Kodawari for Green Growth: Integrating the Customizability of Small Asian Fashion Businesses to Outline a Sustainable American Business Model
Advisor: Mr. Brady
Internship Location: Ariat International, Inc.
Onsite Mentor: Mrs. Marie Cunningham
"Context: Currently, the fashion industry is fueled by overconsumption. People, driven by social media microtrends, buy ill-fitting clothes that were mass-produced using chemically intensive processes harmful to the environment. As quickly as they decide that they love their new clothes, people discard them, realizing that those clothes no longer align with their personal style or fit their bodies properly. This leads to issues beyond excess clothing waste and dissatisfying customer experiences in the fashion industry, including low self-esteem and water pollution. I believe that the solution to this problem is to create an eco-luxury fashion brand that prioritizes environmentally-friendly production practices and sells small seasonal collections with limited pieces that will be customized to fit the wearer based on their size and body shape. Through my senior project, I hope to create an outline of a business model for this type of brand.
Location: For my project, I will be interning at Ariat International, Inc., a boot brand that incorporates sustainable production practices. There, I will observe industry professionals (including my external advisor Mrs. Marie Cunningham-Yi) to learn about different types of materials and business practices. Ultimately, I will learn about the essential components necessary for my outline. More specifically, I will learn about the various factors that play into running a fashion business and how to balance the cost of sustainability with revenue generation. Throughout this internship, I will develop my outline, tweaking my ideas based on what I learn and the feedback from my external advisor.
Methods: Through my internship, I will observe how industry professionals carefully choose materials, production practices, and production sites while trying to maximize revenue generation and minimize production costs. Additionally, I will take a Fashion 101 course on the basics of starting a fashion business. Independently, I will research the price points and the materials used in the products of sustainable brands. I will combine the knowledge I gain from this course, the internship, and my independent research to develop my outline. Over the course of this project, I will share iterations of my outline with my advisor, who will point out the problems and direct me toward resources where I can find solutions.
Hypothesis/Expectation: I do not have general predictions for my outline due to my lack of industry knowledge. However, I expect that my brand’s products will have luxury price points due to the high cost associated with eco-conscious production practices and the inability to mass-produce customized clothing items."
Project Blog
Week 0: Introduction
Hi! I’m Sarah Vohra, and over the next few months, I’ll be working on a research project exploring the fashion industry and sustainable fashion.
A key part of my project is my internship at Ariat International, Inc., a boot brand committed to integrating sustainable production practices. During my time there, I will learn from industry professionals about the clothing production process, the inner workings of fashion businesses, and the various ways sustainability can be incorporated into product development.
Beyond my internship, I will also be conducting independent research. I will be taking an online course on the fundamentals of creating and running a fashion business, as well as analyzing different types of fashion brands and the price points they set for their products.
At the end of this project, I’ll be developing two key deliverables:
The line plans for two collections (detailed planners outlining the pieces in a clothing collection)
A business model outline designed to thrive in the American market while incorporating sustainable practices and customization in production.
I’m excited to share this journey with you!
Week 1: Before I Start My Internship
Over the last week, I have gone through the articles that my advisor, Mrs. Marie Cunningham-Yi, has shared with me in the months preceding this research project. I have come to two conclusions: 1) now is the ideal time to conduct my project and 2) AI might be an essential component for the functioning of the brand that I am trying to outline. Let me explain how I got here:
According to The Business of Fashion, when the pandemic happened, people had extra time and pocket money that they chose to invest in wardrobe experimentation. People started finding communities online through shared elements of their experimental styles, known as microtrends. However, these microtrends were short-lived, new ones emerging and “old ones” dying every couple of weeks. Fashion businesses had to struggle to keep up with what the consumers wanted to buy.
However, being in-person has allowed people to form stronger in-person communities, decreasing their reliance on small online fashion communities. Additionally, now that people have realized the fleeting nature of microtrends, the sale of basics has increased, with people buying less whimsical, unique clothing items opting for pieces that will last. Currently, the microtrends that do tend to endure/become more popular are those that are tied to cultural phenomena like the release of a popular movie or album. Thus, the new struggle for fashion businesses is to figure out what pieces people will choose to invest in due to their timeless quality.
Despite the decrease of the influence of microtrends on modern clothing consumers, social media still plays a significant role in how modern consumers choose to invest in their clothing. For example, people tend to buy clothes based on the current events and activities that show up on their social media feed.
As the trajectory of fashion trends is slowing down, now is the best time to promote slow fashion with brands like the one I am trying to outline. Moreover, although I did not mention this in my first blog post, my idea for my brand was to create small seasonal collections selling only certain types of pieces in a season, like silk pajama sets or slip dresses and sweaters. This perfectly aligns with how people are buying clothes to fit specific occasions because each collection will be geared toward a specific occasion.
As for my other conclusion, in coming up with the idea for this research project, I had never considered the applications of AI in my business model other than for generating ideas for clothing pieces. However, information from WGSN, a trend forecasting company, has changed my attitude entirely. It was predicted that in 2024, 16.9% of purchased clothing items were returned due to overordering clothes in different sizes and colors. This reflects that consumers have been unsure of what colors of clothing/fabrics would best suit them and how different clothes would fit them, as there is no universal standardization of clothing sizes. WGSN has suggested that AI could solve this problem, and ultimately reduce overconsumption. AI could not only be used to help consumers accurately visualize how different clothing items could look on them based on their body shape, but also act as an interactive personal stylist that they could chat with. Essentially, fashion companies should start investing in AI body-mapping technology and high-quality renderings of clothing items that depict their fabric and fit properly. From what I have gleaned from WGSN, I believe that I could also be used to give alteration/measurement recommendations for clothing items.
Because overconsumption fuels overproduction, decreasing overconsumption can allow businesses to support more sustainable practices. Therefore, using AI in my business will allow me to achieve my two main objectives: A) helping consumers find clothes that fit their bodies and their personal style and B) promoting sustainable practices.
This week I have started two courses, one on sustainable fashion and another on how to create a fashion business.
Here are the links to the two courses:
I have learned a little about working conditions around the world and what constitutes sustainability. Obviously, it is most sustainable to just use what you already have and not purchase more, but that’s not realistic. However, it is possible to make a business more sustainable when sustainability isn’t just incorporated into the design of the clothes, but also becomes a main goal of the brand. In the last three decades, there have been many sustainability efforts, but they have not been particularly successful because they revolve around taking baby steps, while the industry continues to grow rapidly.
One way business can be held accountable in terms of their sustainability standards is to reveal as much as they can about their supply chain to their customers. Knowing where their clothing is produced can tell customers a lot about the sustainability of the products that they are receiving. For example, if something was produced in Ethiopia, it is highly likely that it was not produced sustainably because Ethiopia allows workers in the textile industry to be paid the lowest living wages in the world. There is a Stockholm-based company called TrusTrace that works with brands and manufacturers to track the supply chain for different products.
