Skip to content
Part of Human Value in AI ↗Family overview
Human Creative Value in AI

Framework

Ten kinds of creative responsibility

This framework is our interpretation, not a validated scale. It names human work that impressive AI output can hide. People judge whether an approach fits those it affects. AI can help with each kind of work. People remain responsible.

  1. 01

    Problem finding

    Notice what is wrong, missing or possible before anyone asks.

    “What are people working around rather than complaining about?”

  2. 02

    Problem defining

    Name the problem so people can act without losing what matters.

    “If we solved this exactly as written, who would still be failed?”

  3. 03

    Imaginative reframing

    Change how you see the problem: its scale, time, audience or metaphor.

    “What if this were a waiting problem, not a speed problem?”

  4. 04

    Surfacing assumptions

    Name what the first idea assumes. Then check it.

    “What would have to be true for our first idea to be right?”

  5. 05

    Synthesis of disparate views

    Hear conflicting views. Find what connects them.

    “Where do the nurse and the administrator actually agree?”

  6. 06

    Alternative futures

    Imagine several possible futures instead of extending just one.

    “What does this look like if it succeeds too well?”

  7. 07

    Discovering rare expertise

    Find people with lived or unspoken knowledge the data does not hold.

    “Who has solved a cousin of this problem in another field?”

  8. 08

    Recognizing talent & creative leadership

    Recognize leaders who choose the direction, bring people together and protect the purpose.

    “Who can hold the vision while others execute?”

  9. 09

    Intent

    Set an artistic or intellectual purpose. Check the work against it.

    “What should this make someone understand or feel?”

  10. 10

    Useful criteria

    Turn the purpose into clear checks for teams and machines.

    “How will we know a beautiful answer is the wrong one?”

AI may assist

Generate options, sum up views, list assumptions and suggest comparisons.

Humans decide

Choose the problem, the people, the checks and the acceptable risk.

Humans answer for

Answer for the effects on people, including those outside the room.

Apply it in the Studio →