The viral eighties trend and wasting water!

It feels as though the 1980s have suddenly descended upon the Gen-Z era. Scrolling through social media, it looks like people climbed into a time machine and gone back to the ’80s? This time machine is nothing but AI, or Artificial Intelligence. With the help of AI, people are returning to their old memories. And those memories themselves are anything but artificial.

Looking back several decades can be an exploration of memories for some, and a discovery for others. After all, many of the people caught up in this AI-generated ’80s photo trend were not even born during that era.
In this trend, the fashion, looks, and nostalgia of the 1980s have all blended together to transform our Facebook timelines into a wonderful canvas. Rising above the various political, social, familial, and personal frustrations of everyday life, society has embraced a shared sense of joy, with memories serving as the common thread.

But why did people want to return to the 1980s in the first place?

Having overcome the despair of World War II, the West entered the “vibrant sixties,” when people were swept up in a hopeful and exhilarating romanticism. But by the 1970s, the final nails were being driven into the coffin of colonial power in different parts of the world. The Soviet era, too, was nearing its collapse. Meanwhile, amid the Cold War, American capitalism and the US economy were reaching new heights.

Against this backdrop, the changes that emerged in people’s thoughts, attitudes, culture, and lifestyles during the 1980s can be seen as an extension of the colourful, energetic spirit of the vibrant sixties.

With the rise of MTV, music became not just something to listen to but something to experience visually as well. Michael Jackson, Madonna, and Queen appeared as dazzling new sensations. In Hollywood, Star Wars embarked on its journey, while Steven Spielberg created pop-culture classics. Alongside metal, rock, and pop culture, the craze surrounding Princess Diana added another layer of fascination.

All in all, the 1980s came to symbolise a unique blend of economic, social, and cultural transformation. Its influence was reflected in people’s lifestyles: sunglasses, denim jeans, wide-buckled belts, jackets, stylish blouses, large earrings, bangles, long or curly hair, and eye-catching hairstyles. Motorcycles and automobiles were part of the picture too.
In this way, the decade ushered in a new wave in the world of fashion. In Bangladesh, however, the impact of this wave began to be felt after the mass uprising of the 1990s. That is why nostalgia for the ’90s tends to be stronger here.

The emotions surrounding the 1980s in the West have been unleashed by influencers there. What began on Instagram moved to TikTok and has now swept across Facebook. And when a trend starts on Facebook, can Bengalis really resist jumping on the bandwagon?

People are using AI to give not only their individual identities but also their romantic relationships, families, and friendships a nostalgic ’80s makeover. With AI, everyone is recreating themselves in the 1980s in whatever way they can, presenting an idealized version of themselves.

New trends come and go on social media, but this fascination with the 1980s has sparked a particular kind of curiosity. People are even looking into whether there is some deeper mystery behind this phenomenon.
Why Did People Become Obsessed With This Trend?

It is often said that Gen Z and millennials have a particular fascination with the pop culture, voluminous hairstyles, denim, and retro vibes of the 1980s and ’90s. This trend has tapped into that fascination. At the same time, the nostalgia of people who lived through, or were born in, the ’80s and ’90s has made the trend even more powerful.

Some also believe that influential figures such as Kim Kardashian have been used by multibillion-dollar fashion brands to kick-start this ’80s-inspired trend.

Since the COVID-19 pandemic, financial pressure and inflation have increased for various reasons, including successive wars and the global energy crisis. As a result, people in the Western world have reduced their shopping compared with before. Brands have turned to the 1980s as a way to attract consumers by introducing new trends into the fashion world. I do not know how true this theory is, but it would not be unusual for commercial motives to lie behind such trends in the Western world.

Among AI researchers on Instagram, there has also been discussion about how AI companies may be strategically introducing various trends, including ’80s-inspired looks, to build enormous databases containing the personal information, behavior, preferences, dislikes, and reactions of hundreds of millions of people.

Such databases could potentially be sold to governments or other organisations, and the information could even be used in politics or warfare. By leveraging this data, any company could gain the ability to control and influence people’s needs and preferences, both at an individual level and on a global scale.

