When AI Maps the Brain, Awareness Maps Itself
When AI Maps the Brain, Awareness Maps Itself The news landed quietly, but its implications may echo for decades: an artificial intelligence, built on the same architecture as ChatGPT, has identified 1,300 previously unknown regions of the brain . What was once divided into 52 areas has suddenly become a city of neighborhoods—each with its own molecular accent, its own rhythm of thought. The system behind the finding, called Cell Transformer , was trained not on language, but on gene-expression and spatial data . In place of words and sentences, it read neurons and their chemical conversations. The result was a high-resolution atlas of the mind, drawn by a machine that doesn’t yet have one. The Grammar of Life The transformer design—originally meant for text—learns by attending to relationships: how one token depends on the others around it. When that logic is applied to biology, a strange symmetry appears. Neurons, like words, only make sense in context. Meaning, whether in a pa...