Lesson 01 · 8 minutes
What AI is, and what it is not
Separate models, products, search, automation and ordinary software so you can choose the simplest suitable intervention.
Why this matters
People often buy an AI product when an existing rule or software feature would be more reliable.
Understanding the layers makes pricing, privacy and capability claims easier to evaluate.
The practical bit
A model generates or interprets content from patterns. A product wraps one or more models with features such as files, search, memory or collaboration. An API lets software call capabilities programmatically.
Search retrieves information. Deterministic automation follows defined rules. Ordinary software stores, calculates or moves information predictably. Modern products may combine all of these, so the label ‘AI’ does not tell you how a feature works.
What it looks like
If invoices already follow fixed fields, accounting software and validation rules may solve the problem more reliably than asking a general model to interpret every invoice from scratch.
A calendar reminder is ordinary automation. Asking a model to draft a thoughtful message is generative assistance. They solve different parts of organising a birthday.
Calling every useful software feature AI, then comparing products by model reputation instead of the complete workflow you need.
Use your judgement
Try it now
- Pick one digital task you do regularly.
- Classify each step as human judgement, software, search, automation or generative AI.
- Circle the step that genuinely needs interpretation or generation.
Use this in a real situation
Improve a familiar updateUse a practical task to separate pattern work from human judgement.Open the practical →Keep this
Choose the intervention, not the fashionable label.