The Hidden Cost Behind Progress: On AI, Creativity, and Our Environment
Original shoot by FAST at UCLA and RefineLA. Photography by Yuuki Hayashi. Editing and Styling by Lianne Calvo. Modeling by Bella Hawthorne and Lianne Calvo.
This Earth Day, RefineLA and FAST at UCLA are teaming up to honor our appreciation for the vast beauty nature has to offer. Though as college students, we spend most of our time attached to our devices, staying connected to our environments remains a priority for both organizations. Together, we envision a future where fashion and conservation go hand in hand.
There are still systems in place that make it difficult for a circular model to thrive. This is especially true since the advent of AI, which has become embedded into every major digital platform. ChatGPT is one of the most widely used large language models (LLMs), with 400 million weekly active users. Though AI is not inherently wrong, its excessive use has dire consequences for our planet’s health.
Our hope is to redirect attention to what is lost when we rely exclusively on a system that cuts out human creativity, compassion, and in the worst case scenario, critical thinking. As student organizations whose missions center around the power of creativity we implore you to weigh the implications of AI against its environmental costs.
The words “decomposition” and “decay” reflexively evoke negative images of rot, waste, and death. But in reality, they are vital processes in nature. Decomposition initiates the breakdown of organic matter, which returns key nutrients to the earth and makes way for a new generation of organisms to thrive. This is what a balanced ecosystem looks like: a closed circuit of birth, life, and death that results in a biodiverse habitat.
Imagine a world where decomposition did not occur: the planet would be littered with waste with no way of regrowth. This dystopian world is eerily reflective of what many developing countries are already living through, their plight shielded from our eyes by brands that intentionally discard their waste overseas.
Each year, the fashion industry lines developing countries with 92 million tons of fabric waste that end up in flames, returning to the Earth as carbon emissions and accelerating climate change. The Kantamanto markets in Accra, Ghana serve as a harrowing example of brands shipping their problems to other parts of the world.
Many of the clothes donated to op shops end up in Kantamanto market, in the Ghanaian capital Accra. For the past few decades, the resale of Western cast-offs has boomed here. They are so cheap, local textile makers can't compete.
Around 15 million garments arrive in Accra every week from the UK, Europe, North America, and Australia. 40% are in such poor condition that they go directly to landfill, cutting into resellers’ profits and overflowing streets with remnants of unwanted fashion.
A common practice to curb this waste is setting landfills aflame, but this only makes the quality of life worse for the 100,000 locals who must grapple with the risk of disease breeding at every corner. Meanwhile, fast fashion brands absolve themselves of any responsibility for this global issue, a rather disturbing manifestation of “out of sight, out of mind.”
The textiles that wash back onto the beach become so tangled in the sand they are almost impossible to dig out. (Foreign Correspondent: Andrew Greaves)
There are three prices we pay for a single item of clothing: the cost of production, the retail cost, and the environmental cost of wasteful destruction, not to mention the millions of people who bear the burden of our overconsumption. These are truths we as consumers cannot ignore, even as we reap the benefits of the end product. AI is one such product we have many reasons to worry about.
The success of ChatGPT has illustrated the increasing profitability of large language models (LLMs). The start-up’s monthly revenue increased by 1700% in just under two years, and is currently valued at 300 billion dollars. Corporations have been racing to enter the market – tech giants like Meta, Amazon, Google, Microsoft, and Nvidia have all since introduced their own LLM competitor.
Videos using AI-generated content have taken the internet by storm. Videos (left to right) courtesy of @redditstoryteller315, @meowmeowarts, and @0nlytxsh on TikTok.
Silicon Valley has opened the floodgates for how AI can be worked into every aspect of our lives. These personal assistants are used as career coaches, receptionists, relationship gurus, and homework machines. AI has even been fueling youth pop culture – when scrolling through short-form video platforms like Instagram Reels or TikTok, you’re more than likely to find videos made using AI. Platforms like raena.ai have been used by Gen Z to “brainrot-ify” their homework assignments. The app condenses text files into narrated video, complete with a backdrop of video gameplay and internet slang.
Although these creative applications of AI may seem like harmless fun, the harsh reality is that these videos destroy our environment.
The internet is powered by data centers, or warehouses dedicated to housing computing infrastructure. Although data centers have been around since the 1940s, the advent of LLMs is causing them to expand at unsustainable rates. Put simply, AI increases demand for processing power: a single query from generative AI consumes about five times more electricity than a simple web search.
Training a LLM exerts an energy burden on these facilities. Scientists from Google estimate that training Chat GPT-3 consumed enough electricity to power 120 American homes for a year, generating 552 tons of carbon dioxide. What’s worse is that companies release new versions of their LLMs every few weeks, creating an endless cycle of consuming increasing amounts of energy to train the next best model.