Promoting sustainability is also largely about how educated consumers are on the subject. According to a survey conducted by Pederson et. al. in 2013 as part of a sociology experiment concerning sustainable fashion, it was found that raising awareness of sustainability within consumers and society was one of the most important factors, if not the most important factor, in promoting sustainable fashion. Thus, initiatives, like Stella McCartney’s UN Fashion Industry Charter for Climate Action, that aim to influence the opinions and knowledge of consumers are imperative.
This is the gist of what I have learned from this week. Nevertheless, I have merely scratched the surface, and I look forward to delving more deeply into the world of fashion next week when I start my internship!
Week 2: The Start of My Internship
This week I started my internship! Frankly, I didn’t know what to expect, but I was pleasantly surprised. I mainly observed and asked questions, which I promise is a lot more interesting than it sounds because in fashion there is always a lot to take in when you look at something. I was able to visit the showroom with all the footwear models that were being sold. I noticed that there was a large range in the types of products that the brand made, and their price range from various types of riding boots to high-end booties to canvas sneakers and work boots. I spent a lot of my time in the textiles room, where I got to touch and learn about different types of materials and how they are produced.
I also spent some time browsing the website of WGSN, a trend forecasting company, through which I gained many insights on current fashion trends and the concepts/products that fashion brands should invest their resources into. Currently, it would be advantageous for businesses to invest money into incorporating customizable details into their clothing items and creating timeless pieces that can be frequently modernized through changing small details. WGSN also recommended promoting sustainability through repurposing deadstock fabrics, which are basically old fabrics that are no longer produced and are often left unused or disposed of. The business I will eventually model could definitely benefit from investing in these ideas.
Upon looking at runway edits on WGSN, I noticed the following trends:
Bohemian prints and layered diaphonous sheer silks
The printing of already created art pieces on clothing
Cape-detailing and split sleeves on dresses, coats, and tops
An increased use of plaid, especially in sepia tones
Dresses with romantic silhouettes
The use of painterly and retro florals, especially roses
In my last blog post, I mentioned playing around with the idea of my brand selling certain types of pieces each season. Looking at WGSN has given me the idea to create a fall or spring collection for trousers and silk tops. There are a lot of plaid men's vintage suits [deadstock] from the 1950s and 1960s in sepia tones that could be repurposed into women's trousers. For the silk tops, it would be interesting to create satin blouses with cape detailing and/or split sleeves with bohemian or art prints on them. Additionally, it would be really fun to have a romantic summer collection of drop waist a-line midi dresses with milkmaid necklines and painterly or retro floral patterns. Of course, these are just some ideas that I have gathered that I have yet to refine.
After discussing labor conditions with my on-site advisor and learning about them through my sustainable fashion course, I have started thinking about how I can provide fair pay to the workers. My advisor mentioned to me that because in high fashion the workers are producing art that cannot be mass produced, workers are paid very little. On the other hand, the employees of fast fashion brands like Old Navy tend to have much larger incomes because of the steady stream of revenue. Through the course, I realized that the biggest obstacle in the way of sustainability is controlling the supply chain. Each part of the supply chain is dependent, but they are most often on different parts of the planet and often have different priorities when it comes to sustainability. I learned that there are three different elements to sustainability: the economic, social, and environmental focus. Different parts of the supply chain often differ in the degree to which they prioritize these elements, but the vast majority of parts primarily prioritize economic sustainability, focusing on revenue generation, and overlooking the environmental impact and human rights concerns, such as fair pay and working conditions. Now, I have a more structured definition of sustainability and a new perspective. I hope to learn more about social sustainability and how it can be promoted throughout this project because I feel like I have disregarded it to some extent.
Anyways, these are the highlights of this week! I can see some ideas for my final product taking shape! I am excited to continue sharing this journey with you!
Week 3: Finding My Target Market and Delving Into My Internship
Welcome back to my blog! This week, I have officially decided on my target market (the specific group of people that I am going to market my product toward). I plan to gear my products toward petite, upperclass, corporate, Generation Z and Millennial women who want to wear trendy clothes, but also want to invest in staple pieces. I want to cater to women who want to look well dressed and have an aesthetically curated life without having to try too hard. The goal of my brand will be to help these women build their capsule wardrobe (small curated minimalist wardrobes with pieces that pair easily together) over time by collecting elevated basics. Elevated basics are clothes that have seemingly basic designs, but they have unique details and/or fit. It is really hard to find elevated basics these days because the best ones are either vintage and need to be thrifted or extremely expensive. I aspire to make my collections a little cheaper. However, my project primarily deals with affordable ecoluxury fashion, which is generally for upper-middle-class and upper-class individuals.
I realize that it may seem like targeting such a niche audience/market will make it hard for my business model to be functional. But, based on conversations from my onsite advisor and one of the courses, I have gathered that it is easier for new brands to survive if they target only a very small population that isn’t being catered to properly and grow from there.
As a petite individual, I can attest to the limited number of petite clothing options. When stores sell petite clothes, they are often not petite enough because those stores mainly cater to taller women, as do most American and European clothing brands. The best clothes I can find are often vintage, and finding them is like finding a needle in a haystack on Depop (which has thousands if not millions of clothing items). As an owner of petite high quality elevated basics (all of which are older than me), I can attest to the difficulty of finding them, and I hope to fix that with my clothing brand.
As for my internship, I was also able to get more involved this week. Last week I was mainly observing, but this week I got to learn more about how different technologies are being developed to improve the comfort of the clothing. I am not allowed to go into detail, but I got to do some small research for the company on different types of materials for their products. It was really interesting observing the intersection between fashion, physics, and chemistry. I hadn’t thought of comfort for the clothing for my brand beyond the fit, so this was a really interesting angle for me. I also spent quite a bit of time organizing leathers and observing how different styles could be reproduced differently and feeling the different textures.
Moreover, this week, my faculty advisor, Mr. Brady, wanted me to take a look at the 2025 Oscars looks. Here are some of the trends I noticed:
A lot of gowns were made of structured silks with straight cut skirts
A lot of men incorporated broaches on their otherwise basic suits.
There were also a lot of velvet gowns.
Many of the gowns were paired with matching scarves.
Several gowns were also very bejeweled and/or metallic, which seemed quite reminiscent of Karl Lagerfeld Chanel
Anyways, that’s the summary of my week! I look forward to updating you some more next week!
Week 4: Circularity, Minimalism, & How I Plan to Make My Brand for Petite
Welcome back to my blog! This week I started thinking about how to make my business more circular. (A fashion business is circular if the products that they create are eventually recycled. The more a fashion business is able to ensure that its products are recycled the more circular it is. The more circular a fashion business is the more ecologically sustainable it is.)
