Understand
How AI works, in plain English.
Clear explainers, a jargon-free glossary, who’s who and the timeline, so you can judge the claims yourself. No maths, no hype.
AI in 10 minutes
Five ideas that explain almost every AI headline. Read it once and the news starts making sense.
- 01Models what they are and what they do
- 02Training how they learned, and what they can’t know
- 03Prompts why how you ask changes what you get
- 04Context what the AI can “see” right now
- 05Agents when AI takes actions, not just answers
Six clusters, one question each
Foundations
What AI, machine learning and LLMs actually are.
How models work
Tokens, training, context windows and why answers vary.
Agents & automation
When AI takes actions, not just answers questions.
Data, privacy & security
What happens to what you paste in.
Judgment & risk
Where AI should draft and where a person decides.
The business of AI
Costs, vendors and what to buy.
Who’s who, and how we got here
Model makers
Companies that build the underlying models.
- OpenAI
- Anthropic
- Meta
- Mistral
- DeepSeek
Assistants you’ll use
The chat apps most people meet first.
- ChatGPT
- Claude
- Gemini
- Copilot
- Meta AI
Where it shows up at work
AI features inside software you may already pay for.
- Microsoft 365
- Google Workspace
- Notion
- Canva
- Shopify
- HubSpot
Names only, no logos or rankings: a neutral map, not an endorsement.
The AI timeline
- 1950Turing asks “Can machines think?”Alan Turing proposes a test: can a machine hold a conversation you can’t tell from a person’s?
- 1955The term “AI” is coinedA proposal for a summer workshop at Dartmouth College names the field “artificial intelligence.”
- 1997Deep Blue beats KasparovIBM’s chess computer beats the world champion, using brute-force search rather than learning.
- 2012Deep learning takes offA neural network wins a major image-recognition contest by a wide margin, and much of the field shifts to neural networks.
- 2017The Transformer paperGoogle researchers publish the design that most of today’s language models are built on.
- 2022ChatGPT launchesA chat interface puts a large language model in front of the public, and usage explodes.
- 2024AI in everyday softwareAssistants appear inside email, documents, spreadsheets and phones most people already use.
- 2026Agents at workMore assistants take actions, such as sorting mail and booking calls, with a person approving the important steps.
Three myths, gently corrected
- Myth
“AI understands what it says.”
RealityIt predicts likely words. It can sound certain and still be wrong, so check facts, names and numbers.
- Myth
“AI will replace your team.”
RealityIt takes over well-defined steps inside work that people still own and are accountable for.
- Myth
“You need to be technical.”
RealityIf you can write a clear email to a new hire, you can write a good prompt.