Top News

These days, the trend of retro-style AI pictures of the 1980s is becoming increasingly popular on social media. Users are uploading their ordinary selfies to AI tools and turning them into portraits with big hairstyles, vintage clothes, old studio backgrounds and filmy textures.
This picture, which is prepared in a few seconds, may look completely digital, but a huge computing infrastructure works behind it. Data centers running AI models consume electricity, generate heat, and require many resources, including water, to keep them cool.

How much electricity does it cost to create an AI photo?
There is no fixed figure for the electricity spent in making images through AI. This can vary significantly depending on different models, hardware, resolution and capacity of the data center.

According to estimates by some experts, the power consumption for creating an AI image can be around one watt-hour, while in different systems it can be much less or even reach several watt-hours.
That means it is not easy to make an accurate estimate of energy consumption just by looking at a picture. Both the size of the AI ​​model and the hardware it runs on determine its impact.

The real challenge is not one photo, but millions of images
The energy consumption of a single AI image does not seem very large. Problems arise when millions of people use such tools simultaneously.
Often the user is not satisfied with the first picture created. Then many new versions are made to change the face, improve the clothes, change the background or correct the lighting. This means that one viral trend can turn into millions or billions of AI inference tasks.
This is why when understanding the environmental impact of AI, it is more important to look at the full scale of use rather than just the cost of a single image.
Water is also spent in creating AI images
The environmental impact of AI is not limited to just electricity. Servers in large data centers work continuously and generate a lot of heat. Water may be required in the cooling system used to control this heat.
However, the exact water consumption figure for an AI image may not be the same everywhere. This can vary significantly due to the location of the data center, weather, cooling system, power generation methods and efficiency of the infrastructure.
Some estimates put the water footprint for an AI image at around a few dozen millilitres. This amount may seem small for a single image, but when millions of images are created daily, the total consumption can increase significantly.
Increasing use of AI is increasing the burden on data centers
With the increasing popularity of AI tools, the demand for data center computing capacity is also increasing. Running text, image, video and other generative AI systems requires a large number of powerful chips and servers.
This is not a matter limited to just AI photo trend. The billions of AI inference tasks performed every day put a huge burden on the data center's power and cooling infrastructure.
This is why AI companies and data center operators are working on more energy-efficient chips, better models and new cooling technologies.
Your selfie data is also an important question
The retro photo trend is not just about environmental concerns. Here user privacy also becomes an important issue.
People upload their clear and high-quality selfies to create AI images. Such photographs may contain facial structure, skin details and other information to identity.
Therefore, before using any AI photo tool, it is important to read its Privacy Policy and Terms of Service. In particular, it should be seen how long the uploaded image is kept, whether it can be used for model training and whether the user has the option to delete it.
Should AI give up making photos?
There is no need to worry about making just one or two AI pictures. The environmental impact of any one image may be limited in itself. The real challenge comes when billions of AI images, videos, and other content are constantly being created at scale.
The solution may not be to stop using AI completely, but to make it more efficient. Smaller and more efficient models, better hardware, low-energy data center technology and cooling systems that reduce water consumption can help reduce this impact.
The AI ​​retro trend of the 1980s is therefore not just a story of social media fun. This is also an example of how huge physical infrastructure works behind the digital facility available in a few seconds. Therefore, with the increasing use of AI, it is important to pay attention to all three – energy, water and data privacy.