AI Literacy & Agent Building

Learning objectives: Know the major categories of AI tools, what each is actually for, and how to pick the right tool for a task.

Topics & tools to demo:

Conversational assistants: Claude, ChatGPT, Gemini — differences in strengths (writing, reasoning, coding, voice mode)

Search-augmented AI: Perplexity, ChatGPT/Claude with web search — when to trust vs. verify

Image generation: Midjourney, DALL·E, Google's image tools — prompting for style, composition

Video/audio: Runway, ElevenLabs, Suno — quick demo of what's possible now

Coding assistants: GitHub Copilot, Claude Code, Cursor — AI as a coding partner, not a replacement for learning to code

Productivity/office AI: Notion AI, Claude for Excel/PowerPoint/Word, Google Workspace AI

Automation platforms: Zapier, n8n, Make — "AI that does things," a bridge into Module 4

Learning objectives: Know the major categories of AI tools, what each is actually for, and how to pick the right tool for a task.

Topics & tools to demo:

• Conversational assistants: Claude, ChatGPT, Gemini — differences in strengths (writing, reasoning, coding, voice mode)
• Search-augmented AI: Perplexity, ChatGPT/Claude with web search — when to trust vs. verify
• Image generation:Midjourney, DALL·E, Google's image tools — prompting for style, composition
• Video/audio: Runway, ElevenLabs, Suno — quick demo of what's possible now
• Coding assistants:GitHub Copilot, Claude Code, Cursor — AI as a coding partner, not a replacement for learning to code
• Productivity/office AI: Notion AI, Claude for Excel/PowerPoint/Word, Google Workspace AI
• Automation platforms: Zapier, n8n, Make — "AI that does things," a bridge into Module 4

-software builder: Lovable

Slides Used by the instructor for The AI Tools Landscape slides presentation

The Top brainer song that was made during this The AI Tools Landscape

Learning objectives: Write prompts that reliably get better outputs. Understand why vague prompts fail.

Topics:

The anatomy of a good prompt: task, context, constraints, format, examples

Being specific vs. vague — side-by-side comparisons

Giving the model a role/persona and why it helps

Providing examples (few-shot prompting)

Asking for step-by-step reasoning ("think step by step") and why it improves accuracy

Specifying output format (bullet points, table, word count, tone)

Iterating: treating the first response as a draft, not a final answer

Learning objectives: Write prompts that reliably get better outputs. Understand why vague prompts fail. Topics: The anatomy of a good prompt: task, context, constraints, format, examples Being specific vs. vague — side-by-side comparisons Giving the model a role/persona and why it helps Providing examples (few-shot prompting) Asking for step-by-step reasoning ("think step by step") and why it improves accuracy Specifying output format (bullet points, table, word count, tone) Iterating: treating the first response as a draft, not a final answer

Slides used by the instructor in Session 3: Prompt Engineering Slides

Learning objectives: Handle more complex, multi-step tasks with AI. Understand prompt limitations and failure modes.

Topics:

Chaining prompts (breaking a big task into smaller steps)

Using AI to critique/improve its own output

Negative examples ("don't do X") and why models sometimes ignore them

System prompts vs. user prompts (relevant once they start building)

Common failure modes: overconfidence, made-up citations, math errors, outdated info

Prompt injection — a light intro to why you shouldn't blindly trust text an AI reads from the web

Learning objectives: Handle more complex, multi-step tasks with AI. Understand prompt limitations and failure modes.

Topics:

• Chaining prompts (breaking a big task into smaller steps)
• Using AI to critique/improve its own output
• Negative examples ("don't do X") and why models sometimes ignore them
• System prompts vs. user prompts (relevant once they start building)
• Common failure modes: overconfidence, made-up citations, math errors, outdated info
• Prompt injection — a light intro to why you shouldn't blindly trust text an AI reads from the web

• Chaining prompts (breaking a big task into smaller steps)
• Using AI to critique/improve its own output
• Negative examples ("don't do X") and why models sometimes ignore them
• System prompts vs. user prompts (relevant once they start building)
• Common failure modes: overconfidence, made-up citations, math errors, outdated info
• Prompt injection — a light intro to why you shouldn't blindly trust text an AI reads from the web

Learning objectives: Understand the conceptual leap from "chatbot that talks" to "agent that acts."

