Imagine, if artificial intelligence (AI) could read what is currently happening in the human brain, would you dare to believe it? A study published online this month in the international scientific journal Neural Networks shows that Japanese scientists have successfully created the world's first psychological image of objects and scenery based on human brain activity using AI technology. This technology, known as "brain decoding," can visualize perceptual content based on brain activity and is expected to be applied in the fields of healthcare and welfare. In today's rapidly developing world of brain computer interfaces and neural engineering, analyzing brain activity and using machines to read out the thoughts or images of volunteers is no longer a science fiction story. The technology of decoding ideas is becoming increasingly mature, and the idea of AI mind reading is becoming closer to reality. At the same time, this has also raised concerns about privacy among people. Previous studies on replicating psychological imagery have shown that brain activity measured by functional magnetic resonance imaging (fMRI) can reconstruct images seen by humans. But this is limited to specific content, such as letters, faces, etc. Based on previous methods, a team of scientists from the Japan Institute of Quantum Science and Technology (QST) and Osaka University have developed a technique that quantifies brain activity and combines generative AI and prediction techniques to create images for reconstructing complex objects. Researchers presented participants with approximately 1200 images of objects and scenery, and used fMRI to analyze and quantify the relationship between their brain signals and visual stimuli. These images were then used to train generative AI to decipher and replicate psychological images from brain activity. As a result, researchers used this method to generate vivid images, such as leopards with distinguishable features such as ears, mouth, and spots, as well as objects such as airplanes with red wing lights. Researchers say that visualizing psychological images into any natural image is an important milestone. QST researchers claim that humans have used microscopes and other devices to observe the world that is invisible to the naked eye, but are still unable to see one person's inner world. This is the first time humans have peered into the thoughts of another person's mind. Essentially, it can be used to help create new communication devices, while also allowing scientists to explore and understand how hallucinations and dreams work in the brain. A new study on decoding language data was published earlier this year in the journal Nature Neuroscience. The research team at the University of Texas at Austin in the United States has developed a modern "mind reader" with the help of a large language model. They used fMRI to collect brain activity information from three participants and converted their thinking into language with an accuracy rate of 82%. The blood oxygen level dependent signals obtained from fMRI scans display changes in blood flow and oxygenation levels in different parts of the brain. By focusing on the activity patterns of brain regions and networks involved in language processing, researchers have found that their decoders can be trained to reconstruct continuous language, including the general meanings of specific words and sentences. Specifically, the decoder captured the brain responses of three participants while listening to the story and generated word sequences that may have produced these brain responses. These word sequences capture the main idea of the story very well, and even