How AI Is Unraveling the Mystery of Human Motion — Insights from Allison Hamilos

Many researchers are using artificial intelligence to probe a central question in neuroscience: how does the brain generate spontaneous, elective movement? In a recent Boston talk, Allison Hamilos, an MIT- and Harvard-trained neuroscientist affiliated with the Harvard‑MIT Health Science and Technology Program, laid out evidence tying coordinated neural activity, dopamine signaling, and probabilistic decision processes to the origins of motion.

Motor Neuron Avalanches and the Spark of Movement

Hamilos described how movement emerges when large ensembles of motor neurons fire together, producing an "avalanche" of activity that travels down the spinal cord and causes muscle fibers to contract. The central question is not how muscles contract, but how those synchronous bursts are initiated.

Reactive Versus Elective Actions

She emphasized an important distinction: many movements are immediately reactive to sensory input, while others are self-generated or elective and often appear capricious. It is these self-generated actions that most intrigue neuroscientists because they are not tied to abrupt external events.

Not all movements result from abrupt sensory events. It is precisely those self-generated actions that provoke the greatest curiosity among researchers.

Clinical Windows Into Mechanism

Hamilos drew on clinical disorders to illustrate how underlying circuits influence movement. In Parkinson's disease, motivational circuits that normally facilitate voluntary motion are impaired, producing slowed movement overall and phenomena such as paradoxical kinesia, where a person may respond quickly to an urgent external stimulus but remain inhibited for elective actions like standing or picking up an object.

She contrasted that with conditions such as Tourette's syndrome, where individuals experience excessive, apparently involuntary motor or vocal tics. These opposing clinical presentations suggest a balance in neural systems that both generate and constrain spontaneous actions.

The Role of Dopamine and Probabilistic Effects

Research points to dopamine as a critical modulator of motor neuron activity. Hamilos highlighted that dopamine appears to influence movement in a probabilistic, not strictly deterministic, way — biasing the likelihood that certain neural ensembles will ignite. This probabilistic influence helps explain both variability in behavior and the persistence of apparent choice or spontaneity.

Animal Studies and Interrupted Movements

Much foundational work has come from experiments in mice, where researchers can observe interrupted movements, evasion strategies, and the neural dynamics that precede or abort actions. These studies support a model in which timing, selection among competing actions, and perception interact to produce adaptive, sometimes unpredictable behavior.

Three Components of Motion Control

• When To Move: Timing and internal triggers that initiate action.
• Which Option To Choose: Selection among multiple competing motor plans.
• Perception And Context: Sensory and contextual inputs that bias or gate choices.

Hamilos also linked delayed motor responses to cognitive slowing (bradyphrenia) and perseveration, where an individual repeats the same response. Clinically, these patterns can appear with flattened affect and reduced spontaneity, underscoring how motor and cognitive systems are intertwined.

Behavioral Stochasticity And A Shared Circuit Hypothesis

To unify diverse findings, Hamilos proposed the language of behavioral stochasticity and a shared dopamine-circuit mechanism. The idea is that a core circuit coordinates self-generated neural activity and that dopamine modulates the circuit's probabilistic dynamics, explaining both too-little and too-much spontaneous action across disorders.

Where AI Fits In

Hamilos argued that AI and computational modeling can help capture the probabilistic, high-dimensional patterns of self-generated neural activity. Machine learning methods allow researchers to model timing, choice, and perceptual gating jointly, uncover latent circuit motifs, and generate testable predictions. Applications could include sharper diagnostics, personalized neuromodulation strategies, and improved brain-computer interfaces that respect the stochastic nature of human action.

Her talk offered a clear reminder: understanding spontaneous movement is not a theoretical curiosity but a practical necessity for people affected by disorders that impair or exaggerate spontaneity. By combining animal experiments, clinical observations, and AI-driven models, neuroscience is charting a path toward therapies and technologies that better reflect how the brain actually decides to move.

This article originally appeared on Forbes.com.