For decades, textbooks described decision-making as a top-down process: raw signals enter the senses, climb through the brain's hierarchy, and only then does a "higher" region announce a choice. A new study turns that picture on its head.
Researchers at the University of Illinois Urbana-Champaign, publishing in the Proceedings of the National Academy of Sciences, recorded from animals as they performed a simple sensory task. They found that decision-related signals emerged in early sensory regions — the parts that first receive input — far earlier than the traditional model predicts, and well before the signals reached higher cortical areas.
The work matters for two reasons. First, it suggests the brain is an active interpreter from the very first processing stage, not a passive relay that only "decides" at the top. Second, it carries a lesson for artificial intelligence: systems that copy the brain's layered architecture often assume decisions live only at the output layer, when biology may be committing to answers much deeper in the stack.
Knowledge takeaways: (1) decision signals can appear in the brain's earliest sensory regions, not just higher cortical areas; (2) the finding comes from a PNAS study by University of Illinois researchers using direct neural recordings during a sensory task; (3) it challenges the top-down model of cognition and has direct implications for how we design layered AI systems.