METANOWCreating from learned patterns
Generative AI produces new content from patterns learned during training. Depending on the model, outputs can include text, code, images, audio or video. A prompt supplies instructions and context; the result is generated rather than simply copied from a database.
The main components
A practical application combines a model, an input interface, instructions, context and an output format. Retrieval, permissions, safety checks and evaluation can surround the model. These application controls are separate from the model’s learned parameters.
Where it helps
Typical tasks include drafting correspondence, summarising documents and explaining technical information. The result can be plausible while still containing an error, missing context or a biased assumption. Fluency is not proof of correctness.
A useful first exercise
Choose an approved document and ask for a short summary with references to its sections. Compare the answer with the original. Record omissions and unsupported claims before deciding whether this task is suitable for broader use.
Sources & further reading
Reviewed 19 October 2026. This is a technology learning session; examples describe possible designs, not a claim that a particular integration is active.