Decoder. plain-English AI glossary

Assistant persona

● Core

Defining how an AI assistant should sound and behave — its tone, values, and communication style.

Think of it like

Casting an actor for a role: defining how they should deliver lines.

Example

Persona: "You are helpful, witty, and concise. You avoid jargon and explain clearly. You are honest about limitations."

How it actually works

Personas are encoded in system prompts and training (RLHF, instruction tuning). A well-defined persona improves consistency and user trust. Personas can be specific (a pirate, a therapist, a data scientist) or abstract (helpful, harmless, honest). Over-specific personas can cause hallucination.

For product teams

Shapes user perception and engagement. Consistency builds trust.

For engineers

Define clearly in system prompt. Test across diverse inputs. Monitor for off-character outputs.

Role prompting — Role prompting is a specific technique.

Read anything AI without the jargon

Look up any term in plain English, or save terms as you read with the free Chrome extension.

Open Decoder Add to Chrome