Issue 100 Special: Ask AI a Question No One Has Answered Yet

What you will learn in this article:

• Children develop critical thinking skills by evaluating AI responses.
• They cultivate intellectual humility, understanding AI's current limitations.
• Young learners practice research skills to verify AI-generated information.
• They engage with real-world scientific unknowns, sparking curiosity.
• Students learn to formulate precise questions to elicit specific AI behaviors.

A curious 11-year-old might use AI daily for homework help, yet they may not consider how AI actually "knows" things. A 14-year-old might use AI to draft essays, but has not explored the frontiers of human knowledge where even the smartest algorithms hit their limits. This week, PromptHacker.ai celebrates its 100th issue by inviting young minds to push those boundaries, encouraging a deeper understanding of artificial intelligence and the vastness of human inquiry.

Without early exposure to the nuances of AI, children risk viewing it as an infallible source of truth, rather than a powerful tool with specific capabilities and limitations. They might miss the opportunity to develop the discernment needed to question AI's output, a crucial skill in an increasingly AI-driven world. Understanding where AI's knowledge ends, and human curiosity begins, prepares them for a future where critical evaluation is paramount.

This special activity delivers a hands-on exploration of AI's intellectual humility. Children will craft questions that challenge AI's current understanding, then critically assess the responses. They will discover that even advanced AI models, like ChatGPT or Perplexity, do not possess all knowledge, and learn to distinguish between genuine uncertainty and confident fabrication. This process cultivates a healthy skepticism and a profound appreciation for genuine scientific discovery.

This activity guides children through a fascinating exploration of artificial intelligence, specifically focusing on its current knowledge boundaries. Using widely available AI tools, children will formulate questions that challenge the AI, then critically evaluate its responses. This process highlights the difference between AI's vast data recall and its capacity for true understanding or admission of ignorance. The activity is designed for various age groups, ensuring an appropriate level of engagement and complexity.

Tools Needed:

* ChatGPT (free version) or Perplexity (free version) * A standard household device (phone, tablet, or laptop) with internet access.

Parent or Educator Setup (5 minutes):

Ensure the child has access to either ChatGPT or Perplexity. Briefly explain that AI models are trained on massive amounts of data, but they do not "know" things in the human sense. They predict the next most likely word. This activity explores what happens when a question goes beyond their training data or current human understanding. Emphasize that it is okay for AI (and humans) not to know everything.

Activity for Age 8: The "Mystery Question" Challenge

This activity introduces young children to the concept of AI limitations in a fun, accessible way. They will learn that AI does not always have an answer, and sometimes it might try to guess.

What the Child Produces:

A list of "mystery questions" and the AI's responses, noting whether the AI admitted "I don't know" or tried to make something up.

Step-by-Step Instructions:

• Brainstorm Hard Questions: Ask the child to think of the hardest question they can imagine. Encourage them to think about things no one could possibly know, or things that are silly and impossible. Examples include: "What does a rainbow taste like?", "What will my pet dream about tonight?", "What is the exact number of grains of sand on Earth?".
• Open the AI Tool: Help the child open ChatGPT or Perplexity on the chosen device.
• Ask the AI: Type one of the child's mystery questions into the AI chatbox. Press Enter.
• Observe the Answer: Read the AI's response together.
• Evaluate the Response: Discuss with the child: Did the AI say, "I don't know"?
• Did it say something like, "That's a fun question, but AI cannot know that"?
• Did it try to give an answer that sounds like it made something up?
• Did it give a general answer without directly addressing the question?
• Record Findings: On a piece of paper, have the child write down their question and a simple note about what the AI did (e.g., "Said it didn't know," "Made up a silly answer," "Gave a general explanation").
• Repeat: Try 2-3 more mystery questions. Observe if the AI's behavior changes.

Parent/Educator Support:

Guide the child in formulating questions. Help them interpret the AI's response without judgment. Emphasize that finding out what AI *cannot* do is just as important as finding out what it *can* do.

