The processing speed of the human brain is an enduring puzzle in neuroscience, a paradox that challenges traditional beliefs about the brain's capabilities and constraints. This post drew on Carl Zimmer's informative December 26, 2024 New York Times article, "The Speed of Human Thought Lags Far Behind Your Internet Connection, Study Finds."
The Unbearable Slowness of the Brain
While our sensory systems collect massive amounts of data, with a single eye processing visual input at an estimated rate of 1.6 billion bits per second, the brain reduces this torrent into a trickle, allowing only 10 bits per second to drive conscious behavior and thought. This remarkable disparity, dubbed the "unbearable slowness of being" by researchers Jieyu Zheng and Markus Meister, serves as both a revelation and a challenge for understanding human cognition.
Understanding Bits Per Second
To appreciate this paradox fully, one must first grasp what is meant by "bits per second" in the context of neuroscience. A bit is a unit of information, representing the resolution of uncertainty between two equally probable outcomes. The term originates from information theory, developed by Claude Shannon in the mid-20th century, which provides a mathematical framework for quantifying information flow through a system. When we say the brain processes 10 bits per second, we are describing its ability to discriminate among 2¹⁰, or 1,024, possible actions or thoughts in one second. This is astonishingly slow compared to the data rates of our sensory organs and even modern internet connections, which operate in megabits or gigabits per second.
The Evolutionary Basis for Cognitive Slowness
Despite its apparent slowness, the human brain's throughput is adequate for survival, an idea supported by evolutionary theory. Early nervous systems in simple organisms evolved to handle single, serial tasks like detecting food or evading predators. The architecture of such systems prioritized accuracy and resource efficiency over speed. This evolutionary blueprint persists in modern humans. Our brains are not designed for parallel processing on a large scale; instead, we focus intensely on one cognitive stream at a time. Even when multitasking, we merely switch rapidly between tasks rather than processing them simultaneously.
Demonstrating the Bottleneck: Typing, Gaming, and Puzzles
The slowness of human cognition is vividly demonstrated in behaviors as diverse as typing, playing video games, and solving puzzles. Consider the case of expert typists, who can produce up to 120 words per minute—a speed requiring approximately 10 bits of information per second. Although this might seem fast to an observer, it pales in comparison to the potential capacity of the nervous system. This discrepancy arises because tasks like typing rely on finely tuned motor skills and the predictability of language. English, for example, is highly redundant, with predictable patterns that reduce the cognitive load on the typist. This redundancy enables us to type relatively quickly without needing to process vast amounts of new information.
Similarly, competitive video gamers, who perform rapid sequences of actions in high-stakes environments, exhibit the same processing limits. Professional players of games like StarCraft or Tetris reach impressive speeds, measured in actions per minute (APM). Yet even at their peak performance, their information throughput hovers around 10 to 16 bits per second, highlighting the inherent bottleneck in human neural processing. Interestingly, these limits are not due to the physical constraints of muscles or fingers but reflect the brain's capacity to process and translate sensory information into coordinated actions.
Exceptional Mental Feats, Same Limitations
Further evidence comes from extraordinary mental feats, such as blindfolded speedcubing and memory sports. In blindfolded speedcubing, competitors inspect a scrambled Rubik's Cube for a few seconds before solving it from memory while blindfolded. The American speedcuber Tommy Cherry, for instance, inspected his cube in just 5.5 seconds and solved it in 7.5 seconds, achieving an information rate of roughly 11.8 bits per second. Similarly, memory champions who memorize thousands of digits or the order of a shuffled deck of cards also exhibit low information rates, typically around 5 to 18 bits per second. These results suggest that even when physical movement is removed from the equation, the brain's cognitive processing remains constrained by its inherent bottleneck.
The Outer and Inner Brain: Dividing the Work
What causes this bottleneck? Neuroscientists propose that the brain operates in two distinct modes: an "outer brain" and an "inner brain." The outer brain, which includes peripheral sensory and motor systems, handles vast amounts of high-dimensional data in parallel. For example, the retina's photoreceptor cells convert light into neural signals at rates exceeding 1 gigabit per second. These data are then compressed as they move through the optic nerve and visual cortex, distilling essential features for further processing. In contrast, the inner brain, which governs decision-making, memory, and attention, operates on a dramatically reduced data stream. This reduction allows the brain to focus on behaviorally relevant information while filtering out noise.
