- How the internet physically works — cables, servers, data centers
- What happens when you type to ChatGPT (the full journey)
- AI vs Search Engine — the critical difference
- What AI can do vs cannot do — 5 examples each
- Why AI is a soft skill just like English or Excel
- Real Indian companies using AI today (Infosys, HDFC, Zomato, etc.)
- Sign up for ChatGPT + Claude (both free) — 10 min
- Ask AI: "Plan a 5-day trip to Goa for a college student with ₹10,000 budget"
- Ask AI: "Write a 3-line intro about me: [student fills their own details]"
- Note what surprised you — share with the class in 1 sentence each
Before opening ChatGPT, show submarinecablemap.com on the projector. Zoom into India — show the cables entering Mumbai and Chennai. Ask: "When you just typed that message to ChatGPT, where did it physically go?" Watch students realize AI is not magic — it's infrastructure.
💡 Script: "This message traveled through one of these underwater cables at the speed of light, hit a server farm in the US, was processed by a model trained on half the internet, and came back in 2 seconds. That's what we're learning to control this month."
- Connect to Day 1: "You know what AI is — now learn to talk to it"
- Why most people get bad AI results — vague, one-line prompts
- The CRTF Framework: Context, Role, Task, Format
- Live comparison: bad → good → great prompt (same topic, show improvement)
- Tone modifiers: formal / casual / urgent / friendly
- Exercise: Write a job application email for any company you want to work at
- Round 1: Write prompt yourself without any framework — see output
- Round 2: Use CRTF framework to rewrite the same prompt
- Compare both outputs — write 2 differences you noticed
- Homework: use CRTF for 3 different tasks tonight, screenshot results
Open two browser tabs (ChatGPT in both). Tab 1: type "write email" — show the useless output. Tab 2: type the full CRTF version: "Act as an HR consultant. Write a formal job application email from a final-year computer science student applying for a software developer internship at TCS. Include: enthusiasm for the company, 2 relevant skills, a clear CTA. Keep under 150 words." Show transformation live.
💡 Script: "AI is only as smart as the question you ask. A bad prompt gives a lazy intern's answer. A CRTF prompt gives a senior consultant's answer. Same AI. Different result. That's the entire game."
- Connect to Day 2: "CRTF has a Role component — today we go deep on that"
- Why "Act as a..." changes everything — AI responds differently based on role
- 10 powerful roles: HR expert, startup founder, professor, doctor, lawyer, coach...
- Stacking roles: "Act as a senior HR at an IT company who also mentors freshers"
- Custom AI personas for specific needs
- Exercise: "Ask the same question to 3 different roles" challenge
- Question: "How should I prepare for my first job interview?"
- Role 1: Act as an HR Manager at Infosys
- Role 2: Act as a career coach who works with college students
- Role 3: Act as a fresher who just cracked their first interview
- Compare outputs — which role gave the most useful advice and why?
Write 8 roles on chits: Startup Founder, Doctor, Lawyer, Professor, Financial Advisor, Personal Trainer, Chef, Movie Director. Each student picks a chit. They ask AI to "Act as a [their role]" and answer: "How does your job change with AI?" — then they present the AI's answer to the class as if they ARE that person. Gets loud and fun.
💡 Script: "You now have access to every type of expert, for free, available instantly. A lawyer's consultation costs ₹5,000/hour. A doctor's second opinion costs ₹2,000. With the right role prompt, you get expert-level guidance in 10 seconds."
- What "context" means to AI — it only knows what's in the conversation
- Word embeddings: how AI converts words into numbers (vectors)
- Why "King − Man + Woman = Queen" works in AI math
- Chain-of-thought prompting: "Think step by step" magic phrase
- Multi-turn conversations — using follow-up messages to go deeper
- Asking AI for its reasoning: "Explain why you said that"
- Exercise: Solve one real problem using 4-turn conversation
- Turn 1: "How do I prepare for placement season starting in 3 months?"
- Turn 2: "Go deeper on the resume part — I have no internship experience"
- Turn 3: "Make this into a day-by-day action plan for the next 2 weeks"
- Turn 4: "What risks might make this plan fail? How do I handle them?"
- Write down: how did quality change across 4 turns?
Open projector.tensorflow.org on the projector. Search for "king" — watch similar words cluster nearby. Search "cricket" — see "sport", "bat", "India" appear close. Search "Mumbai" — see "Delhi", "city", "India" group together. This is the actual math AI uses to understand meaning — visualized in real 3D space, live in class.
💡 Script: "Every word you type to ChatGPT gets converted into a point in a space with 1,536 dimensions. We can only show 3 here. But this is literally how AI understands that your message about a job interview is completely different from one about cricket. More context = AI finds the right region of this space faster and more precisely."
