📊 View Lecture Slides Full-screen presentation with navigation

Session 1: The AI-Native Workspace

Session Duration: 2 Hours     Block: 1 — Foundations & The Exocortex

Learning Objectives

By the end of this session, students will be able to:

  • Describe how cloud AI services eliminate local hardware barriers to software development
  • Set up a functional AI-assisted development workspace using GitHub Copilot and Google AI Studio
  • Articulate the difference between using AI as a search engine versus using it as a collaborative engineering partner
  • Define a concrete AI product idea they will develop throughout this course

Hour 1: From Local Compute to Cloud Intelligence (Instructor-Led — 60 minutes)

1.1 The Old World vs. The New World

For decades, building intelligent software required access to expensive hardware, specialised teams, and years of mathematical training. A university computer vision project might have needed weeks of GPU time and a team of three PhD students. A recommendation engine required data scientists, ML engineers, and production infrastructure.

That world no longer exists for most practitioners.

The modern reality: Foundation models trained on vast datasets are now accessible via a simple API call. You send text (or an image, audio file, or document). You receive a sophisticated response. The model runs on infrastructure you never touch.

This changes the question from “do I have the skills and hardware to build AI?” to “do I understand the problem well enough to direct an AI system towards solving it?”

1.2 What Cloud AI Actually Means

Concept What It Means Practical Implication
Foundation Model A large model pre-trained on massive data You start from enormous general capability
API Application Programming Interface You communicate with the model using structured requests
Token The unit of text the model processes Your costs and context limits are measured in tokens
Inference Running the model to get a response You pay per use, not for the hardware
Context Window How much the model “sees” at once The length of conversation or document it can process

1.3 AI as a Collaborative Engineering Partner

There are three levels at which most people interact with AI:

Level 1 — Consumer: “Give me a recipe for pasta.” The AI answers. The conversation ends.

Level 2 — Prompt Writer: “Act as a nutritionist. Suggest a high-protein pasta recipe for someone training for a marathon.” Slightly better output, but still a one-shot request.

Level 3 — Engineering Partner: A continuous, iterative conversation where you describe a system, ask the AI to generate a component, review the output, identify the flaw, ask for a revision, test the result, and repeat. This is the level this course operates at.

The critical shift is understanding that AI does not replace the engineer’s judgment — it accelerates the engineer’s execution. You still need to:

  • Define the problem precisely
  • Evaluate whether the output is correct
  • Identify failure modes and edge cases
  • Integrate components into a coherent system

1.4 The Course Framework: THINK → KNOW → BUILD → CONNECT → EQUIP → EVALUATE → SHIP

This module is organised around seven engineering stages that mirror how a professional builds an AI-powered application:

Stage What You Do
THINK Use AI to explore problems and reason about solutions
KNOW Build a personal knowledge base (exocortex) for AI retrieval
BUILD Use AI as a coding partner to scaffold applications
CONNECT Wire applications to cloud AI APIs
EQUIP Give AI tools and agentic capabilities
EVALUATE Test for reliability, hallucination, and security risks
SHIP Deliver a functional, demonstrable prototype

Every session maps to one or more of these stages. By Session 15, you will have passed through the full cycle.

Hour 2: Practical — Environment Setup & Product Idea Definition (60 minutes)

Lab 1.1 — Set Up Your AI-Native Development Environment

Step 1: GitHub Account and Copilot

  1. Create or log in to your GitHub account at github.com
  2. Activate GitHub Copilot (free for students via GitHub Education Pack, or use the free trial)
  3. Install Visual Studio Code if not already installed
  4. Add the GitHub Copilot extension from the VS Code Marketplace
  5. Verify Copilot is active: open a new .py file and start typing def get_ — Copilot should offer a suggestion

Step 2: Google AI Studio

  1. Navigate to aistudio.google.com
  2. Sign in with your Google account
  3. Create a new prompt: select “Create new prompt” → “Freeform prompt”
  4. Get your API key: click “Get API key” in the sidebar — you will need this in Session 7

Step 3: Verify Your Setup

You should now have:

  • ✅ VS Code with GitHub Copilot active
  • ✅ An AI Studio account with a generated API key (store this safely)
  • ✅ A GitHub account for storing your project code

Lab 2.2 — Define Your AI Product Idea

You will build one project throughout this module. Choose wisely: the idea should be specific enough to be buildable in 30 hours, but meaningful enough to be worth building.

Product Idea Framework

Answer these four questions:

  1. Problem: What specific, concrete problem does this application solve?
  2. User: Who uses it? (Be specific — “students at my university” is better than “people”)
  3. AI Role: What does the AI specifically do in this product? (e.g., “summarises documents”, “answers questions about a knowledge base”, “classifies support tickets”)
  4. Success Metric: How would you know the product is working correctly?

Example Responses:

Question Weak Answer Strong Answer
Problem “Helps people learn” “Students waste time re-reading lecture notes; this tool answers questions from those notes”
User “Anyone” “First-year BSc Computer Science students at my university”
AI Role “Uses AI” “Embeds lecture notes and answers natural language questions using retrieved context”
Success “Works well” “Correctly answers 8 out of 10 test questions from the notes without hallucinating”

Recommended Project Areas:

  • Course Knowledge Assistant — upload lecture notes, ask questions
  • Internal Document Navigator — search through a collection of reports or policies
  • CV-to-Job Matcher — compare CVs against job descriptions and explain gaps
  • Multi-Document Summariser — paste multiple articles, receive a structured synthesis
  • Personal Research Assistant — chat with a curated collection of saved web articles

Record your four answers in a document. You will refine this idea in every session.

Key Takeaways

  • Cloud AI services provide foundation model capability via simple API calls — no local hardware required
  • AI is most powerful when used as an iterative engineering partner, not a one-shot question-answering tool
  • This course follows a seven-stage engineering framework: THINK → KNOW → BUILD → CONNECT → EQUIP → EVALUATE → SHIP
  • Your AI product idea should have a specific problem, a defined user, a concrete AI role, and a measurable success criterion

Further Reading

  • Google AI Studio documentation: ai.google.dev
  • GitHub Copilot for students: education.github.com
  • “The Pragmatic Programmer” (Hunt & Thomas) — foundational software engineering mindset