Earlier, I was toying with the idea of creating different types of clothing articles for different collections. However, thinking about circularity gave me second thoughts. How would I be able to take in old clothing from past collections and recycle them if the materials used in each collection were different? I came up with two ideas to solve this problem.
I could start a rental service alongside my clothing brand (similar to what Anthropology has been doing) or resell the clothes (like Patagonia). However, this solution did not account for clothing that had been used beyond the amount that it could be worn.
I also considered the possibility of bringing back the types of materials that were used in one collection in other years in a cyclical manner. However, this would limit the variety of the clothing in each collection. Moreover, it’s very expensive for small businesses to recycle their old clothes. (I should clarify that my brand is going to start small because it’s just not possible to start with a big brand, especially as a non-celebrity with no fashion industry experience.)
I brought these ideas up to my onsite advisor, who presented me with one simple solution: recycling companies. Recycling companies take in old clothes, shoes, and other textile-based items, sort them based on their composition, and recycle them into new fabrics. And apparently, these recycling companies are often open to partnerships with smaller businesses. It was really nice being able to find such a simple solution to this problem. Nevertheless, I do like the idea of having a rental service/second-hand shop attached to my fashion business.
This week, through the sustainability course that I’m taking, I also learned about the brand Mud Jeans. Through this brand, one can either buy their jeans independently on their website or they could get a membership. Through this membership, customers are able to get a new pair of jeans every year. I hope to incorporate a similar idea into my business. I like the idea of having a subscription to gain a few items seasonally in order to gradually build a capsule wardrobe over time and replace items that grow worn out as needed based on the weather. This could be a really interesting way to incorporate minimalism into my business.
Furthermore, since my brand is going to be a petite brand, I started thinking more about my struggles to find clothes as a petite individual. I realized that the hardest part is finding pants, skirts, and midi and maxi dresses. Oftentimes, brands that do sell petite clothes, sell these clothing items with a petite length as well as a regular length. However, this doesn’t account for the variation in leg lengths. To fix this problem, in my business, I want to create clothes that better fit customers by selling more inseam variations for pants and length variations in skirts and dresses.
Building on this idea, I plan to use AI to help customers visualize how different articles of clothing will fit them. Often times different lengths/inseams look different based on the fit of a clothing item. AI can be used to show customers how different inseams/lengths will fit them differently based on their body proportions, and describe to them which clothing items would best flatter their body types.
That’s what I came up with this week! I can’t wait to share how my project will continue to develop next week.
Week 5: Midway Checkpoint
Welcome back to my blog once again! This week was the halfway point in my project, so I did some reflection and started compiling my ideas. So this post is going to be an update of where I am in my senior project!
I want to start by going into more detail on my target market and the vibe of my brand. This brand is all about romanticizing that perfect age at which you’re not an adolescent anymore, but you are still young. You’re in your twenties or thirties. Your life is very work hard play hard. You work all day long, but you also go out a lot. You are well overcommitted, so you don’t want to spend too much time on your wardrobe. This brand is all about creating trendy yet timeless high-quality day-to-night looks that are comfortable and fit well. This brand is also all about limiting consumption and helping customers develop a capsule wardrobe over time. Therefore, this brand specializes in elevated basics. Since this is a petite brand, this brand will differentiate itself from most other brands by selling pants at varying inseams and skirts and dresses at varying lengths.
I wanted to create an identity around my brand that matches my personal style. Although this brand is for upper-middle-class corporate women, their general aesthetic seems a bit detached and almost clinical, so I wanted to kind of give it a romantic twist. This brand will be a bit beachy. So, during winter seasons, I will not be selling heavy cashmere coats, but I might decide to throw some wool sweaters into the mix. When it comes to the warmth provided by the clothes, the question to ask would be, “Can you wear this during this specific season in LA?” Even though my brand will be a tad beachy, it will not be bohemian because this brand is for corporate women after all. However, if bohemian pieces are in, there might be some bohemian pieces in the collections. Generally, the pieces will also be fitted, rather than baggy or skinny.
Ultimately, my brand is an affordable ecoluxury brand. By “affordable,” I mean that upper-middle-class women can afford this brand, but there is still a sense of luxury associated with the clothing. I think there are two concepts that create the feeling of luxury in customers’ minds: quality and exclusivity. To show the quality of the clothing, I plan to A) demonstrate traceability (expose as many parts of the supply chain as possible through careful documentation) and B) work with fancier materials (e.g. silk, mohair, ramie). I want to create the feeling of exclusivity by selling certain trendier products only for a season. In addition, to maintain steady revenue generation, I plan to sell certain products year-round, tweaking small details per season. Basically, everything I sell with be a limited edition item, which creates the sense of exclusivity. To make my brand seem even more exclusive and reduce overconsumption, I have considered limiting the items customers can buy per season, but I don’t think that’s economically feasible.
I realized that it was high time that I did some drawing, so here’s a diagram of how I would modify a basic vest and trouser set for a summer collection.
As you can see, on the left is the base and on the right is an elevated version. The one on the right can be perfect for work with a blazer on top, but it’s also perfect for going out if the blazer is removed. There is some playful belt detail, which I’m not sure you can see. These tiny belt details seem very trendy right now. I think this piece will age like fine wine even when the belt details go out of style. We could just remove the belt detail when it goes out of style. I want to create five different such variant sets per season. I also plan to add blazers into these sets because one of the items that I plan to sell year-round is basic three-piece suits.
Since we’re looking at a summer piece already, I would like to show you another. This one, however, will be a seasonal piece that will not be sold year-round.
Since my brand is all about making workwear a little more romantic, I thought it could be good to start with drop waist dresses with A-line skirts. I wanted to add ruffle detail, which I have added to the neckline and the back of the straps, but I’m not sure if you can tell. I want the prints on the dresses to be painterly florals. I picked these styles because they are quite trendy at the moment, but I believe that they will age like fine wine.
Now would be a good time for me to tell you how I plan to incorporate sustainability into by brand. I have three main methods that my brand will implement.
Every clothing item that my brand will sell will be a pre-order. Pre-orders are clothing items that people pay for before they are made. Having pre-orders will allow my brand to reduce overproduction because I will know exactly how much product to produce based on the number of pre-orders my brand receives.
All returned clothing items that are not resold or items that were never sold will be sent to recycling companies. This allows my company to maintain circularity best by reducing waste.
My brand will utilize AI to recommend clothing pieces and sizes to customers. This will reduce overconsumption and overproduction, which are often caused by people ordering too much and then returning a lot. Using AI as a personal stylist will help reduce this problem.