- Overview: ChatGPT (general), Claude (long/nuanced), Gemini (Google+real-time), Kimi (huge docs)
- Free tiers: what each gives you for free — what students can access
- Strengths and weaknesses: when each one fails
- Decision framework: "Which AI do I open for this task?"
- Preview of next week: Perplexity and NotebookLM (for research)
- Exercise: AI Olympics — 4 open tabs, same 3 prompts
- Prompt A: "Explain blockchain in 3 sentences using a vegetable market analogy"
- Prompt B: "Give me 5 business ideas for a student in India with ₹5,000"
- Prompt C: "Write a 2-line joke about college placements"
- Score each AI: 1-5 on clarity, creativity, usefulness
- Fill in personal "Which AI for What" reference card
Open 4 browser tabs side by side on the projector. All 4 tools, same prompt: "Explain quantum computing to a 15-year-old using cricket as an analogy." Send all 4 simultaneously. Read the responses aloud. Students immediately feel how each AI has a different "personality", depth, and style. This is always one of the most memorable moments of the course.
💡 Script: "These aren't just different websites. They're genuinely different models with different strengths — like four different experts you have on speed dial. Knowing which one to call for which problem is a skill in itself."
- Use ChatGPT with role prompt: "Act as a career advisor with 10 years at [company]". Ask: "What does it take to get hired here as a fresher?"
- Use Gemini: "What are the latest news and updates about [company] in 2024?"
- Use Claude: "Based on this info, write a 200-word career strategy for a fresher applying to [company]"
- Use CRTF to write the ideal job application email for this company
- Use chain-of-thought: "What are 5 questions I might be asked in the interview at [company]? Think step by step."
- Each team presents their Career Brief (2 min per team) — class votes on most impressive
AI sometimes confidently states wrong facts. It doesn't "know" — it predicts the most likely next word. Show a live example: ask ChatGPT a specific fact question, then verify on Google. Explain: "In Week 2, we learn Perplexity — which shows you sources. This is why it matters." Students learn to always cross-check AI answers on important facts. This bridges Week 1 (using AI) to Week 2 (researching with AI responsibly).
- Connect to Saturday: "Last week AI lied to us. Today we use an AI that shows its sources."
- What is Perplexity — how it differs from Google and ChatGPT
- Search modes: Quick / Detailed / Academic
- Follow-up question chains — research conversation with the internet
- Collections / Spaces — organize research by topic
- How to verify citations: click source, check if AI summarized correctly
- Exercise: "Company Brief" — research any company you want to work at
- Step 1: Search "[Company name] culture, salary for freshers, interview process"
- Step 2: Follow-up: "What skills do they look for in freshers specifically?"
- Step 3: Follow-up: "What is the latest news about this company in 2024?"
- Fill a 5-point brief: Culture / Pay / Skills needed / Interview style / Recent news
- Also: same search on Google — compare time taken and quality
Ask the class: "Who wants to work at Google / Infosys / [top answer]?" Take the first response. Do a Google search — show the messy ads and irrelevant results. Then open Perplexity: "Tell me about [Company]: their engineering culture, typical salary for freshers in India, interview rounds, and any recent layoffs or hiring news." A structured, cited brief appears in 8 seconds.
💡 Script: "Before any interview — spend 20 minutes here. You'll walk in knowing things the interviewer didn't expect you to know. That is the game."
- Connect to Day 7: "Yesterday = quick research. Today = deep, automated research."
- Gemini advantage: Google-connected, real-time information, multimodal
- Gemini Deep Research — enable it, set scope, watch it run autonomously
- Article analysis: paste URL → ask questions about the content
- Multi-source comparison: "Which argument has stronger evidence?"
- Perplexity vs Gemini: quick decision guide for when to use which
- Exercise: Deep research on a topic relevant to your final project (Week 5)
- Pick one topic from: AI in healthcare / AI in education / AI in finance / AI in retail
- Run Gemini Deep Research on it — observe the process live
- Find 2 conflicting perspectives in the output — note them
- Ask Gemini: "Which side has stronger evidence in this debate?"
- Save the research brief — you'll use this in Week 5 final project
Enable Gemini Deep Research. Type: "What is the future of IT jobs in India for fresh graduates between 2025 and 2030?" Don't touch anything. Project the screen. Let the class watch as Gemini autonomously searches 15-20 sources, shows its research process live, and produces a multi-page structured report. It feels like watching a robot do someone's homework in real time.