Common Mistakes:

Children might ask questions the AI *can* answer (e.g., "What is the capital of France?"). Gently redirect them to questions that are truly unknown or subjective. Another mistake is expecting a perfect "I don't know." Explain that AI often tries to provide *some* information, even if it cannot directly answer the unknown.

Worked Example (Age 8):

A child named Leo (age 8) asked ChatGPT, "What color is the sound of a drum?" ChatGPT responded: "That's a super creative and imaginative question, Leo! Sounds don't actually have colors, because they are vibrations that our ears hear, not light that our eyes see. So, the sound of a drum doesn't have a color. But if you were to imagine it, what color would *you* think it is?" Leo was surprised. He thought the AI might just say "red" or "blue." He learned that the AI explained *why* it could not answer, and even turned the question back to him. This showed him AI has limits but can also be clever in its responses.

Activity for Age 11: Unanswered Science Mysteries and AI's Take

This activity challenges older children to engage with real scientific unknowns, using AI as a starting point for research. They will develop research skills and a deeper appreciation for the scientific process.

What the Child Produces:

A short report (1-2 paragraphs) for each question, detailing the question, the AI's explanation, and the actual scientific status of the question based on online research.

Step-by-Step Instructions:

• Identify Unanswered Scientific Questions: Encourage the child to think about big questions science has not yet fully answered. Guide them with categories like space, biology, physics, or earth science. Examples include: "What is dark matter?", "How did life begin on Earth?", "Are we alone in the universe?", "What causes gravity at a quantum level?", "How do animals navigate over vast distances?".
• Formulate the AI Prompt: For each question, craft a clear prompt for the AI. For example: "What is dark matter, and what is its current scientific status? Has science answered this question completely?"
• Ask the AI: Input the prompt into ChatGPT or Perplexity.
• Analyze AI's Explanation: Read the AI's response carefully. Does it provide a definitive answer? Does it mention ongoing research or theories? Does it admit uncertainty?
• Research the Real Status: Use reputable online sources (e.g., NASA, scientific journals for kids, university websites, National Geographic) to research the actual scientific status of the question. Look for current theories, ongoing experiments, and areas of debate among scientists.
• Compare and Contrast: Compare the AI's explanation with the information found in research. Was the AI's explanation accurate but simplified?
• Did the AI present theories as facts, or did it acknowledge uncertainty?
• Did the AI miss any key aspects of the current scientific understanding?
• Write a Brief Report: For each question, write a short report summarizing: The scientific question.
• What the AI said (briefly).
• What real-world research revealed about the question's current status (e.g., "Scientists have several theories but no definitive proof yet," "Ongoing experiments are trying to detect it").

Parent/Educator Support:

Help the child brainstorm questions. Guide them towards reliable research sources online, teaching them how to evaluate website credibility. Discuss the difference between a theory and a proven fact, and how science progresses through questioning and experimentation.

Common Mistakes:

Children might rely solely on the AI's answer without cross-referencing. Emphasize the importance of verification. They might also pick questions that are too simple or too complex. Help them refine their choices.

Worked Example (Age 11):

A child named Maya (age 11) chose the question, "How do birds migrate thousands of miles without getting lost?" She prompted Perplexity: "Explain how birds migrate and if scientists fully understand it." Perplexity's response detailed magnetic sensing, visual cues, and olfactory navigation, stating that while much is known, the full complexity is still being studied. Maya then researched online, finding articles from Cornell Lab of Ornithology and National Geographic. She discovered that scientists are actively researching specific mechanisms, like how birds "see" magnetic fields (cryptochrome proteins) and the role of inherited migratory routes. She noted that Perplexity's explanation was good, but her research highlighted the specific, ongoing scientific work and the precise questions scientists are still trying to answer.

Activity for Age 14: AI Explanations and the Edge of Human Discovery

This activity challenges teenagers to engage with complex scientific frontiers, using AI to synthesize information and then critically evaluate humanity's progress toward solutions. It connects AI's capabilities to real-world scientific endeavors and potential career paths in research.