Artists have started poisoning artificial intelligence!

Artists are using software developed to protect their work in an attempt to neutralize artificial intelligence and safeguard their creations. What began as general dissatisfaction with generative AI has now turned into a full-scale offensive by artists.

WHY IS THERE AN ATTEMPT TO BLIND ARTIFICIAL INTELLIGENCE?

The primary reason for this hostility is that artificial intelligence is trained on original images from platforms without the artists' permission. Specifically, some new products being developed aim to confuse AI so that it makes mistakes. In other words, they are being released to "poison" the "artwork" data used by AI.

For example, the free software called Nightshade was developed by researchers at the University of Chicago and has been widely adopted by artists. This software works by using AI against AI; in other words, another AI is being used to neutralize artificial intelligence.

Platforms like Midjourney AI operate by scraping data from the web, including commercial art. According to reports on Midjourney, this platform was trained on more than 100 million images from original creators without their permission.

This AI software works by using tags to make other AI programs misidentify objects in images, subtly altering the image at the pixel level to make other AI programs see something different from what is actually there. This means that the AI may produce results that do not match user instructions.

News Source: 12punto