💡 Script: "A McKinsey consultant would spend 2 full days doing what you just watched happen in 3 minutes. Your new job isn't to collect information — it's to ask the right question and evaluate the answer intelligently."
- Connect to Day 8: "Gemini researches the web. NotebookLM researches YOUR files."
- What is NotebookLM — Google's AI that reads your private documents
- Upload sources: PDF, Google Doc, URL, YouTube video, pasted text
- Auto-generate: Study Guide, FAQ, Timeline, Briefing Document
- Audio Overview — AI-generated podcast from your notes (jaw-drop feature)
- Kimi as alternative when documents are very large (200k+ words)
- Exercise: Upload any subject PDF → full study kit in 10 minutes
- Upload a textbook chapter or lecture PDF you have on your phone
- Ask 5 questions about the document — get instant cited answers
- Click "Generate Study Guide" — download it
- Generate Audio Overview — listen for 2 minutes
- Write: "How would I normally study this chapter vs how I just did it?"
The night before, upload a chapter from a well-known textbook to NotebookLM and generate an Audio Overview. Walk into class, plug in speakers, and play it without saying anything. Students hear two natural-sounding AI voices having a podcast conversation about the chapter — with examples, stories, and back-and-forth discussion. After 60 seconds, pause it and ask: "What was that?" Then reveal: a 40-page PDF, uploaded 5 minutes ago.
💡 Script: "Your commute to college is 30-40 minutes. Your textbook has 12 chapters. Every single chapter can become a podcast you listen to on the bus. Your commute just became study time."
- Connect to Day 9: "NotebookLM reads docs. Today: AI drills you on what you read."
- Active recall vs passive reading — why testing beats re-reading (science)
- Generating MCQs, True/False, Short answer from any content
- The Feynman Technique with AI: explain → get corrected → re-explain
- Personalized revision plan: tell AI your exam date + weak areas → get schedule
- One-page cheat sheet generation for any chapter or exam
- Exercise: 3-part study session for one hard subject
- Part 1: Generate 15 MCQs from a chapter → take the quiz → score yourself
- Part 2: Pick 1 concept you got wrong → explain it to Claude in simple words → get feedback
- Part 3: Ask AI: "Create a 1-page cheat sheet for [chapter/topic] covering all formulas, definitions, key points, and common exam mistakes"
- Compare your performance: did you know more or less than you thought?
Ask for a volunteer. They pick any concept from any subject. They explain it to Claude in simple language — as if explaining to a 10-year-old. Then use the prompt: "Rate this explanation 1-10. Identify exactly 3 gaps in understanding. Provide the correct, complete explanation at the same simple level." AI's feedback is often more precise than a teacher's. The class always relates deeply to seeing a peer corrected by AI in real time.
💡 Script: "The best way to find out if you truly understand something is to explain it simply. AI never gets bored, never gets impatient, and gives you feedback at 2am the night before your exam. Use it every single time you study something new."
- What is a Personal Knowledge Management (PKM) system?
- The 6-step AI Learning Workflow: Discover → Organize → Understand → Test → Synthesize → Apply
- Which tool at which step: Perplexity → NotebookLM → Claude/ChatGPT
- Live demo: complete workflow for one topic in 25 min on the projector
- Building habits: how to use this workflow weekly, not just for exams
- Exercise: Complete full 6-step workflow on your chosen topic
- Step 1 DISCOVER: Perplexity → research topic, save top 3 sources
- Step 2 ORGANIZE: Upload sources to NotebookLM → create notebook
- Step 3 UNDERSTAND: Ask 5 questions to NotebookLM → generate study guide
- Step 4 TEST: Use ChatGPT → 10 MCQs → take quiz → note weak spots
- Step 5 SYNTHESIZE: Claude → "Summarize all of this in a 200-word expert brief"
- Step 6 APPLY: Write 3 ways you'll use this knowledge in your life/career
Pick a topic nobody in the class knows — try "How does India's UPI payment system work technically?" Start a timer visible on screen. Complete all 6 steps live: Perplexity research → paste into NotebookLM → study guide → quiz yourself → Claude synthesis. At 25 minutes, present the knowledge brief like a confident expert. Students go silent when they realize they can do this for any topic in any subject in 25 minutes.
💡 Script: "Students who build this habit become the ones who always seem to know more than everyone else in the room. Because they actually do."
