TL;DR - "AI" is an umbrella word. Today's headline tools (ChatGPT, Claude, Gemini) are generative AI built on machine learning. Knowing the difference tells you when AI will help and when it will flop.
Why it matters
People call everything "AI" - spam filters, Netflix picks, ChatGPT. If you can't tell them apart, you can't predict what a tool is good at, so you'll either over-trust it or dismiss it. This mental model is the foundation everything else builds on.
The family tree
- Artificial intelligence (AI) - the broad goal: machines doing things that seem to need intelligence.
- Machine learning (ML) - systems that learn patterns from data instead of being hand-coded (your spam filter).
- Deep learning - ML using large neural networks; powers image and language understanding.
- Generative AI - deep learning that creates new text, images, audio, or code (ChatGPT, Claude, Midjourney).
Generative AI is a small branch of a big tree - but it's the branch that changed everyday work.
Worked example - same word, different tech
"AI flagged this as spam." -> a classic ML classifier. It sorts; it doesn't write.
"AI wrote me a draft reply." -> generative AI. It produces new text.
"AI recommended this show." -> a recommendation system. Different again.
Same word, three different tools, three different failure modes.
Steal this - the 2-question test
When someone says "it's AI", ask:
1. Is it learning from data, or following fixed rules?
2. Is it generating something new, or sorting/scoring things that exist?Those answers tell you what to expect and where it might break.
Common mistakes
- Thinking all AI is ChatGPT. Most "AI" in products is narrow ML, not a chatbot.
- Assuming "AI" means accurate. It usually means "a pattern-matcher trained on data" - which can be confidently wrong.
- Believing the label. "AI-powered" is sometimes a rules engine with a logo.
Good to know - the tools you'll meet
The big generative assistants are ChatGPT (OpenAI), Claude (Anthropic), and Gemini (Google). They're all LLMs and behave similarly, so the skills here transfer across all of them - no need to pick a favorite yet.