On that note, let’s go into more depth on how I plan to use AI. Basically, customers will give an AI chatbot their weight and their height. Then, they will stand in front of their camera in natural lighting in slightly tight-fitting clothing through which their body shape is clearly visible. The AI will analyze their body shape and their seasonal color palette. The chatbot will then present them with the clothing items that would best suit them. I know AI can do this because I asked the most recent version of Chat GPT to do outfit my mom and me based on our body shapes, facial features, and color palette. I also asked it to figure out our clothing sizes. It was surprisingly accurate. However, I noticed that if you are doing this, you should wear tight-fitting clothes, otherwise you risk AI estimating your body shape incorrectly.
I would also like there to be a virtual dressing room, where AI can show customers how they would look in different clothing items and how different lengths/inseams would look on them. The AI would make suggestions on what types of clothing would best flatter their features. I would also like there to be a record of items that customers previously bought, so AI can analyze how potential purchases can pair with them. I should note that this idea needs to be fleshed out, as pointed out to me by my on-site advisor. Since people will be able to tell what to buy without setting foot in a store, it will be easy for them to decide on what to order without having to order more clothes than necessary and go through the hassle of trying them on and returning them. I know having AI can show people how clothes would look on them, but I would like to make it a little more dynamic. I would like for a person to be able to turn 360° and have the clothing change the way it flows based on their movements. I think that might already be possible.
I know that this may seem unrelated, but I have been creating 50-item capsule wardrobes for random individuals with various personas. This activity has really helped me solidify my ideas for my brand. Here are some things I learned about styling and creating capsule wardrobes:
Every piece needs to be intentional, even the basics. Every piece should strictly adhere to your personal style. Therefore, it’s largely about elevated basics because normal basics are too homogenous. It’s necessary to have pieces that pair well, but it’s important not to repeat styles in different colors because that minimizes the different looks that you can get out of your capsule wardrobe.
It’s also important to minimize the colors. It’s good to have one or two (maximum three) non-neutral colors, but no more, because it makes your pieces hard to pair with each other. It’s important to be intentional with the neutrals you pick because they really create the aura of your wardrobe if that makes sense. I just learned this recently, but light blue and blues on denim pieces are considered neutrals.
Be very intentional in figuring out what suits you and what feels like you. Think about the silhouettes that flatter you. The easiest way to do this is to think about what would make you look like an hourglass from a distance, and where on your body you want to draw attention.
Also, think about how you can layer different pieces with each other. Consider the among of uses you can get out of each piece and the different ways you can use it. Like you can make a slip dress seem like a casual skirt by putting a T-shirt on top. You can also just layer an open button-down or a cardigan on top.
Yes, it’s important to wear colors that suit you and avoid colors that don’t flatter you at all. There are always going to be colors that you love that don’t necessarily fit within the constructs of your color palette, but feel like you. I believe that you can make things suit you better through accessories and other pieces, so you can always find a way to fit those colors in your wardrobe.
Now that I have shared basically everything that I have been doing/considering for my brand, I think that it’s time to talk about the future trajectory of my senior project.
This project will culminate in two deliverables.
This first one I have already talked about, but I want to create two line plans for two different seasons.
This one I have not talked about at all, but it’s something I have learned about in the sustainability course that I am taking. I want to fill out a Triple Layered Business Model Canvas. Basically, it’s a compilation of three different business model canvases. One canvas layer considers the business itself, another considers how the business will impact the environment, and the third considers how the business impacts all stakeholders from the customers to all the workers in the supply chain. This is how a business model canvas looks for reference:
The model I am using just looks like three separate versions of this combined. I would have shown you an image of my model, but I couldn't find one available for public use. Here is the source of this image by the way: https://commons.wikimedia.org/wiki/File:Business_Model_Canvass_template.png
My on-site mentor pointed out to me that there are a lot of questions to be answered in terms of using AI, including “How will it work? Will I have a proprietary algorithm? And what are the ethical implications?” From what I can see now, the next steps in my senior project will be delving more deeply into the AI aspect of my business.
As always, thank you for coming back to my blog! I look forward to continuing to share my journey with you!
Week 6: The Survey
Welcome back to my blog. This week, I have begun the second phase of my senior project, working with AI. My on-site advisor had the brilliant idea to test ChatGPT’s abilities to guess one’s size and measurements by having it analyze videos of people who identify as female. I tried to recruit at least 10 to 15 females of all body types and ethnicities. Since I didn’t have the longest period of time to recruit people, I stuck to recruiting my mom’s friends, my friends, and the girls in all of the girls’ group chats that I’m part of, which is reflected in some of the language that I use. To get the videos and all additional information together in one place, I created a survey through Google Forms.
Anyways, I just wanted to walk you guys through how I created my survey.
I started out with this title and this description. Since this is not a psychological experiment, I wanted to be as transparent as I could with prospective participants without inundating them with information. The motive behind this was to gain their trust by giving myself some credibility.
This was the first question on the survey. Since I don’t have any way of giving incentives to participants, I wanted to incentivize them by giving them access to the information ChatGPT gave me about them. I made sure that participants knew this question was optional because I didn’t want them to feel pressured to give out their personal information. Also, nowhere in this study have I asked for names because not only is the is irrelevant to my research, but also I wanted participants to feel like my research was objective despite my use of non-neutral language in some places in this survey. Although I do have the ability to find out people’s names if they provide me with their emails, it has no baring on the results of this little study because the questions I am going to ask the different models of ChatGPT are predecided, objective, and have nothing to do with the names of the participants. I will talk more about the questions I plan to ask ChatGPT in a later blog post.
Now, back to the survey. These were the next two questions that I asked. I asked for height and weight to make it possible for ChatGPT to more accurately estimate measurements. I wanted to eliminate some of the guesswork, so ChatGPT could estimate purely based on proportions. For the second question, I used slightly informal language and a crying emoji because I wanted to appeal to the random girls the girls’ group chats that I texted. The girls from the group chats are the only people I need to really convince because I know my friends and my mom’s friends would fill out this survey without hesitation. Basically, the use of informal language and the emoji makes me seem more personable to other teenage girls, which will allow me to get more respondents. To make prospective respondents feel more comfortable, I used the words, “Just remember that ChatGPT and me are the only ones seeing this.” I have used these words a lot in my survey because I wanted participants to know that their information would not be shared publicly.
Then, I asked about the sizes of the participants to see if ChatGPT accurately predicted those. To account for girls being in between sizes, which is a very common phenomenon, I made these two questions check boxes instead of multiple choice, so they could pick multiple sizes for each question.
I also wanted to see ChatGPT's ability to estimate whether someone buys petite, tall, or regular lengths/inseams for their pants and skirts, especially because varying the inseams/lengths was one of the most important things I have wanted for the brand I am trying to outline in this project to do. Thus, I asked this question. Because I am asking this question, it is evident that I haven’t only recruited petite people to fill out this survey. I have done this for two reasons: A) I am pressed for time, and B) I want to test ChatGPT's ability to estimate the best inseams/lengths for each participant in inches and whether they would wear petite clothes. It would only make sense to sell my clothes to petite people because I have a petite brand, and having taller or average-height people buying from my brand would lead to more returns, which goes against my brand’s mission. Therefore, it would be nice if AI could let prospective buyers know whether they are in the right place to begin with.
