For decades, neuroscience has maintained that the human brain processes information in a sequential manner: first, it receives stimuli through sensory regions, and then higher-order areas, such as the frontal cortex, are responsible for making decisions. However, new research from the University of Illinois Urbana-Champaign challenges that explanation and proposes a far more dynamic model. Scientists discovered that the brain begins making decisions much earlier than previously believed, thanks to a complex feedback system connecting different brain regions.

READ ALSO. The fentanyl vaccine that could change everything: it prevents the drug from reaching the brain and causing an overdose

The findings were published in the journal Proceedings of the National Academy of Sciences (PNAS) and, beyond expanding our understanding of how the human brain works, they could pave the way for a new generation of artificial intelligence (AI) inspired by the biological architecture of the nervous system.

The brain does not process information in a single direction

The study was led by Yurii Vlasov, professor of Electrical and Computer Engineering at the Grainger College of Engineering, who explains that the classical model of the brain has served for years as the inspiration behind technologies such as the convolutional neural networks used in artificial intelligence.

According to this traditional view, information follows a linear path:

• The senses receive external stimuli.
• The information moves through different layers of the brain.
• It ultimately reaches the frontal cortex, where decisions are made.

However, the new findings reveal that this process is far more complex.

Researchers argue that different brain regions remain in constant communication through feedback loops, allowing information to travel both forward and backward before a decision is fully formed.

The brain starts making decisions in the earliest sensory regions

To test this hypothesis, the team studied the neuronal activity of mice as they navigated a virtual reality environment and made decisions based on the stimuli they observed.

Scientists focused particularly on the activity of the primary somatosensory cortex (S1), one of the first brain areas responsible for processing sensory information.

Surprisingly, they found that this region was not merely transmitting information to higher brain areas.

Instead, it also received signals from other parts of the brain, altering its activity almost in real time.

This discovery suggests that the decision-making process begins much earlier than the brain regions traditionally associated with reasoning.

Artificial intelligence could learn from the brain

One of the main goals of the study is to use these insights to improve the next generation of artificial intelligence systems.

Today, AI models consume enormous amounts of energy to perform tasks that the human brain executes with remarkable efficiency.

While an adult human brain consumes approximately 20 watts of power, many data centers dedicated to training artificial intelligence require megawatts of electricity.

According to Vlasov, understanding how evolution has optimized the brain’s architecture over hundreds of millions of years could help researchers develop far more efficient algorithms.

“We want to learn from a billion years of evolution. How is biological intelligence architecturally organized? Can we emulate that organization to create AI that is more efficient, consumes less energy, and is smarter than today’s systems?” the researcher explained.

Scientists still do not understand the brain’s language

Although the study represents a significant breakthrough, researchers acknowledge that much remains to be discovered.

The so-called “neural code”—the way neurons represent and transmit information—remains one of neuroscience’s greatest mysteries.

For this reason, the team aims to develop new tools capable of recording brain activity with greater precision and analyzing how these feedback circuits emerge within fractions of a second.

The next step is understanding the timing of decisions

Researchers believe that understanding how quickly different brain regions interact will be key to explaining how decisions emerge.

Their goal is to study the temporal dynamics of these signals in much greater detail and determine how different levels of brain processing coordinate with one another.

According to Vlasov, this knowledge could become the foundation for future artificial intelligence architectures capable of reasoning in a way that more closely resembles the human brain.

A discovery that changes our understanding of the brain

The study also challenges one of the most widely accepted ideas in neuroscience: that sensory regions are simply passive receivers of information.

Researchers now believe these areas actively participate in constructing decisions from the very earliest stages of information processing.

Although the study does not yet provide a blueprint for building artificial intelligence comparable to the human brain, it offers a new perspective on how biological intelligence works.

Scientists believe that understanding these mechanisms could mark the beginning of a new era for both neuroscience and the development of smarter, more adaptive, and more energy-efficient AI systems.