Topics:

Definition: an agent = an LLM + tools + memory + a loop that lets it take actions and observe results

The difference between a chatbot, a workflow/automation, and a true agent

Core components: instructions/system prompt, tools (APIs, search, code execution), memory, planning/reasoning loop

Real-world examples: customer support agents, coding agents (like Claude Code), research agents, personal assistants

Why agents are powerful and why they're risky (they can take real actions — send emails, spend money, delete files)

**Learning objectives:** Understand the conceptual leap from "chatbot that talks" to "agent that acts."

**Topics:**

• Definition: an agent = an LLM + tools + memory + a loop that lets it take actions and observe results
• The difference between a chatbot, a workflow/automation, and a true agent
• Core components: instructions/system prompt, tools (APIs, search, code execution), memory, planning/reasoning loop
• Real-world examples: customer support agents, coding agents (like Claude Code), research agents, personal assistants
• Why agents are powerful *and* why they're risky (they can take real actions — send emails, spend money, delete files)

**Activity:** Watch/discuss a short demo of an agent completing a multi-step task (e.g., booking research, filling a form, writing+running code). Map out loud what tools and steps it must have used.

Module 5 (Week 3, Session 1): What Are AI Agents?

• Definition: an agent = an LLM + tools + memory + a loop that lets it take actions and observe results
• The difference between a chatbot, a workflow/automation, and a true agent
• Core components: instructions/system prompt, tools (APIs, search, code execution), memory, planning/reasoning loop
• Real-world examples: customer support agents, coding agents (like Claude Code), research agents, personal assistants
• Why agents are powerful *and* why they're risky (they can take real actions — send emails, spend money, delete files)

Watch/discuss a short demo of an agent completing a multi-step task (e.g., booking research, filling a form, writing+running code).

We build a website using Lovable

Learning objectives: Build a first simple agent/automation without writing code, understanding triggers, actions, and logic.

Topics:

Intro to no-code automation platforms (n8n, Zapier, Make)

Triggers → actions → logic (if/then) as the basic building blocks

Connecting an LLM step into a workflow (e.g., "summarize this form response and send it to Slack/Discord/email")

Where agents get their "memory" (a spreadsheet, database, or note-taking step)

Learning objectives:** Build a first simple agent/automation without writing code, understanding triggers, actions, and logic.

Topics:

• Intro to no-code automation platforms (n8n, Zapier, Make)
• Triggers → actions → logic (if/then) as the basic building blocks
• Connecting an LLM step into a workflow (e.g., "summarize this form response and send it to Slack/Discord/email")
• Where agents get their "memory" (a spreadsheet, database, or note-taking step)

Build a simple working automation together, e.g.:

> "When someone fills out a Google Form, send their answers to an AI model to generate a personalized response, then email it back to them."

Everyone leaves with one functioning no-code agent.

Slides For Agent building No-code

Slides For Agent building No-code

Slides For Agent building No-code

Various examples of AI automation scenario and a tutorial fo how to build n8n automation with Google form

Learning objectives: Understand what's happening "under the hood" of a no-code agent by building a minimal version with code.

Topics:

Intro to calling an LLM API directly (using Python or JavaScript, whichever the group is more comfortable with — or teach both are similar)

The request/response loop: system prompt, user message, model response

Giving the model a "tool" it can call (e.g., a weather lookup, a calculator) and how tool-calling works

The agent loop: model decides → tool runs → result goes back to model → model decides again

This lesson is understanding last session workflow what is realy happening under each component and how they communicate to each other we had a time to troubleshoot issue with sending email back to the user

TAO delivery ready

Learning objectives: Understand what agents can and can't (or shouldn't) do, and how to build them responsibly.

Topics:

What agents are good at today: research, coding, scheduling, data entry, customer support, content drafting

What they're bad at: long-horizon planning, high-stakes irreversible actions, situations needing real judgment or accountability

Guardrails: human-in-the-loop approval steps, permissions/scopes, rate limits, logging

Real failure stories (agents that spent money incorrectly, leaked data, or looped infinitely) — brief, concrete examples

Cost awareness: API calls cost money; agents can rack up bills fast if not monitored

Understand what agents can and can't (or shouldn't) do, and how to build them responsibly. have an observation on Openclaw

Understand what agents can and can't (or shouldn't) do, and how to build them responsibly.

