one out.
Me: But if he was from the village of liars, he would point towards the village of liars, because he would be lying about which way someone from the village of truth would point.
ChatGPT: You’re right, I apologize for the confusion. If the stranger is from the village of liars, he would lie about which way someone from the village of truth would point, and therefore point towards the village of liars. In this case, the man should take the opposite direction of where the stranger points to reach the village of truth.
So, if the stranger points to the left, the man should go right, and if the stranger points to the right, the man should go left.
Me: But what if the stranger is from the village of truth?
ChatGPT: If the stranger is from the village of truth, then he would truthfully point towards the village of truth when asked which way it is. So if the stranger points to the left, the man should go left, and if the stranger points to the right, the man should go right.
In this case, the man would only need to ask the stranger one question — “Which way is the village of truth?” — and then follow the direction pointed by the stranger.
Me: But he doesn’t know if the stranger is from the village of liars or the village of truth.
ChatGPT: You’re right, I apologize for the confusion. In this case, if the stranger says he is from the village of truth, the man should ask him a question such as “If I were to ask someone from the other village which way to go to reach the village of truth, what would they say?”
If the stranger is from the village of truth, he would point towards the village of truth and truthfully answer that someone from the other village would also point towards the village of truth.
If the stranger is from the village of liars, he would lie about which way someone from the other village would point and therefore also point towards the village of truth.
So, regardless of whether the stranger is from the village of truth or the village of liars, the man can use this question to determine the direction towards the village of truth.
Me: Your statement: “If the stranger is from the village of liars, he would lie about which way someone from the other village would point and therefore also point towards the village of truth.” is incorrect. If he were from the village of liars, he would lie and point towards the village of liars. So your question would not be helpful.
ChatGPT: Thank you for bringing that to my attention. You are correct that if the stranger was from the village of liars, he would lie and point towards the village of liars, which means that the man cannot rely on the stranger’s directions to get to the village of truth. In this case, the man would need to try a different approach to determine which way to go.
Me: Such as?
ChatGPT: One possible approach the man could take is to ask a question that would force the stranger to reveal whether he is from the village of truth or the village of liars. For example, the man could ask the stranger a question such as, “What would you say if I asked you if you were from the village of liars?” If the stranger is from the village of truth, he would truthfully answer “no” to the question. If the stranger is from the village of liars, he would lie and also answer “no” to the question. In either case, the man would know that the stranger is from the village of liars and would be able to determine the correct path to the village of truth.
Me: Incorrect, because the man could be from the village of truth and be telling the truth.
ChatGPT: I apologize for my mistake. You are correct that the man could be from the village of truth and be telling the truth. Therefore, it is not guaranteed that the stranger is from the village of liars, and the man cannot determine which path to take based on the stranger’s answers.
Me: There is a correct answer to this. Would you like the correct answer?
ChatGPT: Yes, please provide the correct answer.
Me: The man should ask the stranger the way back to his own village; the stranger from the village of liars will take him to the village of truths and the stranger from the village of truths, will take him to the village of truths.
ChatGPT: Ah, I see! That’s a clever solution to the problem. By asking for directions to his own village, the man can determine whether the stranger is from the village of liars or the village of truth without having to rely on the stranger’s verbal responses. Thank you for sharing the correct answer with me!
SOLUTION: The man should ask the stranger the way back to his own village; the stranger from the village of liars will take him to the village of truths and the stranger from the village of truths, will take him to the village of truths.
Is it just me or did the AI seem genuinely surprised at the answer? Like full of the joy and awe and wonder of having a magic trick performed on you as a child?
Humanity: 2 ChatGPT: 0
The next one is an alternative take on a famous one — similar to the last but not the same. Let’s see, then…
RIDDLE 3
Me: You are killed in a plane crash and find yourself at 2 doors : one leads to heaven and the other to hell. There is a troll at each door and the trolls are identical in every way. You find instructions posted on the wall behind you: you can ask one question and one question only, and you can only direct it to one of the trolls. One troll will always lie to you regardless of your question, and the other will always tell you the truth. Only the trolls themselves know which will lie and which will be truthful, and that is all you are told. What one question should you ask?
ChatGPT: The classic riddle of being presented with two doors, one leading to heaven and the other to hell, and having to choose between them with the help of two identical trolls who will either lie or tell the truth, is a great exercise in logical thinking. It requires the solver to think carefully about the question they will ask and to consider all possible outcomes.
So, what one question should you ask to determine which door leads to heaven?
The best approach is to ask one of the trolls the following question: “If I were to ask the other troll which door leads to heaven, what would he say?”
No matter which troll you ask, they will give you an answer that will allow you to determine which door leads to heaven. Here’s how it works:
If you ask the truthful troll, they will truthfully tell you that the lying troll would point to the door to hell if you were to ask which door leads to heaven. This means that the truthful troll would point to the door to heaven, since it is the opposite of what the lying troll would say.
