AIX DESIGN CHALLENGE
OBJECTIVE
Decorating a home is an emotional, logistical, and aesthetic experience that homeowners have the task of doing. Many homeowners struggle with decorating their homes because of a lack of space visualization, disunified and confusing stylization, and discrepancies in finding and purchasing the perfect decor for their homes. These struggles can keep buyers from purchasing items, lead to overspending, and cause disunity in the room style. In the AIX portion of this project, my goal is to investigate the different challenges homeowners face to further understand how AI might enhance their decorating experience. This blog post captures the process I took to frame the design challenge and the research I conducted. Along with this, I utilized AI as a research collaborator to understand what abilities AI has to understand the human experience and provide insight as to how I may contribute to possible solutions.
Through external research and research conducted directly with AI, I am able to adequately outline specific pain points within the decorating process. This framework is the foundational stepping stone into how I may best solve the different problems that arise from my research.
AI USAGE
AI INTERVIEWEE
I conducted research directly with AI utilizing ChatGPT as an interviewee. To do this, I asked it questions that I thought might shed light to some potential problems that may occur with real humans. I began by giving it a prompted background explaining the problem, and I even told it what kind of personality to take on as it answers the questions to give me accurate answers. This would allow me two things: 1) It helps me understand what the extent AI understands about the human experience and how much information it already has about said topic, and 2) It allows me to anticipate what real people may struggle with which allows me to refine my questions to be specific for interviews.
To begin, here is the prompt I gave ChatGPT to step into character as an interviewee:
“Can you be an interviewee for me? I am researching for an interior design app that helps homeowners to visualize their homes with decorations and styles before purchasing. Create a personality that can match my interview questions”
I then gave it these following questions to answer while in the character I just prompted it to be:
What are homeowners’ most frequent struggles with decorating their homes?
How often to home owners spend money on decorations they don’t use in their home?
What is the biggest issue with creating a well decorated space?
What do you feel like is the biggest struggle when deciding a style for the interior design of your home?
How helpful would an app be to home owners if it allowed users to generate design styles for their homes?
How do you solve the problem when you have no idea what to buy for your home?
What do you need help most with for a well interior decorated home?
Following this interview, I utilized ChatGPT to break down and analyze the responses of each interviewee. To achieve this, I asked it to create an empathy map based on what this user “SAY”, “THINK”, “FEEL”, and “DO” regarding decorating his homes. Here is the response given by AI:
SAYS
“The biggest struggle is visualizing how everything will come together in the space.”
“Sometimes we buy things impulsively, but they don’t fit the space or match the style.”
“Finding balance is the biggest issue—avoiding clutter or making the room feel incomplete.”
“It’s hard to narrow down to one cohesive style with so many beautiful options.”
“An app that generates tailored design styles would save so much time and effort.”
“I often use Pinterest or Instagram for inspiration when I don’t know what to buy.”
FEELS
Overwhelmed by the variety of styles and decisions involved in decorating.
Frustrated when purchases don’t work out in the space as imagined.
Excited about tools that simplify the design process and make it accessible.
Confused about proportions, layouts, and how to create a cohesive look.
Empowered when finding inspiration or guidance that simplifies decision-making.
THINKS
Visualization is key to creating a cohesive design.
Mixing different styles is tempting but can get overwhelming or messy.
Proportions, textures, and layout are critical for a well-decorated home.
Budget constraints and existing furniture complicate design choices.
Apps with personalized recommendations could bridge the gap for homeowners.
DOES
Uses Pinterest and Instagram to gather inspiration and create mood boards.
Visits stores like IKEA or Target for ideas and to see how items look in real setups.
Sometimes buys decor impulsively, leading to unused or mismatched items.
Seeks advice from friends with good design sense when stuck.
Struggles with balancing functionality and aesthetics in a space.
