Indonesian Political, Business & Finance News

80s Photo Trend Using ChatGPT Goes Viral, But It Is Making the Earth Warmer

| Source: CNBC Indonesia Translated from Indonesian | Technology
80s Photo Trend Using ChatGPT Goes Viral, But It Is Making the Earth Warmer
Image: CNBC Indonesia

Social media is currently abuzz with 1980s-themed photos created using artificial intelligence (AI). Users simply need to upload a photo and enter a prompt into ChatGPT to achieve a retro look, ranging from hairstyles and clothing to the characteristic colour tones of that era.

However, behind this seemingly simple process lies a substantial demand for resources. Every time a user requests AI to generate an image, the computational process actually takes place in data centres filled with servers, requiring vast amounts of electricity to operate.

These data centres also require cooling systems, water, and various hardware to continue functioning. Consequently, creating images using AI leaves a physical and environmental footprint, including energy consumption and emissions that contribute to global warming.

Generating AI Photos Requires Energy

According to reports via FirstPost, when users request ChatGPT or other AI services to create images, the process does not occur directly on the user’s phone or device. Instead, the request is sent to servers located in data centres.

There, specialised processors run AI models to understand user commands, process the image, and then generate a new visual. The more complex the task assigned, the greater the computational requirement.

Research from Carnegie Mellon University and Hugging Face found that the energy consumption for generating AI images can vary significantly depending on the model used. In their testing, some models required far more energy than others.

Studies of various other AI image-generation models also found wide discrepancies in energy consumption. The model used, image resolution, and generation method can all influence the amount of energy required.

Therefore, there is no single figure that can conclude every AI-generated photo requires the same amount of electricity. However, the issue becomes more significant when such activities are performed on a massive scale.

A single user, for instance, might not just generate one image. Viral trends encourage people to try various prompts, create multiple variations, change styles, and increase resolution, repeating the process until the desired result is achieved.

If this habit is adopted by millions of users, computational needs swell accordingly. A photo that appears simple on a screen can be part of millions of simultaneous computational processes occurring across various data centres.

Furthermore, generative AI is no longer used solely for generating photos. The technology is also being utilised to create illustrations, animations, videos, and various types of content with increasing levels of complexity.

AI Also Requires Water

In addition to electricity, AI operations require another resource that is often invisible to users: water. Servers in data centres generate massive amounts of heat during operation, necessitating cooling systems to prevent hardware overheating.

Water is one of the resources used in data centre cooling systems. Research from the University of California, Riverside, regarding the water footprint of AI usage shows that water consumption can be a significant part of the environmental impact of artificial intelligence systems.

However, the amount of water used depends heavily on the data centre’s location, cooling technology, and the energy source used. Therefore, the claim that one AI photo definitely consumes a specific amount of water cannot be applied universally.

Data centres with different cooling technologies, locations, or power sources will produce different water and carbon footprints. The environmental impact of AI is not limited to the electricity consumed when an image is generated. Data centre infrastructure requires GPUs, servers, chips, storage systems, and networking equipment, all of which must be produced beforehand.

This means there is a long supply chain before a prompt transforms into a photo on a user’s screen. This process includes raw material extraction, the production of semiconductors and hardware, data centre construction, transportation, and the management of devices at the end of their lifespan.

In other words, the term ‘cloud computing’ does not mean that computation actually occurs ‘in the clouds’. AI still operates through data centre buildings filled with machines that require energy to function.

The growth of generative AI is making data centre energy consumption a concern in many countries. The International Energy Agency (IEA) estimates that global data centre electricity consumption will rise sharply by 2030, with AI being one of the key drivers of this growth in computational demand.

Therefore, the real issue is not limited to the trend of 1980s-style photos. AI retro photos are merely one example of a major shift in computational usage. Today, users are busy creating photos, while in the future, AI usage may shift further towards video, 3D content, and simulations that require even more complex computation.

So, should users stop creating photos with AI? The issue is not that simple. Creating one or a few images using AI certainly cannot be identified as the primary cause of environmental damage.

The larger impact arises from the accumulation of usage on a massive scale and how the AI industry manages its resource requirements. More efficient AI models can reduce computational needs, while data centres powered by low-carbon energy can help reduce emissions.

Furthermore, water-efficient cooling technologies and the use of recycled water can reduce the pressure on water resources.

Thus, the responsibility for reducing the environmental impact of AI lies in…

More from 12 September 2026 →

Cite this article
Source: jawawa.id (https://jawawa.id/newsitem/80s-photo-trend-using-chatgpt-goes-viral-but-it-is-making-the-earth-warmer-1789199710)
Originally published by CNBC Indonesia (https://www.cnbcindonesia.com/tech/20260912103759-37-767387/heboh-tren-foto-80-an-pakai-chatgpt-ternyata-bikin-bumi-makin-panas), translated from Indonesian
Print