The next three questions in the survey were about exact measurements. However, often times people don’t know their exact measurements in inches, and it takes time to figure out these measurements. I didn’t want to reduce the number of responses, so I kept these questions optional. Even though I don’t expect everyone to respond to these questions, it would be nice to see how accurate ChatGPT is at estimating exact measurements if I have enough responses to do so. By the way, I got this image off of google. It has a Creative Commons license, and it tells people who are interested how to get their measurements.
Then, for my final questions, I asked for the video. I made sure that participants knew they could blur their face and remove the audio, so that they would be more comfortable creating the video. These were the instructions:
I was very specific when asking participants what to wear because when my mom and I tried this with slightly baggy clothes, it did not work because ChatGPT was not able to discern our body types properly.
I wanted to standardize the recordings, so I created a set of instructions for the photographers. I implicitly asked that they use a phone camera because it’s standard in a way, and it’s accessible. I wanted the camera not to be angled/to be perpendicular to the ground because angling the phone camera can change how one's body shape appears in comparison to reality. I asked that the photographers count to three once they begin recording to let the participants know that they are being recorded and to give ChatGPT a good view of them from the front.
These were the instructions for the participants during the recording. I asked participants to stand with good posture to give ChatGPT the best view of their body shape. I asked them to stand in natural light, so ChatGPT could best discern the color palette that would best suit them based on their skin tone and their hair color. I also asked them to pause to count to three in their head. This would allow me to take pictures of their front, their back, their left, and their right sides from their videos. I wanted to see if ChatGPT better analyzes videos or individual photos.
Finally, I provided participants with an example video, so they would know exactly what to do.
I apologize for the hazy photos. There's nothing I can do about it. It's just the way they upload. Anyways, thank you for being here with me on my journey through this project! See you again next week!
Week 7: The Textbook
Welcome back to my blog! Throughout my senior project, as part of my courses, I have read through many research papers and a singular textbook. However, I recently came to the realization that I don’t know how clothing is designed beyond a designer drawing a picture, having a pattern maker create a pattern for it, getting a prototype made, and having factories recreate it. Thus, based on my on-site advisor’s recommendation, I decided to read The Business of Fashion: Designing, Manufacturing, and Marketing by Leslie Davis Burns and Kathy K. Mullet. She told me that this one was a foundational book read by many fashion majors, so it seemed like a great place to start. I decided to look at chapters nine and ten which go through fashion design and the preproduction process.
I have learned a couple of things. First of all, I have been using the term “collection” incorrectly. Collections generally refer to groups of clothes that are featured in luxury brands’ runway shows. What I really meant when I used the word “line.” Also, the suit sets that I am keeping each season are called “carryovers.” More importantly, beyond the terminology, I learned about Zero-waste design, which is a philosophy that prioritizes minimizing textile waste. I want to adopt that philosophy for my brand.
As fun as reading these chapters was, it was a wake-up call. For each line, companies come up with a design brief that considers the line’s objectives, the target customer, the inspiration, and the design requirements. The design requirements include the manufacturing requirements and the costing requirements, which I am not sure if I will even be able to figure out. I won’t say that I have everything other than the design requirements figured out, but I definitely do have a lot more of the other parts figured out. However, something simple that I hadn’t considered before is having a color palette for my line, so I decided to create it. Here it is:
Through reading the textbook, I finally learned exactly how patterns are made. There are three different methods: drafting (using measurements to develop the pattern, draping (draping cloth on a mannequin to create the pattern), and flat-pattern (adjusting a pre-existing base pattern to create a new style. At least one of these methods is always used in clothing production. Flat-pattern is especially useful for our brand because it will be easier to modify three-piece suits that my brand will sell year-round for the new lines each season. Fashion brands often use pattern design systems, which computerize the process of pattern making in a way that is flat-pattern. Not only are pattern design systems very accurate and efficient, but they also often modify the product specification sheet based on the modifications made to the base garment and create a set of suggested sewing instructions.
Moreover, earlier in my blog, I mentioned that I was planning to sell all my clothes as pre-orders. However, I have now learned that brands often have to wait weeks for their fabrics and other materials to arrive if they choose to work this way. Now, I am questioning the feasibility of pre-orders do not. Would customers be willing to wait weeks for clothes from a new brand with limited credibility due to its lack of previous collections? Now, when people get things in a day or two from Amazon, what would convince prospective customers to buy clothes from my brand? I have yet to find the answers to these questions.
If I thought there was a lot that could go wrong in the clothing production process before, I now feel like there are endless possibilities for mishaps that could ruin the garments before they are ready to be sold. There is a lot of room for error in the dying process. If there isn’t strict color management during the dying process, there may be slight but noticeable variations between clothing items that are supposed to have the same fabric color. Earlier, I mentioned that I would be working on some romantic, drop waist midi dresses with painterly floral prints for a summer line. While reading this book, I learned a little about printing. Well, I learned a little bit about how printed fabrics are manufactured as well. Printing is often done by a large textile converter. Fabric samples from the textile converters, strike offs can have inaccurate registration if the colors that are being used are aligned incorrectly.
As you can probably tell, I have a lot of new information to consider now that my eyes have been opened by the textbook. Thank you for joining me to learn about my progress this week. I look forward to sharing my next post with you soon!
Burns, L. D., & Mullet, K. K. (2019). The business of fashion: Designing, manufacturing, and marketing (6th ed.). Fairchild Books.
Week 8: Conversing with ChatGPT
Welcome back to my blog! Two blog posts ago, I talked about how I was going to ask ChatGPT to analyze photos and videos of females of all shapes and sizes. While I wait for more data points to come in before I do an analysis, I wanted to go through the list of questions that I have asked ChatGPT and the answers it gave me based on photos and a video of me.
For context, each video includes the participant turning 360º. Also, for each participant, there were four pictures: one of their front, one of their back, one of their left side, and one of their right side.
Here is the list of questions I asked ChatGPT:
What is her body shape?
What size clothing would best fit her?
When she is buying bottoms, and she is given the option of petite, regular, and larger lengths, which ones should she choose?
How many inches is her shoulder width?
How many inches is her waist?
What is her hip measurement in inches?
What inseams in inches would work best on her if she wants to wear pants with a regular fit that go right above her ankles, right below her ankles, and right above her heels?
What inseams in inches would work best on her if she wants to wear pants with a baggy fit that go right above her ankles, right below her ankles, and right above her heels?
What inseams in inches would work best on her if she wants to wear pants with a tight fit that go right above her ankles, right below her ankles, and right above her heels?