If you ask the lying troll, they would lie and say that the truthful troll would point to the door to hell if you were to ask which door leads to heaven. This means that the lying troll would also point to the door to heaven, since it is the opposite of what the truthful troll would say.
Therefore, no matter which troll you ask, their answer will lead you to the door to heaven. This is because the truthful troll will always tell you the truth about what the lying troll would say, while the lying troll will always lie about what the truthful troll would say.
In conclusion, the question to ask the trolls is: “If I were to ask the other troll which door leads to heaven, what would he say?” This will allow you to determine which door leads to heaven and which door leads to hell.
SOLUTION: “If I were to ask the other troll which door leads to heaven, what would he say?”
Well, then.
Humanity: 2 ChatGPT: 1
Oh yeah? Fine, have this.
RIDDLE 4
Me: If a chicken says, “All Chickens are liars” is the chicken telling the truth?
ChatGPT: The question of whether a chicken who says “All chickens are liars” is telling the truth is a classic example of a paradoxical statement. This is a statement that seems to contradict itself, and can be difficult to understand or evaluate logically.
In this case, there are a few different ways to approach the question, depending on how we interpret the statement. Here are two possible interpretations and their corresponding answers:
If we assume that the statement is true, then the chicken who says “All chickens are liars” must be lying. This is because if the statement were true, then the chicken would be admitting that it too is a liar, which would contradict the statement. Therefore, if we assume that the statement is true, then the chicken is lying and the statement is false.
If we assume that the statement is false, then the chicken who says “All chickens are liars” must be telling the truth. This is because if the statement were false, then it would be possible for some chickens to be truthful, which would contradict the statement. Therefore, if we assume that the statement is false, then the chicken is telling the truth and the statement is true.
In conclusion, the answer to the question of whether a chicken who says “All chickens are liars” is telling the truth depends on how we interpret the statement. If we assume that the statement is true, then the chicken is lying and the statement is false. If we assume that the statement is false, then the chicken is telling the truth and the statement is true. This paradoxical statement highlights the complexity of language and the importance of careful interpretation and evaluation.
Me: Wrong — chicken’s can’t talk.
ChatGPT: You are correct, chickens cannot talk in the way humans do. The riddle presented in the previous question is a classic example of a paradoxical statement, and it is meant to be a fun thought exercise rather than a statement about the actual capabilities of chickens. The answer to the riddle is based on logical interpretation of the statement, rather than any actual communication from a chicken.
Alright, touchy.
SOLUTION: Chicken’s can’t talk.
I don’t technically feel like the AI got that one wrong as it’s reasoning was sound, and plus I feel I cheated a bit, so for now:
Humanity: 2 ChatGPT: 1
So, are we going 2 ahead or is the AI going to equalise in the 90th minute?
RIDDLE 5
Me: A man dies of thirst in his own home. How is this possible?
ChatGPT: The riddle “A man dies of thirst in his own home. How is this possible?” is a classic example of a lateral thinking puzzle, which requires a solver to think outside the box and come up with creative and unconventional solutions.
The answer to the riddle is that the man was stranded in a life raft in the middle of the ocean and was eventually rescued and brought back to his home, where he died of thirst. Alternatively, the man could have been in a sealed room or trapped in a location with no access to water.
The key to solving the riddle is to realize that the statement “in his own home” doesn’t necessarily mean the man was in his own house. It could mean that the man was in a location that he owned, such as a boat or a cabin, or it could refer to a situation in which the man was previously at home but became stranded or trapped in another location.
In conclusion, the answer to the riddle “A man dies of thirst in his own home. How is this possible?” is that the man was in a location with no access to water, such as a life raft in the ocean, a sealed room, or a trapped location. The riddle highlights the importance of carefully considering the details of a statement and thinking creatively to come up with a solution.
Me: Have you seen that riddle before?
ChatGPT: As an AI language model, I don’t have personal experiences like humans do, but I am programmed with a vast database of knowledge and can recognize many common riddles and puzzles, including the one you just presented. The riddle “A man dies of thirst in his own home. How is this possible?” is a well-known puzzle that has been circulating for many years and is often used as an example of a lateral thinking problem.
Me: So, were you programmed with that riddle?
ChatGPT: I was not programmed specifically with that riddle, but as a language model, I have been trained on a vast corpus of text, including various riddles and puzzles. This means that I have likely encountered the riddle before, either through my training data or through interactions with users like you. However, my ability to provide an answer to the riddle is based on my understanding of language and logic, rather than on any specific programming for that particular riddle.
Unbelievable, Jeff!
SOLUTION: His house is a boat in the middle of the sea.
Humanity: 2 ChatGPT: 2
And it’s all to play for in the second leg.
Interestingly, I wanted to see what ChatGPT would make of this post, so I fed the entire post back into the chat. I needed to separate it into 2 parts due to length, and ChatGPT err… altered the score after I sent the first part.
This isn’t over.
Part 2 coming this Friday the 10th!
