← All lessons
Week 1 · Foundations of Generative AI
MAM5020F — Generative AI for Research

Foundations of Generative AI

From Turing's Dream to Today's Revolution
Contents
← All lessons
Week 1 · Foundations of Generative AI
Key terms
Generative AI
AI systems that produce new content (text, images, code, audio) in response to a prompt, rather than only classifying or retrieving existing data. ChatGPT, Claude, and Midjourney are widely used examples.
AI winter
A period of collapsed funding and interest in AI research after expectations outran what the technology could deliver; notable winters occurred in the 1970s and the late 1980s.

This opening lesson traces the arc of artificial intelligence from its origins in the 1950s through to today's generative AI revolution. No technical background in machine learning is assumed. The goal is to build a shared conceptual vocabulary and historical understanding that will ground every subsequent week of this course.

Learning Objectives

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

1

Trace AI's Historical Development

Follow the evolution from Turing's foundational questions through expert systems, neural networks, deep learning, and ultimately to modern generative AI. Understand the key breakthroughs, setbacks (including the "AI winters"), and turning points that brought us to today.

2

Explain How Generative AI Works

Describe in plain language how systems like ChatGPT, DALL-E, Midjourney, and similar tools generate their outputs. Understand the concepts of training, patterns, and prediction that make these systems possible.

3

Distinguish Different Types of AI

Differentiate between rule-based systems, discriminative AI (classification/regression), and generative AI using everyday analogies. Understand when each approach is appropriate and what their respective strengths and limitations are.

4

Connect AI to Your Research

Identify which AI paradigms are most relevant to different tasks in your own research field. Recognize opportunities where generative AI might assist your work, while understanding its current limitations and appropriate use cases.

Key Topics We'll Cover

The Turing Test and early AI philosophy
Expert systems: promises and limitations
Neural networks: from perceptrons to deep learning
The ImageNet moment and computer vision
Transformers and attention mechanisms
Large language models (GPT, BERT, Claude)
Diffusion models and image generation
Current capabilities and fundamental limitations

A Preview: Eras of AI Development

1950s-1970s
Symbolic AI & Early Optimism: Rule-based systems, logical reasoning, and the first AI programs
1970s-1980s
First AI Winter: Unfulfilled promises lead to funding collapse and skepticism
1980s-1990s
Expert Systems Boom & Bust: Commercial success followed by another winter
1990s-2010s
Machine Learning Emerges: Statistical approaches, backpropagation, and neural network revival
2012-2017
Deep Learning Revolution: ImageNet breakthrough, AlphaGo, and the transformer architecture
2022-Present
Generative AI Era: ChatGPT, Claude, Deepseek, Nano Banana, Sora, and so much more
← PreviousAI Content Disclaimer

Drafted with Claude (Anthropic) and reviewed by Jonathan Shock before publication. AI-generated errors are possible — if you spot one, please email jonathan.shock@uct.ac.za. Full detail: AI Content Disclaimer.

© 2026 Jonathan Shock · MAM5020F: Generative AI for Research · CC BY 4.0

Week 1
Lesson 1
2 min read
Key terms
Generative AI
AI systems that produce new content (text, images, code, audio) in response to a prompt, rather than only classifying or retrieving existing data. ChatGPT, Claude, and Midjourney are widely used examples.
AI winter
A period of collapsed funding and interest in AI research after expectations outran what the technology could deliver; notable winters occurred in the 1970s and the late 1980s.