Prompt Engineering - Persona, Role and Tone

Three words are often mixed up, but they mean different things:

• Role = the job (what AI does).
• Persona = the behaviour (how AI behaves).
• Tone = the voice (how AI speaks).

Put together, they define the complete behaviour of AI.

The school example makes it clear: all three people are teachers (same role), but one is strict, one is friendly, one is funny (different personas). Same job, different behaviour — that behaviour is the persona.

Here is one prompt evolving level by level. Each step adds one layer of identity — Role (the job), Persona (the behaviour), Tone (the voice) — and the output becomes more controlled with every level.

Persona control means controlling how AI behaves — its strictness, its depth of explanation, and its attitude. A persona is not just a role; it is behaviour + depth + attitude.

Notice how the same topic gets very different outputs from these two personas — one challenges, the other encourages. Same role (interviewer/mentor), different behaviour — that is persona control in action.

A persona alone gives identity, but to make output stable and professional you combine three layers:

• Persona — behaviour.
• Constraints — rules.
• Format — structure.

This combination is what turns a one-time prompt into a reusable AI tool. The school-exam example shows why:

• "Write about water" → different answers from every student.
• "Write 5 points, each one line" → the same pattern from everyone.