Saige

2023, London

Team

_{Ruoxi Song, Jakob Prufer, Bex, Akriti Goel, Tanya Signgh,

Charlie Hou and project partner R/GA.}

Deliverables

Storyboards
Video production
Branding design
Interactive installation
User experience
Speculative design

Research Methods

Affinity mapping
Bodystorming
Popup survey
Workshop
Storyboard

An AI partner helps rehearse difficult conversations

In a world rapidly shaped by the AI revolution, it becomes imperative to consistently reflect on how these systems impact our connections with the world and among ourselves.

This project delves into fundamental questions: what is human and how can we distinguish between human system and machine? What can we learn by thinking that? What do we gain and lose when finding infomation from human as opposed to machines? How to turn humans into AI in an everyday setting, and how this benefits to humans?

Process

Affinity Mapping

This project starts with categorizing information, finding similarities and differences.

Text-based Turing test

Can people recognize human or AI from their
answers of the same question?

Bodystorming - human try to be an AI

Human act as an AI with or without external assistance. And what would wearing masks change the perception of the person pretending to be an AI? And how do humans think of “human pretended AI” just by listening to the voice?

“No matter what you do, you still need the human touch”

AI learning process is like a collective brain, a collaboration building chips into a brain.

What people feel about AI and voice assistant?

Most people envision the positive advancement of AI but they feel it cannot have the human attributes.

People do not want AI to make decisions for them or influence their way of thinking.

AI Exploration

We also tried apply filters to sound to make it sound more robotic.

Human in a conversation - The most irregular patterns, some are out of the area.

Human reading - Irregular patterns but in a consistent area.

“You cannot add human attributes to a machine”

Ideation

How to make human sound more AI?

1. Text to speech

We created a low fidelity prototype demonstrate text to speech keyboard, we made this by following insights on text and sound:

Text:

Use polite words
Long answers
Always provide answer “ask human for help" when it's about emotional

Sound:

Often assumed polite tone, formal language
Detectable inorganic patterns

2. Toolkit idea

Having a scoring zone or line while human trying to speak, and get a score saying how the human sound like an AI.

Testing & more research

What scenarios would AI be involved in everyday setting?

We created an activity of acting out a dating scenario in the workshop, we had two people pretending to be AI and gave responses to the human.

Ideation & storyboard

Talk to AI for dating specific conversations (AI as a group of human pretending).

More testing

Not feel comfortable if it’s human behind AI, and rather to talk with real person.
AI voice is very confident, would prefer more peaceful voice.
Answer is generic.