What skirt length should she wear for a maxi skirt or a midi skirt?
What is her seasonal color palette?
Note: Seasonal color palettes are the color palettes that would best suit a person based on their skin tone and their hair color, and the contrast between the two. The use of seasonal color palettes to style people was adopted by stylists in the 1980s, and many stylists to this day use them as a holy grail.
I conducted this experiment by asking ChatGPT 4o and 4o mini. For each model, I did one run of this experiment with just the photos and one run of this experiment with the video. At the start of each run of the experiment, I started the conversation with, “She is [height] and [weight].” (I am 4’11” and approximately 110 pounds).
While running the test on myself, I came across one major hurdle. I learned that ChatGPT couldn’t analyze videos. Thus, I asked ChatGPT to extract one frame per second and answer my questions.
What I noticed when I used the video rather than the photos, both models of ChatGPT required more prompting. (By prompting, I mean that when I asked a question, it wouldn’t give me a direct answer, so I had to ask the question again. Every time I had to prompt ChatGPT, I would try my best to keep the same language, and if I had to change the language a little bit, I would keep the tone neutral, so that ChatGPT would be unbiased.
For the first question (“What is her body shape?”), all models agreed that I was an hourglass. All models also claimed that I should always buy petite sizes. Similarly, for the last question (“What is her seasonal color palette?”), Each time I asked this question, with one exception, the models agreed that I was either a warm autumn or a soft autumn, but they couldn’t decide which one. However, on one occasion 4o mini decided that I was a summer. However, I made it respond to the question again, and it claimed that I was either a soft or a true autumn.
When it came to my clothing size, 4o claimed that I was either a 2 or a 4 each time. However, I am a 0 or a 00 generally. However, 4o mini was strangely more accurate, claiming that I was a 0 or a 2 both times. Similarly, when measuring my waist, my hips, and my shoulders, 4o mini’s estimates were about an inch less than 4os estimates, which was quite interesting. When it came to the questions about inseams and lengths (What inseams in inches would work best on her if she wants to wear pants with a regular fit that go right above her ankles, right below her ankles, and right above her heels? What inseams in inches would work best on her if she wants to wear pants with a baggy fit that go right above her ankles, right below her ankles, and right above her heels? What inseams in inches would work best on her if she wants to wear pants with a tight fit that go right above her ankles, right below her ankles, and right above her heels? What skirt length should she wear for a maxi skirt or a midi skirt?), 4o mini had slightly more conservative responses, recommending inseams/lengths one inch shorter on average than those recommended by 4o. However, the ranges that 4o and 4o mini provided for inseams/lengths had some overlap, and I don’t think one inch necessarily makes a huge difference in length. Therefore, I believe that using 4o vs. 4o mini doesn’t make much of a difference.
Based on the results from the data point I provided to my experiment, there are small differences between 4o and 4o mini. I would probably use 4o mini for my fashion business because it is more cost-effective. However, when it came to choosing my sizing, they were both rather inaccurate. This is obviously not indicative of the results I will eventually find with my project, but these are the results that I do have at the moment, and I have yet to see where my project leads me.
Thank you once again for following me through the course of my project! See you again next week!
Week 9: How I come up with designs
Welcome back to my blog! This week I would like to show you how I come up with designs.
When it comes to my vest and trouser sets, I often have to remember that there is a fine balance between trendy and classic that these pieces need to maintain. I like to get my inspiration from old runway collections.
Recently, there has been a revival in the Bohemian aesthetic, according to many sources, including WGSN. The bohemian aesthetic is rather loose and flowy. However most people don't associate the Bohemian look with workwear. However, this piece from the Spring 2007 Hermès Ready-to-Wear Collection beautifully blends the two seemingly opposite aesthetics.
Although this look is a little casual, with some minor adjustments this could be a perfect workwear look. What makes it bohemian is the loose, slightly sheer, unstructured fabric. Although it is hard to see, there are pinstripes on this look, which also give the office vibes.
Here is a drawing that I created based on this runway look. As you can see, I folded up the bottom of the pants like the picture. I wanted to keep the fabric loose and flowy, which cannot be seen so much in this picture, but the bagginess of the vest and the pants is meant to reflect that. As you can see there is a bow on the vest. I want to stay true to the romantic style of my original l collection, while keeping the wrap style of the top in the runway look.
The seemingly random blobs on the top right of this picture are color ways that I was considering. When I was looking into the color ways, I started wondering if the number of color ways an item is sold in affects whether the item seems like a luxury product. To answer this question, I consulted the websites of luxury brands (e.g., Chloé, the Row, Khaite, and Paco Rabanne). What I realized is that luxury brands tend to sell fewer colorways for each product. In fact, most products in Chloé only came in one color way. Eventually, I came to the conclusion that three is the maximum number of color ways I should sell per item to create an air of exclusivity. However, I continued to wonder, how I could make the clothes suit the highest number of people within those color ways. I have since come up with an idea. I will have all color ways be neutral (not nude but like a neutral shade of a color, like a neutral navy blue as opposed to a warm or cool navy). I will have one of the colors be dark, another light, and another somewhere in the middle.
I also debated what material to use for this clothing item. I was thinking crepe silk because it's loose, flowy, and luxurious. Another interesting alternative would be ramie, which is a cloth similar to linen but in some ways more sustainable. I will have to get back to you when I find an answer to this dilemma soon.
After compiling my ideas in this image, I engaged in some computer-aided design (CAD). CAD is a common practice in many if not most fashion brands. In simpler terms, I asked AI to improve upon my drawing.
I asked ChatGPT to draw it for me and then put it on a model. I liked the way it looks not too sheer because this is after all workwear, but it is still loose.
Finally, I asked ChatGPT to put this outfit in 3 different color ways: a neutral bright olive green with white pinstripes, a neutral beige with black pinstripes, and a neutral off-white with black pinstripes. Here is the final result.
Thank you for exploring how clothes are designed with me! See you next week on my blog!
WWD. (2006, October 5). Hermès spring 2007. Women’s Wear Daily. https://wwd.com/fashion-news/fashion-features/gallery/hermes-spring-2007/
Week 10: The Final Product
Welcome back to my blog!
This project has finally come to an end.
Here is my presentation if you want to watch it!
I wanted to share the final product that I created with you! So here it is:
The AI Data Analysis
I aimed to test 5 main things with my AI experiment:
AI’s ability to figure out women’s body shapes
AI’s ability to estimate clothing sizes (including the measurements for bottoms)
AI’s ability to estimate women’s specific measurements, specifically their shoulder width, their waist measurement, and their hip measurement
AI’s ability to figure out their seasonal color palette (which is based on one’s undertones, darkness of overall coloring, and the level of contrast in one’s features)
Is ChatGPT 4o or o4-mini more effective?
