Teaching Kids About AI With 1960s Bots: A Curriculum For Creator-Educators

Teach AI Basics With a 1960s Bot: A Reproducible ELIZA Lesson Plan for Creator‑Educators

Hook: You build audience-first learning experiences, but teaching AI to kids often feels scattered: too many tools, jargon-heavy slides, and zero hands-on moments that spark curiosity. Adapt the 1960s ELIZA experiment into a compact, repeatable lesson that creator-educators and workshop leaders can run online or in person to teach AI affordances, computational thinking, and digital ethics—without needing advanced ML knowledge.

Why ELIZA Still Works in 2026

ELIZA — the original rule-based “therapist” chatbot — exposes core truths about conversational AI: simple pattern matching can feel surprisingly human, but it lacks understanding. That cognitive dissonance is a powerful teaching tool. In late 2025 and early 2026, educators and creators leaned into low-tech demos like ELIZA to counter AI hype while meeting new AI literacy requirements and parental privacy concerns. Use ELIZA to show students what AI actually does, not the marketing around it.

“When students chat with ELIZA they uncover how AI works (and doesn’t),” — EdSurge, Jan 2026.

What this lesson teaches: pattern recognition, rules vs. learning, prompt sensitivity, bias and hallucination, privacy basics, and how to prototype simple language systems. It’s perfect for ages 10–16 (with adaptations for younger kids).

Learning Objectives (Workshop-Level)

• Explain the difference between rule-based and statistical/LLM-based chatbots.
• Identify five affordances and three limitations of conversational AI.
• Build or modify a simple ELIZA-style bot and test conversational behavior.
• Reflect critically on where AI should and should not be used.

Before You Teach: Materials & Setup

Choose one delivery mode:

• In-person workshop: Laptops/tablets for pairs, printed worksheets, projector for live demo.
• Online live session: Shared screen for demo, breakout rooms for pairs, a simple web ELIZA or GitHub-hosted sandbox for each student.
• Self-paced course: Video walkthroughs, downloadable code or Scratch projects, quizzes, and a facilitator guide.

Tools options (pick two): a browser-based ELIZA emulator; a simple JavaScript file that runs in-browser (no server); or a Scratch/Snap project for block-based coding. Offer a downloadable worksheet and a one-page teacher checklist.

90‑Minute Reproducible Lesson Plan (Detailed)

0–10 minutes | Hook & Norms

• Start with a one-line provocation: "Can a program be a good listener?"
• Set norms for kindness, privacy (no personal data), and experimentation.

10–25 minutes | Live Demonstration

• Demo ELIZA live on a projector/screen. Use simple prompts: "I am sad," "My teacher...", "I feel..."
• Ask students what felt real and what felt weird. Capture answers on a shared doc or whiteboard.

25–40 minutes | Quick Explain: What ELIZA Is

Use plain language and visuals to explain:

• Rule-based pattern matching: ELIZA looks for keywords and applies canned responses.
• No understanding: It doesn’t model meaning or intent; it matches patterns.
• Contrast with modern LLMs: Large models predict text statistically from data and can produce more flexible but also less predictable answers.

40–65 minutes | Hands-On Pair Activity

1. Pairs alternate: one student chats with ELIZA, the other monitors and notes what triggers certain responses.
2. Each pair records three surprising responses and three obvious failures.
3. Optionally: give students a worksheet to map triggers → response patterns (helps computational thinking).

65–80 minutes | Build & Modify (Maker Phase)

Provide one of two scaffolds depending on skill level:

• No-code/Scratch: A prebuilt Scratch ELIZA project where students can add new trigger-response rules and test.
• Code-light (JavaScript): A 20-line script that matches regex patterns to responses and runs in the browser console. Students edit rules and refresh to test.

80–90 minutes | Share & Reflect

• Pairs present one example of success and one failure. Ask: What did we learn about where this system can help people? Where would it be harmful?
• Close with one action: “Tomorrow I will try…” (e.g., test a family member using the worksheet).

Ready‑to‑Use Script Snippets & Templates

Keep the snippets small and editable. Below are high-level templates creators can drop into slides or code sandboxes.

ELIZA Rule Template (Plain English)

• Trigger: If user message contains "I feel" or "I'm",
• Respond: ask a reflective question or mirror back: "Why do you feel X?"

This template teaches pattern recognition—students list triggers and craft responses, then test how robust the triggers are.

Assessment & Computational Thinking Connections

Assess progress with a simple rubric (score 1–4):

• Understands rule-based systems vs. ML models
• Can map trigger → response rules
• Reflects on ethical/real-world implications
• Can propose a small improvement or extension

Tie each activity to computational thinking skills:

• Decomposition: Break a conversation into parts (greeting, problem, emotion).
• Pattern recognition: Identify repeated keywords that trigger responses.
• Abstraction: Create generalized triggers (e.g., "I feel *")
• Algorithms: Sequence the steps ELIZA follows to choose a reply.

