The incursion of artificial intelligence throughout the Internet continues to cause unexpected effects and different reactions by users. Although a large part of the people who are not interested in the matter simply ignore everything related to AI, some users have taken the negative effects of this current very to chest.
In recent weeks, there have been several stories that we have been able to know about different web creators who have designed strategies and resources to catch, capture or block the AI ​​or Crawlers bots, as they are known, that various companies use to sweep all Internet collecting data to train its AI models.
There are many bloggers, youtubers, content creators or other network users who not only want to prevent these companies from using their creations or information to train AI models, but want to take revenge for being disgusting with the new trend that seems to be seizing the Internet.
An example of this is that of Aaron B, a programmer who has created a website with a special code, called Nepenthes, designed to catch, so to speak, to these scraping bots in an infinite cycle that they cannot escape. The objective is to leave them useless, make them spend fooling resources to harm the companies behind, or even contaminate the AI ​​when providing it with meaningless data that will worsen their training.
Bots traps
This programmer, in conversation with 404 Media, explains that he decided to create this tool as a form of protest, in the face of a political situation that worries him and that he believes that the Internet is worsening as we know him.
What it does Nepenthes It is to generate non -stop links on the same website, which makes an AI not to enter them even if they only redirect themselves, without providing any valuable information. They are only random words or numbers, with which to create a kind of labyrinth for robots.
In networks, many users have celebrated this initiative, although many comment that, for practical purposes, it is probably not so useful. Many suggest that this kind of virus for AI will not affect the most advanced bots, capable of detecting when they have been trapped in an infinite loop. In addition, they blame that it is counterproductive for the increase in the energy expenditure that this causes, and therefore, the damage to the environment. Despite this, Aarob B presumes that some bots can be imprisoned for months.
On its website, Aaron warns: «This is a deliberately malicious code that aims to cause harmful activities. I don’t implement it if you don’t feel completely comfortable with what you are doing.
Intoxicating the AI
Another example is that of the F4MI YouTuber, creator of content that, to take revenge on other users who use AI to steal the scripts and ideas of their videos and make cheap replicas quickly, has uploaded a video to YouTube with totally surreal subtitles, To use any attempt to theft.
This youtuber, which makes videos about nostalgic technology, has uploaded a video with a file of subtitles in “.ass” format full of absurd text to complicate the work to AI, although any human can correctly see the video since the subtitles are Transparent, that is, they are not seen. Only AI can detect them.
Even so, it is known that advanced bots such as that of Chatgpt O1 is able to focus more on the audio of the video, and thus achieve a subtitled version with meaning, ignoring the chaotic text. Therefore, we see an open war between Internet activists and artificial intelligences, which will have to become more advanced to overcome these tests with which they could be increasingly usually.
