Term: Generative AI
~6 min read
Estimated time: ~6 min read — for the in-app brief plus opening the primary source.
What this is
Generative AI creates new content — text, images, code, sometimes audio — by predicting what should come next. That is why it can draft well and also invent a citation.
Everyday example
ChatGPT drafts a customer letter; Claude proposes contract language; Grok outlines a board note; Gemini comments on a storyboard. Each is generating new text — not looking up one stored answer.
Generative AI creates new text, images, or code — it does not look up a single stored answer.
- ChatGPT, Claude, Gemini, and Grok are generative: they compose the next likely tokens.
- That is why they can draft a memo and also invent a citation.
- Analytics AI (fraud score, demand forecast) is usually not generative.
- Most 2023–2026 board conversations about “AI” actually mean generative AI.
Next action: When someone says “we use AI,” ask: does it generate content, or does it score/predict?
What changes in how you lead
How decision rights, process, and ownership should change.
- Treat generated output as a draft until a person verifies facts that matter.
- Do not apply chatbot rules to a fraud model, or fraud-model rules to a chatbot.
Compare related ideas
Generative AI vs AI (broad)
AI is the whole category. Generative AI is the subset that creates content. A fraud-score model is AI without being generative.
Open AI (broad)Deep dive
Analytics AI scores or forecasts (fraud, demand). Generative AI composes.
Products you already know — ChatGPT, Claude, Gemini, Grok — are generative.
Treat output as a draft until a person verifies facts that will be decided on.
Marketing, Legal, and comms feel this first because language is the product.