Data centers also create an unprecedented demand for freshwater as an integral part of cooling systems that prevent servers from overheating. Data centers are currently in the top ten most water-consuming industries in the U.S, and AI will only cause centers to climb in the ranks. Global water consumption from AI use is projected to reach between 4.2 and 6.6 billion cubic meters by 2027, which is 4 to 6 times the annual water usage of Denmark.
While the expansion of AI – and thus, data centers – is accepted as necessary for advancing revolutionary technology, this innovation comes at the expense of the global south. Tech companies have been strategically placing their data centers in Latin America and sub-saharan Africa, exacerbating freshwater scarcity and impacting 720 million people worldwide. As these companies profit off the cheaper real estate and energy prices overseas, they turn a blind eye from the strain they put on local water resources. Despite these massive levels of waste AI’s seamless integration into everyday life encourages users to dismiss its environmental cost.
Beyond environmental destruction, the convenience of AI reinforces the notion that creativity can be engineered, not cultivated. For example, the topic of AI was a major sticking point in negotiations throughout the 2023 SAG-AFTRA strike. Although the union was able to instate protections regarding the use of AI, SAG-AFTRA workers still perceive the technology as a potent threat to their livelihoods.
AI has even made its way to the top ranks of entertainment: the film Emilia Pérez won Best Original Score at the 2025 Oscars, despite using AI to enhance lead actress Karla Sofía Gascón’s voice. Although the process did not involve the use of generative AI, critics remain concerned about the technology’s potential to replace creatives.
However, it’s not just writers that are at risk of being replaced. Fashion giant H&M recently announced that it will be developing 30 AI-powered “digital twins” of its model based on existing works with no guarantees of crediting original contributors. Sara Ziff, a model and labor activist, stated to CNN she has “serious concerns about the use of digital replicas without meaningful protections in place.” H&M’s move also sparks worries about consent and compensation for all fashion workers involved in a shoot, including hair and make-up artists, stylists, and production assistants. As more fashion brands hop on this trend, these technologies could render human creatives disposable once their likeness is effectively captured.
The tension between creatives and AI reflects a larger anxiety about how AI will be regulated to ensure the artistic integrity of freelancers and brands alike. We’ve already seen major mishaps from brands using AI-generated art without human oversight, such as the most recent AI Barbie trend. Meanwhile, smaller creators worry that the lack of regulations surrounding training data used for LLMs will allow large corporations to profit from their work without proper credit.
Tighe Flatley on Linkedin comments on LaCroix’s use of AI to generate social media posts that unintentionally reinforce ethnic stereotypes.
Others suppose that AI will render us creatively incompetent or cut out the need for creatives entirely. While this does underscore an urgent need for copyright protections, this nihilistic stance lacks faith in the artists who continue to find creative inspiration that de-centers artificial production. These artists remind us to hold onto the things that lead to true ingenuity: our imagination, and most importantly, our values.
Ella Mae is one brand that forwards this counter-narrative: in their S/S 2025 collection “Rock Collecting”, which features themes of nature’s elements. As the brand puts it, their collection is “like holding the world in your palm.” The runway is a vivid rendering of a journey through a mossy forest, with models carefully making their way through arched branches and across a stone path. Rather than a sterile showcase, the space invites the audience to morph themselves around nature, a gentle reminder that nature is its most beautiful when we allow it to unfold without human interference.
Ella Mae, S/S 25.
More established fashion brands have rolled out similar runways in the past, with Jacquemus’s S/S 2020 show staging models in sprawling fields of lavender in the South of France. Collina Strada’s spring 2025 “Touch the Grass” runway featured models playfully interacting with dirt and grass during their catwalks, asserting the idea that a digital detox is a necessity to achieving balance in today’s world.
All of these shows point to the idea that any imitation of nature will fall short; natural colors, sensations, and textures achieve a vibrance that manufactured mass-produced items cannot attain. Still, the execution of some runways leaves much to be desired, with brands using nature as aesthetic positioning rather than taking tangible actions to protect environmental spaces, either through donations or internal commitments to sustainable production.
Hosting a show in a natural setting barely scratches the surface of true environmental activism. The real need for change lies in the systems of production, and this is what we must continue to question. In what ways are we returning the wealth of beauty and inspiration we get from nature back into the land?
Fast fashion thrives on rendering its customers blind to environmental realities that we simply cannot accept. It doesn’t need to take a village to champion the environment. Advocacy can start with individual actions, self-education, and the courage to reckon with the outcomes of overconsumption.
