Loading…
TechNest
Use AI assistants effectively
Explore and get curious
2 steps
Try things, experiment
2 steps
Go deep, master it
2 steps
Explore & Discover
Open ChatGPT (chat.openai.com — free), Google Gemini (gemini.google.com — free), and Microsoft Copilot (copilot.microsoft.com — free) and ask all three the exact same question: "What is the best hiking trail near Salt Lake City for a beginner?" Compare the three answers side by side. Notice how they write differently, give different recommendations, and have different levels of detail. Then ask something totally random — "Can you explain black holes like I'm 10 years old?" — and see what you get. You're not looking for perfect answers right now; you're just getting a feel for how these tools respond and where they seem confident vs. uncertain. You're ready for the next step when you can describe one difference you noticed between two different AI assistants after asking them the same question.
Learn the Basics
Learn the difference between a vague prompt and a specific one. A vague prompt is "write a story." A specific prompt is "write a 200-word adventure story for a 12-year-old set in Arches National Park where the main character discovers a hidden trail." Better prompts get you better results — this is called prompt engineering. Watch the free Khan Academy course lesson on "How AI works" at khanacademy.org to understand what's actually happening inside these tools. Then learn about hallucination — when an AI confidently states something that's completely wrong. Test it: ask an AI about a fake event you made up and see if it plays along. This is why you always fact-check AI answers. You're ready for the next step when you can write one vague prompt and one specific version of the same prompt and explain why the specific one would get better results.
Build Your First Project
Use an AI assistant to help you with a real project you actually need to do this week — studying for a test, planning a project, writing an email, or understanding something confusing in class. Don't just accept the first answer — have a back-and-forth conversation. If the answer is too long, say "make it shorter." If it's confusing, say "explain it differently." If it's missing something, say "add more about X." This conversation style is called "iterative prompting" and it's the key skill that separates people who get great results from people who give up after one try. Screenshot your best conversation exchange. You're ready for the next step when you've had a multi-turn AI conversation that solved a real problem and you can show the conversation thread.
Experiment & Iterate
Explore three specific types of AI prompts: the role prompt, the format prompt, and the constraint prompt. A role prompt tells the AI to act as someone specific ("act as a tour guide for Utah National Parks"). A format prompt tells it how to structure the output ("respond in bullet points with three examples each"). A constraint prompt sets limits ("explain this in under 50 words using no jargon"). Practice all three techniques this week by picking three different tasks — one creative, one informational, one practical — and applying a different prompt type to each. Document what changed when you used these techniques vs. when you didn't. You're ready for the next step when you can demonstrate all three prompt types with real examples and explain what each one controls.
Advanced Techniques
Learn to think critically about what AI can and can't do well. AI is great at generating ideas, summarizing, explaining, and drafting — but it's bad at math, recent news, personal opinions, and knowing things specific to your life. Use Google's free tool "Teachable Machine" (teachablemachine.withgoogle.com) to actually train a simple AI yourself — this gives you a real peek inside how machine learning works. Then try using an AI assistant to help you do something it's bad at (like solving a multi-step math problem) and verify its answer yourself. Notice where it went wrong and why. Also check out AI-powered tools you might already use: Spotify's recommendations, YouTube's suggestions, and Google Maps' routing all run on AI. You're ready for the next step when you can list three things AI does well and three things it does poorly, with a real example of each from your own testing.
Final Project Showcase
Create an "AI Tools Guide" for someone your age who has never used AI before. Think of a younger sibling, a friend, or a classmate. Your guide should cover: what AI tools are free to use, how to write a good prompt (with three examples), what to watch out for (hallucinations, privacy), and one thing AI is actually useful for in school. Write it using Google Docs (free) and include real screenshots from your own conversations as examples. Keep it under two pages — clear and friendly, not a textbook. Share it with at least one real person and ask them if they could actually follow it. If they get confused anywhere, revise that section. You're ready for the next step when your guide has been read by at least one other person and you've made at least one revision based on their feedback.
Recommended materials and resources for this quest.
Composition Notebook for AI Experiments
RequiredKeep a running log of every prompt you try, what happened, and what you'd change. This becomes your personal prompt library. Composition notebooks are durable and don't fall apart in your backpack.
amazon
$3–7
Sticky Notes Multi-Pack
RequiredUse sticky notes to capture quick observations as you test AI tools — what surprised you, what was wrong, what was great. Stick them in your notebook next to the prompt that caused the reaction.
amazon
$4–9
AI and Machine Learning for Coders Book
If you want to understand how these tools actually work under the hood — not just use them — this book explains neural networks and machine learning in plain language with real code examples.
amazon
$25–40
Some links may be affiliate links. We may earn a small commission at no extra cost to you.