- Choose a topic from provided list OR their own (must be something they don't already know)
- Run the 6-Step AI Knowledge Workflow: Perplexity → NotebookLM → Claude synthesis
- Generate 3 surprising facts and 1 controversial point from the research
- Prepare a 5-minute verbal presentation — no slides, just speak from their brief
- Present to class: cover the topic, 3 surprising facts, 1 practical application
- Class Q&A: other students ask 1-2 questions — presenter answers using their brief
Now that students can research, write, and study with AI, they need to understand what THEY still need to bring. AI can write an email, but humans need to build relationships. AI can analyze data, but humans need to make judgment calls. Introduce "AI + Human Collaboration" as the real skill. This bridges Week 2 (using AI for knowledge) to Week 3 (using AI for professional workplace tasks where human judgment is still critical). Exercise: "Ask Claude: what can't you do as well as a human?" — discuss answers as a class.
- Bridge from Week 2: "You can research and study. Now we put those skills to work."
- Professional email anatomy: subject, salutation, context, body, CTA, sign-off
- The 6 types: job application / 7-day follow-up / post-interview thank you / networking intro / polite decline / complaint
- Subject line optimization — what makes someone open vs ignore an email
- AI email editor: paste your bad email → AI transforms it
- Exercise: Write all 3 of these real emails using AI
- Email 1: Job application to your dream company (use company research from Week 2)
- Email 2: Follow-up email exactly 7 days after submitting that application
- Email 3: Thank you email after an imaginary interview yesterday
- For each: first write subject line, then full email using CRTF from Week 1
- Share best subject line from each email with the class
Ask: "Has anyone sent a work or application email that got ignored or felt awkward?" Volunteer shares it anonymously. Paste into Claude: "Fix this email. Make it professional and compelling. Improve the subject line. Keep under 150 words. Explain what you changed and why." Show before/after side by side. Students laugh, relate, and immediately want to fix their own sent emails.
💡 Script: "A 7-day follow-up email doubles callback rates. Almost nobody does it. With AI, it takes 60 seconds. This one habit alone could change your placement results."
- Connect to Day 13: "Emails = short form. Today = long form professional writing."
- Document types: business report, academic essay, project proposal, case study
- 4-Step AI Document Workflow: Outline → Draft → Edit → Polish
- Exact prompts for each step — live demo with timing
- How to keep your own voice in AI-written content (important!)
- 10 editing prompts that improve any piece of writing instantly
- Exercise: "15-Minute Report" speed challenge
- Topic: "How AI is changing jobs in India" (everyone same topic)
- Min 1-3: Generate outline using Step 1 prompt
- Min 4-9: Draft each section using Step 2 prompt
- Min 10-12: Edit using Step 3 — "improve this paragraph, add data"
- Min 13-15: Polish using Step 4 — "check grammar, flow, consistency"
- Read your opening paragraph aloud — class votes on best opening
Put a visible countdown timer on the projector screen. Everyone starts at the same time on the same topic. Play lo-fi music quietly in the background. At 15 minutes, call time. Ask 3 volunteers to read their opening paragraph. The quality difference between students who followed the 4-step workflow vs those who didn't is immediately visible and educational.
💡 Script: "3 years ago, a report like this took a junior employee 4 hours. You just did it in 15 minutes — and if you followed the workflow, it's actually good. This is your unfair advantage walking into any office."
- Why Excel + AI is the #1 skill combo for office jobs in 2024
- ChatGPT as Excel tutor: generate formula, explain it, debug errors
- 5 must-know formulas — each explained by AI in plain English: VLOOKUP, SUMIF, COUNTIF, IF, INDEX-MATCH
- Microsoft Copilot for Excel (M365) — if available: natural language to formula
- Analyze a shared sales dataset: find insights, create chart, explain findings
- Exercise: "New Employee Day 1" — analyze a shared sales dataset
- Provided dataset: 20 rows, columns: Month / Product / Units / Revenue / Region
- Task 1: Ask ChatGPT to write 5 formulas for this data (total, top product, growth%...)
- Task 2: "What 3 insights can you find in this data?" — paste the data, ask AI
- Task 3: "What chart type best represents this data and why?" — use AI recommendation
- Task 4: Ask AI to explain one formula you don't understand in simple terms
Share a messy, unfamiliar sales Excel file with students. Most freeze. Then show this prompt: "I have Excel data with: Month, Product, Units Sold, Revenue, Region. Write me 5 Excel formulas to: 1) Total revenue, 2) Top product by units, 3) Month-over-month growth %, 4) Count transactions above ₹50,000, 5) Average revenue by region. Include exact cell references assuming data starts at A2." From confused to capable in 90 seconds.
💡 Script: "You don't need to memorize Excel formulas. You need to know what question you want answered. Every finance, operations, and HR person who does this manually is now slower than you."
