Attachments
  • Mark as Completed
  • 10
  • More
Previous class
Section 1: Examples vs instructions

Section 2: Personas

Purpose of This Section

This section explains what personas are, what they are not, and how to use them safely and effectively when working with AI systems.

Personas are one of the most powerful—and most misunderstood—tools in prompting. When used well, they improve clarity and usefulness. When misunderstood, they can create false authority and over?trust.

The Core Idea

A persona is a behavioral frame, not an identity.

When you give an AI a persona, you are not giving it knowledge, judgment, or lived experience. You are guiding how it responds—its tone, priorities, and style.

Personas shape delivery, not truth.

What Personas Do

Personas influence:

  • tone and voice
  • vocabulary choices
  • level of formality
  • types of reasoning emphasized

For example, asking an AI to respond "as a lawyer" encourages cautious language and structured arguments. Asking it to respond "as a teacher" encourages explanation and clarity.

The AI is mimicking patterns of language, not assuming professional authority.

What Personas Do Not Do

Personas do not:

  • grant expertise
  • provide legal, medical, or financial authority
  • transfer responsibility
  • guarantee accuracy

A convincing persona can sound authoritative even when the content is wrong.

The Self-Reference Principle

This section intentionally demonstrates personas by using one.

The video avatar’s tone—confident, playful, or sultry—is a chosen persona, not an emergent personality. This makes the mechanism visible instead of hidden.

Seeing the persona in action helps users recognize how easily style can be mistaken for authority.

Common Failure Mode

The most common mistake with personas is deference.

Users hear a confident, role?appropriate response and stop questioning it. Phrases like “the AI lawyer said it was fine” or “the AI expert confirmed this” signal a loss of human oversight.

This is not safe use.

The Conjugo Rule

You assign the persona. You retain authority.

Before acting on an AI response, ask:

  • Do I understand this answer?
  • Would I trust it without the persona framing?
  • Am I still accountable for the outcome?

If the answer is unclear, pause and verify.

Best Practices

Use personas to:

  • improve clarity
  • adjust tone for an audience
  • explore different perspectives
  • generate drafts and options

Avoid using personas to:

  • outsource judgment
  • replace professional review
  • justify decisions
  • avoid responsibility

Section Takeaway

  • Personas are performance frames, not identities
  • Tone can feel authoritative without being accurate
  • Responsibility never transfers to the AI
  • Human judgment remains essential

Personas are powerful when used consciously.

This concludes Section 2.