Micro UX - Week 1

Brief: Design a way for a person to pass as a ‘generative AI’ in an everyday setting.

Methods: Bodystorming, Drawing, Secondary research, Artefact analysis, Turning Test

Team: Tanya Singh, Akriti Goel, Jakob Prufer, Ruoxi Song, Changlin Hou, Hong Zhou, Xiaole Zhang, Slavi Kaloferov, Qibin Cheng, myself.

---

Design away for a person to pass as a generative AI in an everyday setting

An unusual brief which took a moment to even understand what we were being asked to do. Paired with R/GA, we sought out their opinions, and they were keen for us to just go through the process - the outcome was still important, but getting there in the right way was more so.

Dissecting the brief, we understood that there were four parts to this brief:

Deconstructing the brief. Images by author.

Our initial discussion developed a number of questions we wanted to research, deciding to divide up the topics and use a communal online space to share what we have been finding.

Initial research questions. Images by author.

We also decided to experiment, explore and play, rather than try and obtain information purely from the written pieces, which wouldn’t have given us a working understanding of the topic. We tried bodystorming, drawing the human form of Siri, and also doing an artefact analysis on Siri.

Visualising AI, body storming scenario and artefact analysis. Images by author and Tanya Singh

As the brief was explicitly the opposite of the Turing Test, running one to understand how to flip it on its head was our next step. A selection of questions were asked, and two answers were returned; one from a human, one from Chat GPT. The outcomes were our first insight to understanding AI’s limitations, and why it is so different from humans. The AI had not been calibrated, something we wouldn’t understand until much later in the project, so the answers were formal, factual and impartial. In the end, we modified the human answers to sound more like AI (example below, answering the voice of a 5-year old) to fool those making the attempts to distinguish one from another.

Sample Turing Test Question: What does 'have' mean? Response 1: 'Have' is like when you have a toy or a snack that you can play with or eat. It's something that belongs to you or that you can use. It can also mean when you need to do something, like when mommy or daddy says, "You have to eat your vegetables." Response 2: To 'have' is to own, possess or hold. It can be used in a sentence such as 'I have a pen in my bag'.

Process of sorting the research. Video by Tanya Singh

We made sure to synthesise and analyse the data we were coming up with as we were encountering it this gave us the insights of the research questions and things that we wanted to know around social skills of AI its personality questions about testing of AI and the responsibility and a reflection back on humans as well.

Process of sorting the research. Images by Tanya Singh

This was a significant amount of research to undertake in the first week, however we needed to ensure we had a baseline for everyone’s understanding. Additionally, this helped us establish different areas of interest for the group split the following week.

The presentation was well received, and we were credited with trying a number of techniques to explore the topic. We replayed our Turing Test questions with the audience, to aid them in understanding the experiences we had encountered.

Slide from the presentation.

We were advised that we would need to consider the size of the group, the area of interest and refining the overall project plan. The feedback was fair; we knew that we would break down into smaller groups for the remainder of the project, we had had agreed to give ourselves the first week to form some opinions collectively to ensure the split went well.

The interesting part of this brief has been the insistence from the tutors and the project partners that we should focus on the process and not the outcomes. It gives us greater scope to follow our instincts, research findings, and interests. However, the lack of specific outcome or direction from the written brief means there is the potential to find ourselves lost quite easily with no clear direction.