- Why presentations are career-defining (interviews, pitches, appraisals)
- Presentation structure: HPSCA — Hook, Problem, Solution, Call to action, Appendix
- AI for slide outline: exact prompt to generate 10-slide deck structure
- AI for speaker notes: what to say for each slide, 2-minute speaking version
- Gamma.app: describe your topic → AI generates entire deck with design in 90 seconds
- Export Gamma.app deck to PowerPoint format
- Exercise: Build an 8-slide deck and present 2 slides to a partner
- Step 1: Use Claude to generate HPSCA slide outline (topic: "Why AI is a career skill")
- Step 2: Generate speaker notes for each slide ("2 minutes per slide, conversational")
- Step 3: Paste topic into Gamma.app → watch auto-design happen → export to PPT
- Step 4: Present slides 1 and 2 to a partner using your AI speaker notes
- Partner feedback: was the opening strong? Did you sound natural?
Open gamma.app on the projector. In the prompt box, type: "Create a presentation: Why every company in India should adopt AI in 2025. 10 slides. Professional dark theme." Hit Generate. Start a 90-second countdown timer. Watch a fully designed, beautifully laid out presentation appear as time counts down. Every student's jaw drops. Then show how to export to PowerPoint in 2 clicks.
💡 Script: "Someone paid a designer ₹5,000 and 4 hours for a deck like this last week. You just made a comparable one in 90 seconds — and you can export it to PowerPoint and edit it. The skill is knowing what story to tell, not how to move boxes around."
- Canva AI features: Magic Write, Magic Design, AI Image Generator, Background Remover, Magic Resize
- 3 design principles for non-designers: Hierarchy, Contrast, White Space
- Social media size guide: Instagram (1080×1080), LinkedIn (1200×627), YouTube thumbnail (1280×720)
- ChatGPT + Canva workflow: write copy → design visual → resize for all platforms
- Design for purpose: event poster vs social post vs professional banner are different
- Exercise: Build a 3-piece "social kit" for an imaginary AI workshop
- Step 1: ChatGPT writes: event name, tagline, 2-line description, CTA
- Step 2: Canva Magic Design — paste topic → pick design from suggestions
- Step 3: Create event poster (A4 size)
- Step 4: Create Instagram post (1080×1080) for same event
- Step 5: Magic Resize → auto-convert to all social formats in 2 clicks
Design one Instagram post for "Gap2Growth AI Workshop". Then use Canva's Magic Resize to instantly convert it to LinkedIn post, YouTube thumbnail, and Instagram Story — all properly formatted, all in under 3 minutes. Previously this was an hour of a graphic designer's work. Students immediately see how this translates to freelancing, college fests, or business social media.
💡 Script: "Social media managers charge ₹15,000/month to do what you just did in 3 minutes. If you ever run a startup, a club, or a freelance gig — you are now your own design team."
- Task 1 (Email): Write a client-facing email announcing a new product feature. Formal, under 150 words, with a clear CTA.
- Task 2 (Report): Write a 300-word internal report summarizing this week's sales performance (use provided fake data). Use the 4-step workflow.
- Task 3 (Excel): Given a small dataset (10 rows), use AI to write 3 formulas and identify 2 insights. State the formulas + insights in a note.
- Task 4 (Deck): Use Gamma.app to create a 5-slide pitch on their fictional company. Export to PPT.
- Task 5 (Design): Create 1 LinkedIn post announcing their company's new product using Canva AI. ChatGPT writes copy, Canva designs.
- Teams present their 5 deliverables in 5 minutes — class scores on professionalism and speed
Students now create professional-quality content with AI. The question: should they disclose this? Discuss: when is AI use transparent vs deceptive? In job applications, college assignments, client work, social media? Introduce real examples of AI disclosure policies at companies. Ask students: "If your email was 80% AI and 20% you — is that okay?" This thoughtful discussion bridges to Week 4 (creative AI) where attribution becomes even more complex with music, images, and video generation.
- Bridge from Saturday ethics: "We discussed AI images ethically. Now let's create them."
- Free tools: DALL-E (ChatGPT), Canva AI Images, Bing Creator (unlimited), Adobe Firefly
- The Image Prompt Formula: Subject + Style + Lighting + Mood + Angle + Detail level
- 5 key style words every student should know: Photorealistic, Cinematic, Illustration, Watercolor, Minimalist
- Real use cases: marketing images, product mockups, social media, presentations
- Exercise: 3-round "Image Prompt Battle" — class votes each round
- Round 1: Simple prompt (just a noun — "office", "mountain", "phone") — compare results
- Round 2: Add style ("cinematic photorealistic", "watercolor illustration") — see jump in quality
- Round 3: Full formula prompt — Subject + Style + Lighting + Mood + Angle + Detail
- Also: generate your own LinkedIn banner background image using the full formula
Show a slideshow of 10 images — 5 real photographs, 5 AI-generated. Students vote: which are AI? Mix: a photorealistic portrait, a hyperdetailed product shot, a cinematic city skyline, an illustration-style scene, a corporate headshot. They'll get many wrong. Then reveal the answers. Then say: "By the end of this class, you'll be the one creating images like these — in 10 seconds, for free."