Subjective Inflation: The Illusion of Rich Experience
The dichotomy between the outer and inner brain explains many peculiarities of human cognition. For instance, why do we experience a vivid, richly detailed visual world when our brains process only a fraction of the sensory input we receive? This phenomenon, known as subjective inflation, arises because the brain integrates sensory data with expectations, memories, and contextual cues to create the illusion of completeness. However, this illusion breaks down under scrutiny, as demonstrated by experiments on inattentional blindness and change blindness. These studies reveal how much of the visual scene we ignore when our attention is focused elsewhere.
Implications for Brain-Computer Interfaces
The implications of this slowness extend beyond neuroscience into technology and philosophy. For example, brain-computer interfaces (BCIs), which aim to connect the brain directly to machines, face significant challenges due to the brain's low information rate. Efforts to restore vision through retinal implants or enable paralyzed patients to control robotic limbs often falter because they attempt to replicate the brain's peripheral processing, which operates at much higher rates than its conscious output. A more effective approach might involve designing BCIs that align with the brain's natural throughput, delivering simplified but meaningful information.
Humans vs. Machines: The Speed Gap
Similarly, the disparity between human and machine processing highlights the limitations of human cognition in an increasingly automated world. While artificial systems process data at kilobits or even megabits per second, humans remain bound by their 10 bits-per-second constraint. This difference explains why machines outperform humans in tasks like image recognition, real-time strategy games, and autonomous navigation. However, it also underscores the unique strengths of human cognition, such as creativity, intuition, and the ability to synthesize disparate pieces of information into coherent narratives.
Unpacking the Paradox
The paradox of the brain's slowness invites deeper questions about its design and function. Why does such a sophisticated organ, with billions of neurons and trillions of synapses, operate at such a glacial pace? And how does this constraint shape our perception, behavior, and culture? The answers may lie in the brain's evolutionary history, its division of labor between parallel and serial processing, and the trade-offs between speed, accuracy, and flexibility. Understanding these dynamics not only illuminates the mysteries of human cognition but also provides a foundation for advancing neuroscience, artificial intelligence, and human-computer interaction.
Glossary
action per minute (APM): a measure of how many actions a player can perform in a minute during a video game, used to assess speed and efficiency.
attentional bottleneck: a cognitive limitation where only a small amount of information can be processed consciously at any given moment.
bit: the basic unit of information in information theory, representing a choice between two equally probable alternatives.
bits per second (bps): a measure of information flow, quantifying how much data is transmitted or processed per second.
brain-computer interface (BCI): a technology that establishes a direct communication pathway between the brain and external devices.
change blindness: a phenomenon in which significant changes in a visual scene go unnoticed by an observer.
Claude Shannon: the father of information theory, who developed a mathematical framework to quantify information transmission.
cognitive stream: the focused line of thought or attention that the brain processes at a given moment.
conscious throughput: the rate at which the brain processes information that reaches conscious awareness, estimated at 10 bits per second.
evolutionary blueprint: the inherited design of the nervous system shaped by evolutionary pressures, favoring efficiency over speed.
information theory: a branch of applied mathematics and electrical engineering that studies the transmission, processing, and storage of information.
inner brain: the part of the brain responsible for conscious thought, decision-making, and memory, operating at a reduced data stream.
inattentional blindness: a phenomenon where an individual fails to notice an unexpected stimulus in their visual field when attention is focused elsewhere.
megabit: a unit of information equal to one million bits, often used to describe data transfer rates.
multitasking: performing multiple tasks in rapid succession rather than simultaneously, as the brain cannot process many tasks at once in parallel.
outer brain: the peripheral sensory and motor systems that process vast amounts of high-dimensional data in parallel.
photoreceptor: a type of neuron in the retina that detects light and converts it into neural signals, with a high data transmission rate.
redundancy: the predictable patterns or repeated elements in language or sensory data that reduce the complexity of processing.
Shannon's entropy: a measure of uncertainty or randomness in information theory, quantifying the amount of information required to resolve uncertainty.
subjective inflation: the brain’s creation of a richer perceptual experience than the actual amount of sensory information processed.
throughput: the amount of data processed or transmitted by a system within a given time frame.
visual cortex: a region of the brain responsible for processing visual information, part of the outer brain.
Google Illuminate Discussion
Please click on the podcast icon below for a lively Google Illuminate discussion of this paper.
Article
Zheng, J., & Meister, M. (2024). The unbearable slowness of being: Why do we live at 10 bits/s?
Neuron. https://doi.org/10.1016/j.neuron.2024.11.008 A preprint is available at arxiv.org.
Please Support Our Friends
Comentarios