Somewhere on a scanner in Seoul, a thought is being mapped onto the body. Somewhere on a server in Buenos Aires, a language model is being trained to describe what a wandering mind feels like from the inside. Somewhere in a lecture hall in Edmonton, a professor is asking what happens to a student’s thinking when an AI writes the first draft. None of these projects know about each other. All of them are studying the same thing, at the same moment, from opposite sides of the world.

The oldest attempt to study thought scientifically may also be the least discussed. Ancient Egyptian physicians performed the first documented anatomical investigations of the brain, thousands of years before neuroscience had a name. That thread nearly disappeared from the modern research map. It is now being deliberately picked back up: a 2026 initiative called BRIDGE has named Africa’s missing neuroscience data a real gap in the field, not a footnote, and researchers across Egypt, Ghana, South Africa, Nigeria, Tunisia, and Morocco are working to close it.

In Seoul, a research group has spent the past several years asking a question that sounds almost too simple: what does a thought feel like in the body, not just the brain? Their 2026 study on spontaneous thought dynamics found something specific, that the pattern of a wandering mind carries a measurable signature of positive or negative feeling, traceable and distinct enough to map. A thought, in this framework, is not confined to the skull. It has a bodily address.

Singapore has become, almost quietly, one of the field’s major hubs. Its 2026 computational neuroscience conference drew researchers from across the Asia-Pacific to compare notes on a shared problem: how large datasets and mathematical models can be used to catch thought in the act, not just after the fact. The region, once a minor presence in this field, is now described by its own participants as a global leader in it.

In Belgium, researchers have been quietly revisiting an older method for a newer problem. Most thought research interrupts a person to ask what they were just thinking, a method that works but breaks the very flow it is trying to observe. A University of Liège team has been testing a return to continuous, spoken narration instead, an attempt to watch a stream of thought without damming it first to measure it.

In Argentina, a research team took an unusual next step: training a large language model to read written descriptions of spontaneous thought and rate their emotional and structural qualities the way a trained human researcher would, then checking those AI-generated ratings against real-time brain activity recorded by EEG. The ratings held up. A tool built to generate language is now being used to help decode the texture of a wandering mind.

In Canada, a different kind of study raised a more uncomfortable question. Surveys tracked a sharp rise in generative AI use among university students, from just over half in 2024 to nearly three-quarters two years later. Close to half of the same students reported that their own critical thinking had grown weaker since. The tools used in one country to study thought with unprecedented precision are, in another setting, being named as something that may be dulling it.

In New Zealand, one of the only dedicated interdisciplinary cognitive science programs in the country continues a decades-long project of trying to hold this whole picture together: psychology, computer science, linguistics, and philosophy, all pointed at the same target, how thinking and behavior can be studied at all.

No single lab, on any continent, currently holds a complete account of what a thought is. What the current research does show, taken together, is a shift in where the questions are even being asked. Not from one dominant center outward, but from Seoul and Buenos Aires and Accra and Liège and Dunedin at once, each running a different instrument over the same, still largely unmapped, territory.