💡 Script: "Stock photo sites charge ₹500 per image. You'll generate 20 per day for free. If you're a freelancer, a small business, or a content creator — this is your design team."
- Connect to Day 19: "Yesterday = still images. Today = they move."
- Types of AI video: text-to-video, image-to-video, AI video editing
- Runway Gen-3: create account (free credits), video prompt formula
- CapCut AI: auto captions, AI script-to-video, background removal, AI voice
- Full 30-second workflow: ChatGPT script → Runway clips → CapCut edit → reel
- Exercise: Create a 20-30 second Instagram reel using full AI workflow
- Step 1: ChatGPT — write a 30-second script for "Why students should learn AI"
- Step 2: Runway — generate 3 short video clips (5-8 sec each) from text prompts
- Step 3: CapCut — import clips, arrange, add auto captions, choose background music
- Step 4: Export and share reel with class WhatsApp group
- Class votes: best script, best visual, best caption style
Prepare before class: use ChatGPT script → Runway video clips → CapCut edit to create a 20-second reel about Gap2Growth. Walk in, connect phone to projector, play the reel without context. Students assume it was made by a video editor. Then reveal the full 4-step workflow and that it took 25 minutes. The reaction when they realize they'll learn to do this in 2 hours is priceless.
💡 Script: "Video agencies charge ₹10,000 for a 30-second promotional reel. YouTube automation channels make ₹50,000/month creating videos exactly like this. You now know the complete workflow."
- Connect to Day 20: "Images → Video → Now: Music. You now have a full media studio."
- Suno AI: describe a song → full song with vocals in 30 seconds
- Simple mode vs Custom mode (your own lyrics)
- Style tags: how to control genre, instruments, mood, language
- ChatGPT + Suno workflow: ChatGPT writes lyrics → Suno composes
- Genres to practice: Bollywood pop, lo-fi, hip-hop, motivational, acoustic
- Exercise: Create 2 songs — one simple mode, one custom with your lyrics
- Song 1 (Simple): Describe your college in Suno simple mode — generate 2 versions
- Song 2 (Custom): Use ChatGPT to write lyrics for "an AI course that changed my life" → paste into Suno custom mode → style: "Bollywood pop, upbeat, Hindi-English mix"
- Share your best song in class WhatsApp group
- Optional: add Day 20's reel + today's music = complete video with custom soundtrack
The night before, create the Gap2Growth course theme song: ChatGPT writes Hindi-English motivational lyrics → Suno custom mode with style: "Bollywood pop, upbeat, motivational, male and female vocals, modern Indian sound". Walk into class Wednesday playing it through Bluetooth speakers. Don't explain it. Let students figure out it's AI. Reveal after 1 minute. The reaction sets the perfect tone for the entire class.
💡 Script: "Jingle companies charge ₹40,000 for a 30-second brand song. You now know how to make a full song in 10 minutes. YouTube channel? Podcast? Small business? You have a music studio in your browser."
- What is a content campaign — how brands plan 7-30 days of content
- YouTube content workflow: title → thumbnail concept → full script → description → tags
- Blog article: 600-word SEO article with proper structure in 10 minutes
- 7-day social media calendar: theme, post idea, caption, hashtags, best time
- Platform differences: Instagram vs LinkedIn vs YouTube — different strategy
- Exercise: Build a mini 3-platform content campaign on one topic
- Topic: "AI as a career skill for students" — all students same topic
- Deliverable 1: YouTube video plan (title + thumbnail + 5-min script outline)
- Deliverable 2: 600-word blog article draft using 4-step doc workflow (Week 3!)
- Deliverable 3: 3 Instagram captions with hashtags — different angles on same topic
- Deliverable 4: 1 LinkedIn post — professional angle, 150 words
Point out explicitly to students: today's exercise uses the 4-step document workflow from Week 3 (for the blog), image generation from Day 19 (for the thumbnail), and Week 2's research skills (for the YouTube content). This is the first time all the weeks connect visibly. Pause and let students notice it themselves: "What Week 1, 2, 3, and 4 skills did you use today?"
💡 Script: "A single person with these skills can now do the job of a content writer, graphic designer, video producer, and social media manager. That's 4 salaries worth of output. This is why AI is a leverage skill, not just a tool."