It is also true that retro styles and fashions repeatedly make a comeback. At different times, people choose or celebrate different periods of the past. In that sense, this nostalgia for the 1980s can be seen as part of an ongoing cycle. The involvement of AI, however, has created an entirely new level of excitement.

Photographs from the 1980s were far from perfect. Photoshop did not exist back then. Images often had a slightly cracked, grainy quality, with a yellowish or almost rust-like tint. That very imperfection has created an aesthetic appeal in today’s highly polished digital world.

Recreating the 1980s requires nothing from that era itself. With just a photograph and a prompt, anyone can create an attractive ’80s-style image. But jumping on the trend has also created a kind of competition to generate unnecessary images. As a result, the environmental concerns surrounding AI use have surfaced once again.

There is talk about how the AI-powered frenzy of recreating the 1980s through images may be consuming an enormous amount of additional water. But how exactly does creating images with AI use water? For many people, this is still an unfamiliar issue.

From the very beginning, there has been widespread discussion about the major shock AI could bring to job markets around the world, as AI has increasingly begun to influence, and in some cases already transformed, people’s lives both directly and indirectly. Now, concerns about environmental damage have been added to the conversation.

The question being raised is this: when the planet is already struggling to meet the thirst of 8 billion people, can we really afford to satisfy the “thirst” of countless AI systems, including ChatGPT and Gemini? The enormous amount of water required to cool AI servers and data centers is far greater than most of us might imagine.

A recent special environmental research report on this issue has been published by the United Nations University’s Institute for Water, Environment and Health.

The study says that by 2030, the annual water consumption of AI data centers could reach approximately 9.3 trillion liters (930 billion liters). That amount of water would be enough to meet the minimum annual water needs of around 1.3 billion people in sub-Saharan Africa.

And it is not just water. By 2030, AI data centers could consume a significant share of the electricity generated worldwide. Data centers are projected to use 945 terawatt-hours of electricity each year, roughly equivalent to the combined annual electricity demand of many countries.

The same reality applies to the data centers operated by technology giants such as Google, Microsoft, and Facebook. In 2023 alone, one technology company’s own data centers directly withdrew 29 billion liters of potable water to cool their processors. More than 23 billion liters of that water was lost through evaporation, with approximately 80 percent of it being drinking water. Since the emergence of AI, the need to train AI systems and support their use has made concerns over water consumption even more alarming.

It appears that a popular AI company receives and processes 2.5 billion prompts a day. This number is expected to continue rising. By 2027, meeting global demand for AI could require the direct and indirect withdrawal of 4.2 to 6.6 billion cubic meters of water, equivalent to the annual water consumption of four to six Denmarks, or roughly half that of the entire United Kingdom. OpenAI’s GPT-3 language model alone required approximately 5.4 million liters of water to train just once.

These figures were highlighted in a study by researchers at the University of California, Riverside, titled “Making AI Less Thirsty,” which examines ways to reduce AI’s water consumption.

The study further found that generating a simple image with an AI system, or exchanging prompts of 100 to 500 words, consumes roughly 0.5 liters (about two glasses) to 1 liter of water, directly and indirectly, for processor cooling. Generating a high-quality 10-second AI video puts considerable strain on servers. On average, around 20 glasses of potable water are invisibly consumed or evaporated to cool the servers or generate the energy required for the processing.

That means that simply going along with the trend, or even creating an ordinary image, can consume the equivalent of two glasses of water. Environmentalists are bound to be concerned about this. The issue has come into the spotlight once again because of the trend of returning to the 1980s, as well as the earlier trend of using Nano Banana AI to create images of women wearing sarees, and the excessive enthusiasm surrounding such trends.

I have to admit, I too had to turn to AI several times while writing this piece. Now I am wondering: how many glasses of water did that end up consuming?

* Rafsan Galib is Editorial Assistant, Prothom Alo. He can be reached at
rafsangalib1990@gmail.com

* The views expressed here are those of the author.