Classroom Variations & Differentiation

Younger students (8–10)

• Shorten session to 45 minutes, use block-based tools, and focus on role-play: kid plays ELIZA and alternates with computer.

Older students (14–16)

• Introduce a mini-research prompt: compare ELIZA to a modern assistant. Have students document failure modes and design an evaluation test.

Remote cohorts

• Use breakout rooms for pairs; share a link to a hosted ELIZA sandbox. Provide a short formative quiz after the session.

Ethics, Safety & Privacy (Non-Negotiables in 2026)

AI literacy now includes understanding risks. Build these guardrails into every session:

• No personal data: Students should never input real names, contact info, or private medical details.
• Explain limits: ELIZA is not a therapist. Reinforce that only trained adults can offer mental health support.
• Data minimization: If you host logs for research, anonymize and get parental/guardian consent—many jurisdictions tightened enforcement after 2024 and 2025 updates to COPPA-like rules.
• Regulatory context: Reference recent policy trends: in 2025–2026 regulators increased scrutiny on deployed AI in schools. Make disclosure and consent part of your lesson pack.

Case Study: Middle School Workshop That Scaled

Example from a real-style run: a creator-educator ran the ELIZA workshop at an after-school club for 60 students. They used a single projector demo, 30 Scratch laptops, and a two-page worksheet. Outcomes:

• 90% of students could explain why ELIZA felt human sometimes.
• Students designed three new trigger rules on average during the maker phase.
• The facilitator repackaged the materials into a paid micro-course and sold seats to two local schools for live runs.

Packaging & Monetization Strategies for Creator‑Educators

Turn this lesson into a sustainable product:

• Sell a DIY workshop kit: slides, worksheet PDFs, code sandboxes, teacher checklist, and a 30-minute recorded demo.
• Offer live workshops for schools or parent groups, priced per session or per student.
• Bundle with competency badges or micro-credentials—partner with local schools or platforms offering certificates.
• License lesson packs to other creators or teachers on marketplaces (provide a clear teacher usage guide and versioning notes to comply with privacy/regulatory changes).

Advanced Extensions (If You Have Time & Skill)

• Build a hybrid: a small rule-based front-end that routes difficult queries to a sandboxed LLM. Use this to teach routing and system design.
• Introduce evaluation metrics: precision of rule matches, false positives, or an ad hoc user satisfaction poll.
• Explore multimodal affordances: what if a voice interface used the same rules? Record and compare expectations.

Practical Tips for Creator‑Educators (From Workshops I’ve Run)

• Keep edits visible: When you change a rule, run a live test so learners can immediately connect code to behavior.
• Encourage hypothesis testing: Have students predict before testing a change—this builds scientific thinking.
• Document failure cases: Make failures part of the deliverable. They drive the best learning moments.
• Offer follow-ups: After the session, email a one-page cheat sheet and a 10-minute challenge to keep engagement high.

Why This Matters in 2026

AI is embedded in kids’ daily lives—from chat features in games to homework assistants. As platforms and regulators evolve rapidly in 2025–26, creator-educators play a critical role teaching practical, hands-on AI literacy. ELIZA-based lessons strip away marketing noise and reveal foundational mechanics in a way that’s safe, cheap, and directly scalable by creators.

Resources to Include in Your Lesson Pack

• One-page teacher checklist (privacy and consent steps)
• Editable slide deck and instructor script
• Code sandbox or Scratch project link
• Student worksheet and rubric
• Follow-up micro-challenge and badge template

Final Checklist Before Your First Run

1. Confirm you have ethical disclaimers and parental consent procedures in place.
2. Pre-test the ELIZA sandbox on devices students will use.
3. Print or upload worksheets and rubrics to your LMS or workshop platform.
4. Prepare a 3-minute wrap-up slide that outlines next steps and additional resources.

Call to Action

Ready to run this as a 45–90 minute workshop or package it as a micro-course? Download the editable lesson pack, code sandboxes, and teacher checklist I use in live runs. If you want a ready-to-run, white-labeled bundle for schools, reply to this post or sign up for the next creator training—space is limited.

Takeaway: ELIZA is more than a historical curiosity—it's a durable, low-risk, high-impact way to teach kids what AI does, why it can fail, and how people design and evaluate intelligent systems. Use the reproducible plan above as your template: demo, hands-on play, build, reflect, and scale.

Related Reading

• Art-Inspired Car Interiors: How to Commission a Bespoke Upholstery Job Influenced by Classical Portraiture
• Portfolio Tips for Interior Design Students: Learn from a $1.8M French Designer Home
• The ‘Very Chinese Time’ Trend: Cultural Sensitivity for Muslim Travelers
• DIY Micro-Business in Your Shed: How a Cocktail Syrup Brand Scaled From a Stove to 1,500-Gallon Tanks
• Composable Quantum Microservices: Building Tiny, Replaceable Q-APIs for Rapid Prototyping