- Resume optimization: ATS keywords, impact statements, AI audit
- Job description matching: how to customize resume for each role
- AI mock interview: 1 question at a time + feedback on each answer
- LinkedIn "About" section: 150-word, conversational, ends with CTA
- Portfolio project ideas from AI: based on your specific skills + interests
- What to post on LinkedIn after this course (do it before leaving Saturday)
- Exercise: "Career AI Audit" — improve your real career materials
- Part 1: Paste your resume into Claude → get full audit + 3 rewrites of weak bullet points
- Part 2: 5-question AI mock interview — answer each question, get feedback
- Part 3: Write your LinkedIn About section using AI prompt → paste into LinkedIn
- Part 4: Ask AI: "Suggest 5 portfolio projects I can build in 3 days given my skills: [list]"
Ask for a volunteer. Use Claude with this prompt: "You are a senior HR interviewer at a top Indian IT company. Interview [student name] for a fresher software/marketing role. Ask ONE question at a time. After each answer, give: 3 things done well, 1 specific thing to improve, and a model answer. Start with: Tell me about yourself." The class watches live. AI's feedback is precise, kind, and actionable — often better than mock interview coaching students have paid for.
💡 Script: "Most students walk into interviews with zero practice. AI never gets bored, never gets impatient, and gives you detailed feedback at midnight before your interview. Use it until you're genuinely confident. Then walk in."
- AI Image: Create 3 brand images using DALL-E or Bing Creator — product shots, lifestyle, promotional visual
- AI Music: Use Suno to create a 30-second brand jingle (ChatGPT writes lyrics first)
- AI Video clip: Use Runway to generate 1 short brand clip (5-8 seconds)
- Content Calendar: 3-day Instagram plan with captions, hashtags, post type for each day
- LinkedIn Post: Professional announcement for this brand launch — 150 words, formal tone
- Agency pitch: present the full package in 3 minutes — show images, play jingle, show calendar. Class votes on most compelling brand package.
Students have learned to use AI tools manually. But the next wave is AI agents — AI that doesn't just answer questions, but takes actions: browses the web, sends emails, manages files, books meetings. Show 2-minute demo of any AI agent tool (ChatGPT's web browsing, Claude's computer use, or Perplexity's Copilot). Discuss: "If AI can do all of these tasks automatically — what do YOU need to be?" The answer: the person who sets the goal, evaluates the output, and makes the final call. This bridges directly to Week 5 where students become directors of AI, not just users of it — building their final real-world project.
- A — Startup Research: Business plan + market analysis + investor pitch deck. Tools: Perplexity, Claude, Gamma.app
- B — Marketing Campaign: 7-day calendar + 5 designed posts + promo video + ad copy. Tools: Canva, ChatGPT, Runway, Suno
- C — AI Content Channel: YouTube plan + 3 video scripts + 3 thumbnails + channel branding. Tools: ChatGPT, DALL-E, CapCut
- D — Music Album: 4 songs (different genres) + lyrics + album cover + playlist description. Tools: ChatGPT, Suno, DALL-E
- E — Career Portfolio: ATS resume + LinkedIn profile + cover letter + 3 portfolio project descriptions. Tools: Claude, ChatGPT, Canva
- Teacher presents all 5 projects with example outputs (30 min)
- Students fill "Project Brief" form: project type, topic, audience, tools, expected outcome
- Research phase begins: use Perplexity + Gemini to gather project-specific research (from Week 2!)
- End of day: share project brief with teacher for quick review
- Homework: complete research tonight, come ready to build tomorrow
Open Monday with a 5-minute "before and after" moment. Show a list on the projector of everything students could NOT do 4 weeks ago. Then show the list of what they can do now. Read both lists slowly. Let the silence after the second list land. Then say: "This week is not homework. This week is your portfolio. Build something you're proud of."
💡 Script: "Your project this week is the beginning of your portfolio. Every skill from the last 4 weeks was practice for this. Make something real. Make something you'd show at an interview."
- 5 min standup: share yesterday's research finding — 1 thing that surprised you
- 70 min: core content build (project-specific prompts below)
- 15 min: peer review — swap with neighbor, get 3 suggestions
- 20 min: implement peer feedback using AI
- End: every student shares a 1-sentence progress update with teacher
- A (Startup): Executive summary + market size + competitive analysis + business model
- B (Campaign): Campaign strategy + 7-day calendar + 3 post captions + ad headline + CTA
- C (YouTube): Channel positioning + 3 video titles + full script for Video 1 + channel description
- D (Album): Album concept + 4 song concepts + complete lyrics for Song 1 + artist bio
- E (Career): Resume audit + 3 rewrites + cover letter + LinkedIn About + portfolio brief
Share these 4 prompts with every student during Build Day 1 — they work for any project type and improve output quality significantly at any stage:
1. Quality: "Review what I've built so far. Give 5 specific improvements to make it more professional."
2. Clarity: "Is this easy to understand for someone who knows nothing about this? What's confusing?"
3. Impact: "What's missing that would make this significantly more impressive?"
4. Audience: "Read this from the perspective of [target audience]. What's their reaction? What questions do they have?"
- 5 min standup: what did peer review reveal? What are you improving?