Topics:

• What agents are good at today: research, coding, scheduling, data entry, customer support, content drafting
• What they're bad at: long-horizon planning, high-stakes irreversible actions, situations needing real judgment or accountability
• Guardrails: human-in-the-loop approval steps, permissions/scopes, rate limits, logging
• Real failure stories (agents that spent money incorrectly, leaked data, or looped infinitely) — brief, concrete examples
• Cost awareness: API calls cost money; agents can rack up bills fast if not monitored
• Understanding Openclaw

Learning objectives: Use AI critically and responsibly — as citizens and future professionals, not just as tool-users.

Topics:

Where bias comes from in AI systems (training data, whose voices are represented) and what it looks like in practice

Academic integrity: using AI to learn vs. using it to avoid learning; how post-secondary institutions in Canada are handling AI policies

Data privacy: what happens to what you type into a chatbot; reading a privacy policy in 2 minutes

Misinformation and deepfakes: how to spot AI-generated images/video/audio, and why that skill matters more each year

Environmental and labor costs of AI (energy use, data-labeling work) — a balanced, non-alarmist overview

Intellectual property: who owns AI-generated content, and copyright basics

Use AI critically and responsibly — as citizens and future professionals, not just as tool-users.

• Where bias comes from in AI systems (training data, whose voices are represented) and what it looks like in practice
• Academic integrity: using AI to learn vs. using it to avoid learning; how post-secondary institutions in Canada are handling AI policies
• Data privacy: what happens to what you type into a chatbot; reading a privacy policy in 2 minutes
• Misinformation and deepfakes: how to spot AI-generated images/video/audio, and why that skill matters more each year
• Environmental and labor costs of AI (energy use, data-labeling work) — a balanced, non-alarmist overview
• Intellectual property: who owns AI-generated content, and copyright basics

This is your handout but it is also an exercise to help you understand this Lesson If you answer different questions in this shared document alongside with the Lecture shared in the video you will understand the legality around AI and the responsible way to use it either in academic or professional world

Learning objectives: Understand how AI is reshaping the job market they're about to enter, and how to position themselves.

Topics:

Which skills are becoming more valuable (judgment, communication, taste, knowing what to ask for) vs. less valuable (rote execution)

"AI-native" vs. "AI-resistant" career paths — an honest, non-hype discussion

Using AI as a career tool: résumé/cover letter drafting (with a human-editing rule), interview prep, learning new skills faster

A quick look at building something (a small business, a side project, a portfolio) using AI tools cheaply — relevant for entrepreneurial-minded students

Where to keep learning: reputable newsletters, courses, communities (avoid hype-chasing)

Understand how AI is reshaping the job market they're about to enter, and how to position themselves. Topics: Which skills are becoming more valuable (judgment, communication, taste, knowing what to ask for) vs. less valuable (rote execution) "AI-native" vs. "AI-resistant" career paths — an honest, non-hype discussion Using AI as a career tool: résumé/cover letter drafting (with a human-editing rule), interview prep, learning new skills faster A quick look at building something (a small business, a side project, a portfolio) using AI tools cheaply — relevant for entrepreneurial-minded students Where to keep learning: reputable newsletters, courses, communities (avoid hype-chasing)

Today's lesson slides

Learning objectives: Apply everything learned to build one complete, working AI agent or AI-powered tool.

Format: Individually you pick and build one of:

A no-code agent that automates something real in their life (school, job search, hobby)

A small coded agent with at least one custom tool

A well-documented "AI tool comparison guide" for a specific use case, with prompt templates included

Learning objectives: Apply everything learned to build one complete, working AI agent or AI-powered tool. Format: Individually you pick and build one of: A no-code agent that automates something real in their life (school, job search, hobby) A small coded agent with at least one custom tool A well-documented "AI tool comparison guide" for a specific use case, with prompt templates included

Learners handouts

Learning objectives: Present work clearly, get feedback, reflect on the course.

Format:

You do a 5-minute demo: what you built, what tools/prompts used, what broke and how was it fixed it

Group discussion: what surprised you most about AI this course? What are you worried about? What are you excited about?