The setup of the research was crucial, as I needed to get a good understanding on how young designers approach, think and interact with generative AI tools. Besides literature review, I also decided to conduct my own ethnographic research.

I decided to first conduct a series of unstructured interviews, in which a free-flow conversation allowed the participants to dive deep into anything they'd like. This allowed me to grasp the initial patterns of interaction, as well as gain understanding into mental models of how designers think about AI.

A follow-up method to this would be a series of Design Fiction assignment. This method was chosen to further see how exactly the interactions with AI work within an actual design challenge setting. The participants had series of vague, yet challenging tasks during which they had to utilise AI tools and push the boundaries of their skills, as well as the limits of AI.

At the end of the design fiction assignments, the participants presented their results (as they couldn't see each others work beforehand, to avoid influencing them), and filled out a short survey to further note down their views, workflows and thought processes.

After the initial interviews, it became clear that my original focus - the interplay of Generative AI, Anthropomorphism in Design and Ethical use of these tools was too broad.

I decided to stop focusing on anthropomorphism in design - something initial research review showed was that the use of human traits is rather for skeuomorphic - trying to replicate "real world" (in this case the human-like conversation) for the sake of easier interaction for users, rather than being genuinely "human".

This allowed to further distill the research and analyse thoroughly the data into insights

After the conducting the research, I analysed the data and came up with a list of insights, which I then used to fuel further steps. I analysed each interview into a node-based system of thoughts and quotes. After that, I started recognising underlying patterns between different stages of "maturity" of the participants - commonalities and differences between 1st, 2nd and 3rd year design students, as well as working junior designers.
Here are some of the insights:

- Most users have very little knowledge about AI and how to work with the tools.
- Associated with the little knowledge, there is big "respect" or "fear" of the tools - some fear of being replaced by AI.
- A prominent use case for generative AI is inspiration and formulation of ideas - acting as a "springboard" for their creativity.
- It is easy to overlook ethical side of using AI tools, and what can happen when users of these tools aren't aware of hidden biases in the AI generations.

All of these insights were then put into a cohesive Design Vision, which you can see below.

The first thing to do in ideation was to purge myself of initial ideas. I brainstormed very quickly about possible solutions and ideas that could solve the main task - informing and engaging users into the world of (ethical) AI.
‍
On the left, you can see the result of the first brainstorm, in which I created in total 21 apps. I noticed that

After the brainstorm, I wanted to understand which of the ideas can be connected and form more complex solutions. To do this, I have decided to "weigh" the ideas on matrix, which you can see on the left.

The X axis of the matrix represented a Feasibility to develop at least a "proof of concept" level of prototype until the end of the set timeframe (roughly 1 month). The value was measured by technologies and tools required to carry out necessary level of detail in the prototype.

The Y axis represented the possible Weight of Impact that the solution would have on the users. This was measured by how much interaction, decision making and conclusions can be drawn from the solution.

Once weighed, I started marking ideas by colours, and linking the potential combinations between ideas to form complex solution.

Having a general idea of what the final solution should be - informative, playful, and with emphasis on ethical use of the AI tools, I set out to further generate, now more specific ideas together with the target users. For this, I created a creative session, during which different techniques were used.

At the start, I would introduce the co-creative session participants to the topic and the main issue - talking about the impact Generative AI has on design practices, and why ethical use is important in this topic.

This was followed up by a short brainstorm to rid ourselves of first ideas. For this brainstorm, I created a question to help generate ideas: HMW communicate an ethical use of gen. AI?

After initial brainstorm, we moved on to our main creative technique - Design Charette. During this technique, the participants were divided into pairs, and silently working on their ideas in short time periods - every 5 minutes, one of the participants from each duo would change their seat, bringing fresh view and ideas from other duo.

In total, this technique would go in two rounds for each of the following HMW questions:
‍
- HMW make an impactful experience prompting students to reflect on use of generative AI?
- HMW use gamification to help development of design skills?

After long idea generation, my plan was to converge with the target group through triading - a method in which three options are presented, and one has to go -> ultimately leading to a "top 3" of ideas.
However, due to time constraints of the session and availability of the participants, we had to cut down on that, as we were already past the planned timeframe. That's why I quickly decided to go in a round with the participants, each stating what they like the most about all the generated ideas.