Practice appears to rewire multitasking in the brain

Practice may change how the brain handles two tasks at once

Why multitasking feels difficult has long been framed as a basic limitation: either the brain can process only one demanding task at a time, or different tasks compete so heavily for the same neural machinery that performance inevitably breaks down. New research points to a more flexible answer. In a study published in Neuron, scientists report that the brain appears to reorganize itself during training, shifting from interference-heavy processing toward a more separated, efficient arrangement for handling parallel demands.

The work was co-led by researchers at City University of Hong Kong and the Chinese University of Hong Kong. Instead of looking only at behavior, the team tracked the activity of the same neurons over weeks while mice learned to carry out a dual-task challenge. That design let the researchers watch how multitasking changed not just in outcome, but in underlying neural organization.

The central conclusion is that multitasking is not governed solely by an immutable bottleneck. Early on, the brain showed both competition for resources and active coordination between tasks. With practice, however, it increasingly recruited task-specific neurons and separated the neural representations of the two activities, making simultaneous performance more effective.

A closer look at how the experiment worked

To probe parallel task processing, the researchers designed a setup in which mice had to keep a continuous lever-movement task going while also listening for auditory cues and making a Go/No-Go decision. That is a meaningful test because it forces the brain to manage an ongoing motor requirement while evaluating incoming sensory information and selecting a response.

The team then used longitudinal two-photon calcium imaging to follow large populations of neurons in the secondary motor cortex, known as M2. Because the same neurons were tracked across several weeks of training, the researchers could observe how patterns changed as the animals improved.

That matters because debates over multitasking often rely on static snapshots. A single measurement can show that tasks interfere, but it cannot easily reveal whether the brain treats that interference as a permanent constraint or as something that can be reorganized through learning. This study argues for the latter, at least in this experimental model.

From interference to separation

At the start of training, the neural picture was messy in a way that will sound familiar to anyone who has tried to do two things at once. Signals associated with the two tasks overlapped more heavily, meaning the same neural resources were engaged across both demands. That overlap can create conflict: activity useful for one task may disrupt the other, or at least force the brain to constantly coordinate competing priorities.

According to the researchers, the brain did not solve that problem merely by becoming faster at using the same shared circuitry. Instead, training gradually changed the organization of the system. More neurons became specialized for one task or the other, and the representations of those tasks became more distinct.

That separation appears to be the key insight. Efficient multitasking, in this account, is less about eliminating limits altogether and more about restructuring the network so that different jobs interfere less with one another. Practice does not magically create infinite capacity. It helps carve cleaner lanes.

The findings also help explain why some kinds of multitasking improve with repetition while others remain frustrating. If training can push the brain toward more task-specific organization, then repeated exposure to the same dual-task demands may make them feel smoother over time. But when two tasks continue to rely on overlapping processes, conflict may remain stubborn.

Why the result matters beyond basic neuroscience

The study adds nuance to a long-running scientific argument. The classic “central bottleneck” view treats multitasking as being constrained by a narrow processing stage that only one task can pass through at a time. Competing theories emphasize shared-resource limits. The new findings do not simply endorse one side and reject the other. Instead, they suggest the brain begins with a mix of competition and coordination, then can move toward greater segregation through learning.

That makes the result interesting for fields far beyond animal neuroscience. Training design, rehabilitation, human factors engineering, and cognitive performance research all depend on a better understanding of when parallel performance can be improved and when it cannot. The work suggests that task structure and repetition may matter as much as baseline capacity.

The researchers also said the biological findings could inform artificial intelligence training. That idea is straightforward: if brains become better at multitasking by differentiating internal representations and recruiting more specialized units, AI systems might also benefit from training approaches that reduce interference between tasks rather than forcing a single shared representation to do everything equally well.

What the study does and does not say

The paper is careful not to claim that all multitasking is easily trainable or that the brain fully escapes limits. The work was done in mouse models, using a specific dual-task design and measurements from one cortical region. That gives the study precision, but it also sets boundaries around what can be concluded.

Even so, the result is notable because it shows a concrete biological mechanism for improvement. Rather than treating better multitasking as a vague behavioral outcome, the study ties it to observable neural change over time. That strengthens the case that learning can reshape how parallel task processing is implemented in the cortex.

• The study was published in

  Neuron.

• Researchers tracked the same neurons across weeks of training.
• Early multitasking involved both resource competition and neural coordination.
• Practice led to more task-specific neurons and more separated task representations.
• The authors said the findings may help guide AI systems designed to multitask.

For everyday life, the message is restrained but useful: practice may not make people universally better at doing many things at once, but it can help the brain reorganize around repeated combinations of tasks. In other words, what feels impossible at first may become manageable not because the brain has no limits, but because it learns to spend less time making two jobs collide.

This article is based on reporting by Medical Xpress. Read the original article.