Generative AI, explained
The request tells the system what kind of result to produce. The result is new, but it is shaped by patterns the model learned from its training data.
Generative AI is artificial intelligence that creates new content, such as text, images, audio, video, or software code, in response to a person's request.
It creates rather than only classifies
Some AI systems sort records, detect patterns, or predict an outcome. Generative AI produces material you can read, view, hear, or use.
A business team might ask for a draft email, a summary of notes, or a first version of software code. The model uses the request and its learned patterns to form the response.
The request and review both matter
A clear request gives the model a job, useful context, and a desired format. That helps the output fit the intended use, but it does not guarantee that every detail is correct.
Treat the first response as material to inspect. Check facts against the source, test code before relying on it, and keep a person responsible for the final decision.
Generative AI terms
- Prompt
- The request or instructions given to a generative AI system.
- Foundation model
- A model trained broadly enough to support many different applications.
- Output
- The text, image, audio, video, or code the system creates in response.
What makes a request useful
State the job
Tell the system what you need the result to help someone do.
Add the context
Include the facts, audience, or source material that should shape the result.
Check the result
Review important facts and rules before using the generated material.
Questions about generative AI
It can create text, images, audio, video, and software code. The available formats depend on the model and application.
No. A large language model is a type of foundation model focused on language. Generative AI also includes models that create images, audio, and video.
It generates output from patterns learned during training. The result still needs review for accuracy, suitability, and proper use of source material.
Related glossary terms
Turn an idea into a useful tool
Caffeine apps run on the Internet Computer, a public network built for this.