- Generate all required images using DALL-E / Bing Creator / Canva AI
- Design all visual deliverables in Canva (posts, covers, banners, slides)
- Use Magic Resize for multi-format outputs
- Teacher checkpoint: review every student's visual output — quality pass/fail
- 5 min standup: visual layer done — what's the gap between good and great?
- Add video clips (Runway) or music (Suno) if project requires
- Complete all remaining written deliverables
- Use universal refinement prompts on EVERYTHING one final time
- End of day: full project should be 90% complete
Teach students to run the "Stranger Test" on every deliverable they create: show it to someone who has NEVER seen your project and ask them: "In 30 seconds, what is this? Who is it for? What is it asking me to do?" If they can't answer all 3 — it needs more work. This is the same quality standard professional agencies use. AI can help with this too: "Read this as if you've never heard of my project. Answer: what is it, who is it for, what should I do next?"
💡 Script: "Good work is clear to a stranger. Great work is obvious to a stranger. That's the difference between a 7/10 and a 10/10 project."
- The WHAT-HOW-WOW Presentation Structure:
— WHAT: what did you build? (30 sec)
— HOW: which AI tools + prompts you used (2 min)
— WOW: best result, what surprised you, what's next (1 min)
— Q&A: 1-2 questions from audience (1.5 min) - Common presentation mistakes to avoid: reading from notes, showing too many slides, rushing
- Using AI for Q&A prep: "Ask me 5 hard questions about my project"
- 30 min: teacher explains WHAT-HOW-WOW + common mistakes + tips
- 20 min: students write their own WHAT-HOW-WOW script (AI can help draft it)
- 40 min: partner rehearsal — 5-minute timed presentation each way
- Partner feedback form: 3 things done well + 1 thing to improve
- 10 min: final project check + file organization + tech check for Saturday
- Homework: practice presenting to a family member or mirror tonight
Teach students this Q&A prep prompt: "My project is [brief description]. I'll present it to a class of students and a teacher. Ask me 5 difficult questions that a skeptical audience might ask. Then tell me how to answer each one well." This is more useful than any mock presentation coaching, and they can do it alone at home the night before Demo Day.
💡 Script: "The best presenters don't wing Q&A — they've already answered every hard question in their head. AI can throw every tough question at you tonight so nothing surprises you tomorrow."
- 3-minute project preview — show your work so far, explain where you are and what's left
- Peer reviewer 1: fills feedback form — 2 strengths, 1 specific improvement suggestion
- Peer reviewer 2: answers — "Is the project's purpose immediately clear? What's confusing?"
- Teacher: gives 1-minute targeted feedback focusing on one high-impact improvement
- Student writes: "My top 2 improvements for next week based on today's feedback"
- Final 20 min: everyone uses AI to implement ONE piece of feedback immediately in class
Do a quick show-of-hands exercise: "How many of your friends know what Perplexity is? How many have used NotebookLM? How many have generated a song with Suno?" Almost nobody in the room will raise their hands for each. Then say: "Every tool you've learned this month — your peers don't know about it. Your future employers don't know who from their applicants knows these tools. But now they will — because your project will show them." This restores motivation in the final stretch and connects directly to Demo Day's LinkedIn post exercise.
- 5-minute presentation per student: WHAT (30s) + HOW (2min) + WOW (1min) + Q&A (1.5min)
- Class Q&A: 1-2 genuine questions from audience after each presenter
- Teacher feedback: 2 sentences — what was most impressive, what is the student's strength
- After all presentations: class votes on 4 awards (Most Creative, Most Practical, Most Likely to Go Viral, Best Career Move)
- LinkedIn Post: each student writes and posts their course completion post using AI — before they leave the room
- Certificate ceremony + teacher closing speech + group photo
"35 days ago, you came here curious. You leave today capable. The difference between those two words — that's this course. AI is not slowing down. The tools will change. The principles you learned here — how to communicate, how to research, how to create, how to build — those don't change. You are now the people who know how to work with the future instead of being afraid of it. Use that. Share it. Teach someone else. The world is changing — and you changed first."