What is AI?

Artificial Intelligence is when a computer mimics human thinking to solve problems.

Machine Learning

Instead of following strict rules (like Python code), the AI learns from Examples (Data).

Show 1000 dog photos
Finds Patterns
Recognises new dogs

Generative AI

AI that can create new content, not just analyse it.

LLMs (Text)

Large Language Models (e.g., ChatGPT) predict the next word in a sentence.

Image Gen

Tools like Midjourney create pixels based on text descriptions.

The Prompt

The instruction you give the AI. Better detail = Better result.

"Draw a cat" vs "Draw a neon cyber-cat in rain"

Training Data

AI is only as smart as the data it was fed.

Garbage In, Garbage Out

If an AI learns from bad (or biased) data, it will give bad answers.

Copyright Issues

AI learns by scraping the internet. It might use artists' work without permission to learn styles.

AI has no "Truth", only Patterns.

Hallucinations

When an AI confidently makes up a lie.

Why?

It doesn't "know" facts; it just predicts the next likely word. Sometimes it predicts wrong but sounds right.

ALWAYS Fact Check AI work!

Algorithmic Bias

When AI treats certain groups of people unfairly.

  • Example 1: An AI hiring tool rejecting female CVs because it was trained on mostly male CVs.
  • Example 2: Facial recognition struggling with darker skin tones due to lack of training data.

Bias is often hidden in the training data.

Deepfakes

AI-generated video or audio that looks/sounds real but is fake.

Dangers
  • Fake News
  • Scams (Voice cloning)
  • Bullying
Spotting It
  • Unnatural blinking
  • Blurry mouth/lips
  • Robotic tone