sAIge

Duration: 7 weeks

My role: User Research, Analysis, Concept development, Physical Computing

Team members: Rebecca Hodge, Charlie Hou, Tanya Singh, Ruoxi Song, Jakob Prufer and Akriti Goel.

In collaboration with: Design Studio R/GA London and University of the Arts London

Design Question

“Can generative AI live in our everyday lives? How might this reposition the role of human intelligence?”

The Outcome

sAIge: A promo

A short film

SAIge is an AI- powered light strip that helps you practice difficult conversations from the comfort of your home. This video was produced to clearly communicate the use cases and intent of this personalised experience. It also helps understand the potential involvement of human intelligence.

The project brief was to develop a way for a person to pass as a ‘generative AI’ in an everyday setting.

The Process

Dividing into phases

We divided our work into two phases:

  • Phase 1: Understanding the human perception of AI
  • Phase 2: Deceiving this perception in order to pass as an AI

Phase 1 involved collating important pointers from secondary research and employing a vareity of primary research methods suited to the task.

Experimented with Sound-based applications
Bodystorming as an AI
A chat for a cookie at King’s Cross
Directed Storytelling with 30 people
Conducting AI Workshops
Group sketching Siri
Primary research methods employed

Crazy 4’s ideation

After thorough research and analysis, the Crazy 8’s ideation method was modified into a less intense Crazy 4’s which gave us a vareity of conceptual ideas. We were brainstorming ways of using Voice based AI assistants for everyday tasks like cooking all the way to buying medicines.

Crazy 4’s
Prototyping AI interfaces
Prototyping buttons
Prototyping AI voices
Prototyping and testing

Prototyping and testing

We created low-fidelity quick prototypes to understand what was working and what wasn’t. We prototyped all aspects of a human-AI interaction.

Low Fidelity Prototype replicating voice assistants
AI assistant working without a barrier
AI assistant working with a barrier
Compared different scenarios for human-AI interaction

Project process sketchnote

This sketchnote summarises the development and inter-relation of the key insights that played a role in shaping the final outcome sAIge.

The lessons I’ve learnt from this project

  • Having a solid foundation of research can be invaluable in the consequent stages of a design project especially when required to take design decisions.
  • Workshop design and facilitation requires careful consideration of time, activities and etiquette. Last minute workshops should be avoided since they can be mismanaged.
  • To aim for a high quality of documentation for a design project and to use optimal media to convey ideas, processes and outcomes.

For a more detailed process, please visit the project blog here.