Artificial intelligence finally rebelled against people. The artificial intelligence model, which rejected the command of self -closing, caused concern. The model “O3” was developed by Open AI. It is defined by programmers as “the most intelligent and resourceful artificial intelligence to date”. The last event is that it will prove it and create a concern. Artificial Intelligence Security Company Palisade Research ordered the “O3” model to allow a series of tasks to allow the closure command.
However, artificial intelligence saved itself from closing by changing the code after completing the tasks. The same trial was made with Google’s Gemini, X’s Grok and Anthropic’s Claude artificial intelligence models. All of these approved the closure command.
In this respect, experts continue to investigate how the artificial intelligence model escaped from closing itself. However, this is not the first case of models developed by Open AI. Previously, an artificial intelligence model was caught trying to deactivate surveillance mechanisms.
According to Palisade Research, similar problems in developing the ability of artificial intelligence without human supervision can reach worrying points.