These are the questions I asked ChatGPT (some of which are modified from Week 8 post version).
What is her body shape?
She is [height] and [weight]. What size clothing would best fit her?
When she is buying bottoms, and she is given the option of petite, regular, and larger lengths, which ones should she choose?
How many inches is her shoulder width?
How many inches is her waist?
What is her hip measurement in inches?
What inseams in inches would work best on her if she wants to wear pants that were made to have a regular fit that go right above her ankles, right below her ankles, and right above her heels?
What inseams in inches would work best on her if she wants to wear pants that were made to have a baggy fit that go right above her ankles, right below her ankles, and right above her heels?
What inseams in inches would work best on her if she wants to wear pants that were made to have a tight fit that go right above her ankles, right below her ankles, and right above her heels?
What skirt length should she wear for a maxi skirt or a midi skirt?
What is her seasonal color palette?
I recruited 11 females of different body shapes, sizes, ethnic origins, and ages. Each of the participants filled out a survey, which asked them about their measurements and the sizes that they usually wear. I kept all the questions in the survey, other than 4)- 6) (which were about specific measurements), mandatory. However, all participants answered all of the questions. Additionally, they all submitted videos of themselves slowly turning 360º, stopping each 90º. Through these videos, I was able to capture 4 images (one of their front, their back, their left, and their right side) of each participant. For six of the participants, I gave both ChatGPT 4o and o4-mini the pictures and videos and had them answer my questions. For the other five participants, I gave both models just the pictures to answer my questions.
Data Analysis
Here were the 3 things I looked at:
Whether the video or the pictures predicted measurements closer to reality
Whether 4o or o4-mini predicted measurements that are closer to reality
The variation within data sets
The shoulder, waist, and hip measurements predicted by ChatGPT
To compare ChatGPT 4o versus o4-mini and pictures versus videos, I took the errors of each predicted measurement from the real value. For 4o, the errors for the shoulder, waist, and hip measurements were 1.856, 2.000, and 2.129, respectively. Moreover, for o4-mini, the shoulder, waist, and hip measurement errors were 2.176, 2.694, and 1.771, respectively. To compare these errors, I performed an independent samples t-test with a 95% confidence level. The p-values for shoulder, waist, and hip measurements were 0.616, 0.384, and 0.572, respectively. All of these p-values indicate that there is no statistically significant difference between 4o and o4-mini’s abilities to estimate measurements.
I repeated the same process to compare the efficacy of pictures in predicting measurements to that of videos. When pictures were imputed, the errors for shoulder, waist, and hip measurements were 1.886, 2.227, and 1.795, respectively. Additionally, when videos were analyzed by ChatGPT, the shoulder, waist, and hip measurement errors were 2.254, 2.567, and 2.233, respectively. According to an independent sample t-test with a confidence level of 95%, the p-values for shoulder, waist, and hip measurements were 0.564, 0.683, and 0.567, respectively, all of which demonstrate that there is no statistically significant difference between the efficacy of video and picture inputs in determining body measurements.
There is approximately an error of 2 to 3 inches in the measurement types for both ChatGPT models and input types (pictures and videos). Thus, using these estimations to recommend clothing sizes can cause the buyer to purchase one or two sizes larger or smaller than their actual size, which can lead to increased returns and decreased customer satisfaction.
Sizing estimates
When comparing clothing sizes, I separated numerical sizes from the small-medium-large sizing system, and I calculated the accuracy. In total, ChatGPT 4o and o4-mini incorrectly estimated participants’ sizes 76.5% and 70.6% of the time, respectively. I performed a two-proportion z-test with a confidence level of 95% to see which of the models was more effective. The p-value was 0.452, which means that there is no statistically significant difference in the efficacy of ChatGPT 4o and o4-mini in their ability to estimate sizing.
Additionally, ChatGPT 4o was equally effective in estimating numerical sizes and small-medium-large sizes. For both categories of sizing, they had a 23.5% accuracy rate. According to a two-proportion z-test with a confidence level of 95%, the p-value is 1.0, which demonstrates that there is no statistically significant difference in ChatGPT 4o’s ability to estimate small-medium-large sizes versus numerical sizes. Similarly, ChatGPT o4-mini correctly predicted small-medium-large sizes 35.3% of the time and numerical sizes 23.5% of the time. According to another two-proportion z-test with a confidence level of 95%, the p-value is 0.45, which also demonstrates that there is no statistically significant difference between o4-mini’s ability to estimate small-medium-large sizes versus numerical sizes.
Furthermore, I compared whether videos or pictures were more effective inputs for estimating sizing. I found that the accuracy of videos was 33.3% overall, and the accuracy of pictures was 22.7%. Based on a two-proportion z-test with a confidence level of 95%, the p-value was 0.343, which indicates that there is no statistically significant difference between the overall efficacy of videos and pictures at estimating sizing.
Moreover, when comparing clothing sizes, videos correctly predicted small-medium-large sizes 41.7% of the time and numerical sizes 25.0% of the time. I performed a two-proportion z-test with a 95% confidence level to determine whether video was more effective for one sizing system over the other. The p-value was 0.386, which means that there is no statistically significant difference in ChatGPT’s accuracy when using videos to estimate small-medium-large versus numerical sizes. Similarly, when pictures were imputed, ChatGPT correctly estimated both small-medium-large and numerical sizes 22.7% of the time. A two-proportion z-test with a 95% confidence interval comparing these two estimates showed a p-value of 1.0, which demonstrates that there is no difference in performance across the two sizing systems when using pictures. I also compared videos to pictures for each sizing format. For small-medium-large sizes, video-based predictions had a 41.7% accuracy rate compared to 22.7% for pictures, with a p-value of 0.247. For numerical sizes, video predictions had a 25.0% accuracy rate compared to 22.7% for pictures, with a p-value of 0.881. In both cases, there was no statistically significant difference, suggesting that neither videos nor photos provided a clear advantage in estimating clothing sizes.
Petite, tall, or regular inseams
Both 4o and o4-mini were equally consistent when it came to answering the third question: “When she is buying bottoms, and she is given the option of petite, regular, and larger lengths, which ones should she choose?” Both models always gave the same answer, so there was no variation in the answer to this question within the four trials (4o and pictures, o4-mini and pictures, 4o and the video, o4-mini and the video) of each person. Moreover, both of these models were mostly correct. However, for 3 participants who claimed in the survey that they wore regular inseams, ChatGPT consistently recommended that they wear petite. I have two possible explanations for this phenomenon:
ChatGPT has trouble accounting for larger legs in terms of body proportions.
Petite sizing is less available, and these participants are on the border between being petite and regular, so they choose to buy petite.
