GLOSSARY

What is Generative AI?

Table of content
Definition

Definition

Generative AI is artificial intelligence that produces new content such as text, images, audio, video, code, or interface structures in response to input and context.

TL;DR

  • Generative AI creates new outputs from learned patterns and current input.
  • Outputs can include text, images, code, audio, video, and product artifacts.
  • The model predicts plausible outputs; it does not automatically know your product truth.
  • Quality depends heavily on context, constraints, tools, and evaluation.
  • Product teams should treat generated output as a candidate to review, not ground truth.

What does generative AI do?

Traditional software often follows explicitly programmed rules. Generative AI can produce a new response from a prompt and context: draft a PRD, generate an interface, write code, summarize research, or create an image.

The system produces output by modeling patterns learned during training and information available at runtime.

What can generative AI generate?

  • text and structured documents
  • images and visual concepts
  • software code
  • audio and speech
  • video
  • UI and product artifacts
  • synthetic data

Generative AI vs. AI agent

Generative AI describes the ability to create outputs. An AI agent can additionally plan, use tools, inspect results, and take multiple steps toward a goal.

Why does context matter?

A generic model can generate a plausible settings page, but without product context it may ignore existing navigation, terminology, permissions, components, or business rules.

That gap is why professional workflows increasingly combine generation with retrieval, product memory, design systems, and validation.

Common mistakes

Confusing plausible with correct

Generated output can look confident while containing factual or product errors.

Expecting one prompt to contain the whole product

Large tasks benefit from structured context and iterative evaluation.

Automating judgment along with production

Human review remains important for consequential product decisions.

The bottom line

Generative AI asks: given this input and context, what useful new artifact can the model produce?

Related terms

AI agent · AI product design · AI prototyping · AI prompt

Relevant Figr resource

Read AI for Product Design for how generative systems are used in real product workflows.

Related Figr Projects

No items found.