Transforming Human-AI Relations: A New Paradigm

Uday Dandavate
2 min readNov 16, 2023

This article proposes using “professional roles” as metaphors for Human+AI interactions, avoiding emotionally charged personal relationships. It introduces five key archetypes and emphasizes essential boundaries for ethical design, focusing on transparency, user control, objectivity, and confidentiality.

I recommend using commonly understood “professional roles” as metaphors for Human+AI interactions, rather than relying on “personal relationships” like friend, parent, or teacher, which imply emotional bonds.

Professional Role Metaphors:

  1. Guide:
  2. . – Conducts background checks
  3. . – Provides tones of voice feedback
  4. . – Offers cultural sensitivity guidance
  5. . – Interprets written/spoken communication

2. Librarian:

. – Creates bibliographies

. – Manages citations

. – Searches for relevant quotations

. – Suggests images/videos/illustrations

. – Provides editorial services

3. Catalyst:

. – Builds customized social/professional networks

. – Enhances productivity in dating

. – Facilitates collaboration

. – Analyzes problem statements

. – Challenges/reframes questions

4. Research Assistant:

. – Conducts literature searches

. – Manages research operations

. – Assists in research design

. – Creates tools for data collection

. – Conducts analysis/synthesis

. – Designs compelling data visualizations

. – Develops visual illustrations of complex relationships

. – Builds presentations

5. Curator:

. – Compiles support materials for creative projects

. – Aids critical thinking/decision-making with relevant information

. – Builds topical archives

6. Concierge:

. – Provides travel assistance

. – Serves as a tour guide

. – Reveals hidden gems of a place/culture

. – Offers translation/interpretation/transcription services

. – Assists in event booking

Boundaries:

  1. Transparency:

Designers and marketers of AI-powered products and services should be transparent about capabilities and limitations.

2. Control:

Users should be in control of the conversation; AI should offer options and interpretations, not prescriptions or moral/ethical judgments.

3. Objectivity:

Engagement with AI should not be equated with interactions and relationships with living beings (people/pets) who have their own EQ.

4. Confidentiality:

Emotional revelations during interactions with AI should not influence the curation/personalization algorithm.

Behaviors a person wouldn’t want known by others should not be tracked, stored, or used by AI without consent.

These observations reflect my exploration into diverse citizen perspectives, unraveling nuanced metaphors and boundaries. Through insightful conversations, I derived meaningful insights, emphasizing the intricate tapestry of societal viewpoints. Ultimately, this journey provided me with a comprehensive understanding of the diverse landscape of citizen opinions.

I hope that these observation trigger design approaches that are mindful of human values and concerns.

--

--

Uday Dandavate

A design activist and ethnographer of social imagination.