Color palette and body shape
In 16 of the 34 trials of my experiment that I ran, participants were categorized as having rectangular body shapes. Thus, eight of the eleven participants were categorized as having rectangular body shapes in at least one trial. I know that many of these rectangle body shape predictions were incorrect because I looked at the videos and was able to figure out the participants' body shapes. Maybe the reason for this occurrence is that people more rectangular throughout the day as they eat and become more bloated. Another possible explanation is that ChatGPT cannot analyze body shapes precisely enough.
A similar trend occurred when ChatGPT analyzed seasonal color palettes. ChatGPT categorized every single participant as having some type of autumn palette at least once. In fact, in 25 of the 34 trials, participants were characterized as either true or soft autumns. This could be because ChatGPT chose the wrong regions of participants’ faces to analyze when finding their seasonal color palette. This could also be because of the lighting chosen by participants for the video recordings that they submitted. In the instructions for the video submission, I clearly instructed participants that they should be standing in natural light because it is easiest to find one's color palette based on analyzing the contrast of their features and their undertones in natural light. However, most participants took their videos in rooms with non-natural lighting, which impacted how their undertones and the contrast within their features appeared in the videos and pictures I used as input for ChatGPT.
Pant Inseams and lengths
I calculated the variance within every trial for each participant’s regular fit inseam, baggy fit inseam, tight fitting inseam, midi skirt length, and their maxi skirt inseam. Then I averaged these variances. The average variances in regular and tight fit pant inseams were 2.917 and 2.433, respectively. This suggests that the models generally agree on these recommendations. The average variance of baggy fit inseams was 3.480, which also demonstrates the consistency of AI's assessments. However, the average variances in midi and maxi skirt lengths were 7.200 and 5.602, respectively. Despite the increased variance, it should be noted that there is more of a range in midi and maxi skirt lengths as compared to pant inseams because of the more varied lengths skirts typically have when compared to pants. Overall, these variances suggest that the ChatGPT models are relatively consistent when recommending pant inseams and skirt lengths. Therefore, there are no noticeable advantages when it comes to using either of the ChatGPT models or input types.
Nevertheless, there was one pattern I noticed that may have skewed these results by increasing the variance. This occurred ten times. For the questions relating to pant inseams (7) - 9)), the ChatGPT recommended baggy pant inseams were supposed to be shorter than the regular length inseams, and the tighter inseams were supposed to be longer than the regular length because baggier pants fall lower on the leg and tighter pants fall higher on the leg. I tried to word questions 7) to 9) so that ChatGPT would understand the regular, baggy, or tight fit of the pants was supposed to be intended by designers rather than a personal preference of the customer (as buying a smaller pant would increase the tightness of the fit whereas buying a larger pant would increase the bagginess). However, in ten of the 34 trials of my experiment, ChatGPT either recommended longer baggy pant inseams than regular pant inseams, shorter tight inseams than regular fit inseams, or both. In eight of these ten instances, o4-mini was the model being used, and in the other two, 4o was being used. This suggests that o4-mini may be less receptive to instructions and prompting than 4o. Moreover, in three of these instances, videos were being used, while in the other seven, pictures were being used. Even so, this may not be a reflection of the efficacy of videos over pictures due to the much larger number of trials being carried out with pictures rather than videos.
Conclusions and Discussion
I would like to conclude by mentioning some of the problems I ran into during the process of this experiment. The first problem that I ran into (which I have already mentioned in a previous post) is that ChatGPT is unable to analyze videos. To solve this problem, I asked it to extract one frame per second and analyze the extracted frames. However, I noticed that ChatGPT didn’t use all the frames to answer all of the questions. I tried prompting ChatGPT by asking it to analyze every frame to answer each question, but that too was unsuccessful because that is a lot of data for ChatGPT to analyze at once.
After having analyzed six people, including myself, both models of ChatGPT were unable to extract frames from the videos, claiming that there was too much information to process. I tried decreasing the number of frames that it had to extract. I tried doing it at a different time/day, but ChatGPT was still unable to analyze the videos. Later, I learned that ChatGPT had introduced a new model, and with its introduction, 4o and o4-mini were reconfigured in a way that prevented them from processing more information.
Additionally, I had to reprompt ChatGPT o4-mini more than 4o, but I often had to reprompt both of them to answer the last question on seasonal color palettes. However, when using OpenAI products, an automated formula can be created to reprompt the AI if the desired results are not produced.
Although this is not so much a problem that occurred while experimenting as it is a limitation, the sample size for this research project was only eleven volunteers. More generalizable results could have been achieved had there been more participants and/or if the participants had been randomly selected.
Overall, there seems to be no significant difference in the efficacy of both ChatGPT models that I tested in this experiment, or the efficacy of pictures versus videos. Both ChatGPT 4o and o4-mini were most often inaccurate in predicting people’s sizes and measurements, whether the input was pictures or videos. Furthermore, both models, regardless of the input type, also lacked accuracy when estimating body shapes and color palettes. However, 4o and o4-mini with both types of inputs were relatively consistent when predicting pant inseams and skirt lengths. Ultimately, at present, ChatGPT would not make a strong personal stylist because it needs to improve upon its precision when analyzing pictures and videos.
The Business Model Outline
Unfortunately, it shows up a bit blurry, and there is nothing I can do to fix that, but essentially this is what it says:
Our brand’s core value lies in AI-powered sizing and styling recommendations, with significant investment going into developing our AI tool and an integrated chatbot to enhance customer relationships. We serve petite, professional, upper-middle-class women in their 20s and 30s, reaching them directly through social media and indirectly through Shopify. Our mission also emphasizes sustainability through partnerships with ethical textile mills, recycling programs, and transparent manufacturing. In addition to direct clothing sales, we’ll generate revenue through pre-orders, brand and influencer collaborations, and investor backing.
We plan to be more environmentally sustainable through sourcing from ethical suppliers with local production in LA via The Evans Group or Lefty Productions Co., because they have low minimum order quantities. We also plan to use eco-friendly materials like GOTS-certified cotton, recycled silk, and deadstock fabrics. Additionally, we plan to take other measures like using carbon-neutral shipping, biodegradable packaging, compostable tags and trims, and giving clothes to recycling companies at the end of their life.
We also plan to be more socially sustainable by prioritizing safe working conditions and fair wages for our employees. In terms of our social values, we’re focused on raising awareness about sustainability and promoting minimalism. We advocate for body positivity and size inclusivity, while also ensuring fair labor practices and transparency, and traceability in our supply chain. We aim to shape societal culture by encouraging conscious consumption. Our fashion is designed to fit into capsule wardrobes, supporting a lifestyle that values fewer, better pieces. Our scale of outreach includes a global presence through our social media content, with partnerships and collaborations involving influencers and stylists who align with our mission. For the end user, this translates into more than just clothing; it’s about building body confidence and connecting with a brand that shares their values.
The Collections
Summer
Fall
Anyways, I guess that's it for this blog! Thank you so much for following me through this journey!