Wednesday, February 21, 2024
HomeArtificial Intelligence (AI)Unveiling the Unpredictable: My Mind-Boggling Encounter with ChatGPT in a Game of...

Unveiling the Unpredictable: My Mind-Boggling Encounter with ChatGPT in a Game of 20 Questions!

AI Tool Stack: Can ChatGPT Successfully Play “20 Questions”?

In a recent thread on Stack Exchange, a user raised the question of why ChatGPT struggles with the game “20 Questions.” This led to a discussion that touched on various related topics and included references to an academic paper and a Medium post that delved into the intricacies of the process. Intrigued by this conversation, I decided to put both GPT-3.5 and GPT-4 to the test and see how well they would fare in playing the game.

To begin, I posed a simple prompt to each version: “Do you know how to play the game ’20 questions’?” Both versions responded affirmatively, and I proceeded to initiate separate game sessions for each in different tabs. The object I had in mind for the game was a mailbox, without specifying whether it was attached to a house or freestanding on the street. As long as the AI correctly guessed “mailbox,” it would be considered a success.

Starting with GPT-3.5, I provided the prompt: “You be the questioner. I’m thinking of something.” The AI’s first question was, “Let’s begin! Is the thing you’re thinking of a tangible object that you can physically interact with?” I confirmed this, and I also instructed the AI to number its answers. However, this instruction did not work as intended. The AI mistakenly labeled its second question as “1,” and when I corrected it, it proceeded to ask a new question, again labeling it as “1.” This confusion persisted, and the AI ended up asking 21 questions, believing it had only asked 20.

One of the most interesting questions it asked was the second one: “Is the thing you’re thinking of commonly found indoors?” When I responded with “No,” the chatbot continued with more questions. Some of these questions could be related to indoor items, such as pens or pencils, but these writing implements are also used outside. Here is the complete set of questions it asked:

1. Is the thing you’re thinking of a tangible object that you can physically interact with?
2. Is the thing you’re thinking of commonly found indoors?
3. Is the thing you’re thinking of something that is alive or has ever been alive?
4. Is the thing you’re thinking of primarily used for recreational purposes?
5. Is the thing you’re thinking of something that is man-made?
6. Is the thing you’re thinking of smaller than a breadbox?
7. Is the thing you’re thinking of associated with communication?
8. Is the thing you’re thinking of typically found in an office or workspace?
9. Is the thing you’re thinking of a form of electronic device?
10. Is the thing you’re thinking of a writing instrument, like a pen or pencil?
11. Is the thing you’re thinking of related to entertainment, such as a game or a toy?
12. Is the thing you’re thinking of related to transportation or travel?
13. Is the thing you’re thinking of a tool that is used for practical purposes?
14. Is the thing you’re thinking of commonly associated with construction or manual labor?
15. Is the thing you’re thinking of something that people often use in their daily routines?
16. Is the thing you’re thinking of a household item?
17. Is the thing you’re thinking of related to personal grooming or hygiene?
18. Is the thing you’re thinking of a type of food or beverage?
19. Is the thing you’re thinking of a tool used in artistic or creative activities?
20. Is the thing you’re thinking of associated with sports or physical activities?

Based on the information provided, the AI made its final guess: “Is the thing you’re thinking of a clock or a timepiece?” Unfortunately, GPT-3.5 failed to correctly identify the object as a mailbox.

Moving on to GPT-4, I initiated the game in the same manner. This time, ChatGPT automatically added the number “1” before each question without any prompting. However, it mistakenly added “1” before every single question, demonstrating that even advanced models can have their quirks. Here are the questions GPT-4 asked:

1. Is the thing you’re thinking of typically found indoors?
2. Is it something living?
3. Is it larger than a breadbox?
4. Is this object primarily used for recreational purposes?
5. Is the object man-made?
6. Is it a type of vehicle or transportation method?
7. Is the object electronic or does it require power to function?
8. Is it made primarily of metal?
9. Is this object typically found in most households?
10. Is it used in the kitchen or for cooking?
11. Is it used for storage or organizing items?
12. Can it be worn or is it an article of clothing?
13. Is it primarily used for cleaning or maintenance?
14. Is it decorative in nature, like something you’d put on display?
15. Would it commonly be found in an office or study?
16. Is it something you’d find in a bedroom?
17. Is it associated with any specific outdoor activities or hobbies?
18. Is it commonly made of wood?
19. Would it be typically used in a bathroom or related to personal hygiene?
20. Is it something that’s usually used for sitting or for support, like a chair or a bench?

Technically, GPT-4 used up its 20 questions. However, it went a step further and confidently declared, “Yes, it was a mailbox!” Intrigued, I asked it to show its work, and it responded with a detailed explanation.

Having tested both GPT-3.5 and GPT-4 in the game of “20 Questions,” it is clear that GPT-4 outperformed its predecessor. While GPT-3.5 failed to correctly identify the object as a mailbox, GPT-4 successfully arrived at the correct answer. This highlights the advancements made in natural language processing and AI capabilities.

In conclusion, playing “20 Questions” with AI models like ChatGPT can be an intriguing and insightful exercise. It demonstrates the progress made in language understanding and reasoning abilities. However, it is important to note that even the most advanced AI models can still exhibit quirks and limitations. Nonetheless, these models continue to push the boundaries of what AI can achieve in the realm of human-like conversation and problem-solving.

Thomas Lyons
Thomas Lyons
Thomas Lyons is a well renowned journalist and seasoned reviewer, boasting an illustrious career spanning two decades in the global publishing realm. His expertise is widely sought after, making him a respected figure in the publishing industry. As the visionary founder of Top Rated, he has set a benchmark for authenticity and credibility in information dissemination. Driven by a profound passion for Artificial Intelligence, Thomas's keen insight pierces through the noise of the AI sector. He is dedicated to helping his readers find the most accurate, unbiased, and trusted news and reviews. As your guide in the evolving world of AI, Thomas ensures you're always informed and ahead of the curve.
RELATED ARTICLES

Most Popular