Along with this experimental interview with AI, I conducted a SWOT analysis of 3 different apps to understand what solutions are currently available. The apps I compared were Houzz, Room AI, and Redecor. I used ChatGPT to analyze the information I provided from my own SWOT analysis so I could better visualize the key takeaways of what I should focus on. Here is the AI response to that request:
SWOT ANALYSIS
1. Real-Time Visualization Tools
Problem Solved: Homeowners struggle to visualize how furniture, decor, and colors will come together in their space.
Feature Idea: Enable users to upload photos of their rooms and use augmented reality (AR) or AI to show real-time, lifelike designs. They can see how colors, furniture, and decor fit together in their actual space.
Value: This eliminates guesswork and ensures homeowners can confidently design their space.
2. Product Recommendations Linked to Real Items
Problem Solved: Homeowners often buy decor or furniture that doesn’t fit or match, resulting in wasted money.
Feature Idea: Link furniture and decor in the app to purchasable items from retailers. Provide size, color, and price details to ensure everything fits their space and budget.
Value: This simplifies the buying process and ensures users can turn their designs into reality.
3. AI-Powered Style Suggestions
Problem Solved: Many struggle to decide on a cohesive style or mix too many styles, resulting in a cluttered or mismatched look.
Feature Idea: Offer AI-generated style recommendations based on user preferences, room layout, and existing furniture. Users can take a quick style quiz or upload inspiration to generate tailored designs.
Value: Provides a virtual interior designer experience, saving time and making design decisions easier.
4. Collaborative and Social Features
Problem Solved: Homeowners often turn to Pinterest, Instagram, or friends for inspiration but lack ways to collaborate in one platform.
Feature Idea: Add features for users to share their designs, vote on others’ creations, or get inspiration from a community gallery. Include collaboration tools so users can design rooms with friends or family.
Value: Builds a creative community and helps users feel supported and inspired throughout the design process.
5. Proportions, Layouts, and Balance Guidance
Problem Solved: Many find it hard to determine if furniture fits or if a room feels balanced and cohesive.
Feature Idea: Add tools to help users calculate furniture proportions, layout arrangements, and visual balance. Include templates for room layouts and tips for mixing textures, colors, and styles effectively.
Value: Helps homeowners create polished, functional spaces without over- or under-decorating.
CHALLENGES
While using AI, there are many benefits such as high-speed analysis and data breakdown, but there are some challenges with it as well. The first challenge I faced had to do with AI’s lack of ability to fully relate and pick up on human responses or behavior. To an extent, AI is great at analyzing based on the words of a human response, but it isn’t able to understand how each response may be reflected from an internal feeling or behavior. There is a disconnection that doesn’t allow AI to fully understand human emotion due to it being technology and humans being complex beings. The information I received, while helpful, was limited to the database of knowledge it currently has which is limited.
Another challenge I faced had to do with the prompts given to AI. AI sometimes misinterprets prompts or questions asked, and it can tend to stray away from what we need because of the lack of understanding. This means that AI needs hyper-specific prompts and detailed background of the project to understand the full picture. Detailed prompts and project needs will help AI know what the initial posture is for the project, the purpose, the tasks, and how to best move forward with user needs.
BENEFITS
The first benefit I experienced while using AI was the efficiency that flowed from this tool. AI is incredible at taking tedious, high volume tasks, such as research, and condensing it down into an understandable portion. When AI breaks down the large volumes of information, it gives it back to me and only highlights the most important parts per my prompt from the beginning. There are options to even ask for it to expand on each highlight when prompted to give me a deeper insight into the idea.
The second benefit I experienced is the time management that is utilized when using AI. Instead of spending hours and hours reading, documenting, and researching different blogs or articles about home decorating, AI is able to find the information for me. This cuts the back-end research down by hours which can allow me as a designer to spend more time on the important part: designing.
REFLECTION
For this section of the project, AI was extremely helpful because it cut my research time down by at least half. It also helped me both understand what direction I want to move in more specifically as I understand more what homeowners may experience. It also helped analyze the data in a simple and organized way. If I had to make any changes to how I utilized and prompted AI I would consider using it more as a collaborator and asking for other unique ways I could research for pain points outside of the typical interviews or article research.