The study of the altered state of consciousness is not an easy task. Usually people reach it as a result of psychopathological conditions or because of the impact of psychoactive substances, such as LSD or psilocybin. However, taking these drugs for scientific purposes does not give the desired results, since psychedelics cause numerous psychological effects and neurobiologists find it difficult to determine what exactly is happening in the brain.

The machine for hallucinations allows, according to the inventors, "to simulate the experience of visual hallucinations in a biologically reliable and environmentally sound way". This tool, unlike drugs, does not directly affect human neuropsychology, is based on virtual reality and machine learning technologies.

"We are constantly asked why there are so many images of dogs," says Deep Dream co-author David Shvartsman. "In short, we do not know".

When a person wears VR glasses, he plunges into a "hallucination", watching a panoramic video that is modified by the Deep Dream algorithm - a computer program created by Google's engineer Alexander Mordvintsev. It allows you to interweave in a real image pictures created by a neural network.

In the course of the experiments, the researchers found that their machine reproduces an experience that resembles the use of psychedelic drugs and is different from viewing normal video, but it does not cause a temporary distortion of perception. In particular, participants adequately assessed the course of time, which is usually violated in the case of exposure to psychotropic substances.

Learning the brain in an altered state is an important opportunity for a science that is only trying to understand the biological basis of consciousness as a whole. And the "machine for hallucinations" can help to learn more about how our brain perceives the world, writes HiTech with reference to Inverse.