E-Paper
Scientifically Speaking | The surprisingly slow speed of human thought
The future could be in expanding our limited bandwidth of consciousness and in developing AI systems that can process the flood of information we filter out
Deep within the epic Mahabharata, Prince Yudhishthira faces a divine riddle from a yaksha. "What moves faster than the wind?" He answers swiftly: "Mind." In the same dialogue, when asked what outnumbers the blades of grass, he responds: “Thoughts.”
This ancient understanding of thoughts as lightning-quick and unlimited in number has persisted through human history. Now, a research perspective published in the journal Neuron reveals a startling paradox that would have given Yudhishthira pause. Our conscious thoughts crawl along at just around 10 bits per second.
Science writer Carl Zimmer sums up the collective disappointment in the snail’s pace of the mind in a headline to a piece in the New York Times covering the discovery – “The Speed of Human Thought Lags Far Behind Your Internet Connection.”
The research, conducted by Jieyu Zheng and Markus Meister at Caltech, unveils a profound mystery. On the one hand, our senses flood our brains with information at a staggering billion bits per second: for example, the retina of the eye alone has a capacity of up to 1.6 gigabits per second under bright conditions. On the other, our conscious thoughts seem to emerge as a mere trickle. It’s as if a Niagara Falls of information is being filtered down to a drinking straw. This staggering reduction by a factor of hundreds of millions highlights the sharp contrast between sensory input and our brain’s conscious output.
The evidence presented in the article by Zheng and Meister entitled “The unbearable slowness of being” spans the spectrum of human achievement. Take the world's fastest typists, hammering away at 120 words per minute. Or "speedcubers" who memorise a scrambled Rubik's Cube in seconds before solving it blindfolded. Even champions of competitive video games, whose fingers blur across keyboards in seemingly superhuman displays of speed. They all work within this modest bandwidth of around 10 bits per second.
"Every moment, we are extracting just 10 bits from the trillion that our senses are taking in and using those 10 to perceive the world around us and make decisions," explains Meister. “This raises a paradox: What is the brain doing to filter all of this information?”
Before moving on, I should mention that this study focuses solely on conscious processing, leaving out unconscious neural activity that keeps us standing, walking, or recovering from a stumble. The study’s focus on task-relevant conscious output suggests that 10 bits per second reflects the narrow bandwidth of conscious thinking rather than the brain's total processing capacity. But it’s still a strikingly low number that raises profound questions. What gives?
Why do we have 86 billion neurons if we're processing conscious information at the speed of a 1980s modem? How does our brain choose which handful of bits to elevate to consciousness from the torrential flood of sensory data? And perhaps most puzzlingly, why can we only think about one thing at a time?
The researchers have a lot of theories, but few solid answers to the conundrum.
They suggest an evolutionary explanation for the simple and slow processing of thoughts. The earliest of our ancestors that had nervous systems needed only to navigate simple environments with simple requirements- finding food, and fleeing predators. This single-track processing would have been sufficient for survival, and our consciousness may have inherited this architectural structure. Modern human thinking, the researchers propose, navigates through abstract concepts bound by the limitations of its primitive origins.
Yet this explanation deepens the mystery. Our peripheral nervous system (eyes, ears, and skin) processes information in massively parallel streams. But somewhere between sensation and consciousness, this vast parallel highway narrows to a rural dirt road. The researchers hypothesize that the brain operates in two distinct modes: an "outer brain" handling the sensory flood, and an "inner brain" distilling it down to the essential few bits that guide our behaviour.
This discovery has immediate implications for ambitious technological ventures. Dreams of high-bandwidth brain-computer interfaces, championed by the like of Elon Musk, would be limited by the fundamental speed limit of consciousness, the researchers argue. No matter how many electrodes we implant or how fast our computers become, we will be ultimately bound by our brain's conscious processing bottleneck.
This bottleneck explains why true multitasking is a cognitive myth. What we experience as juggling multiple tasks is rapid switching between single threads of thought, each constrained by the same 10-bit limit. When we attempt to simultaneously process multiple streams of information like reading emails during a video call, we are not expanding our bandwidth, rather, we are fragmenting it, often at the cost of performance and comprehension. The total processing pie remains the same, we’re only slicing it thinner.
On the flip side, what this research shows is that the brain’s architecture is tuned for deep, sequential processing rather than parallel streams of conscious thought. This is not a new observation, though our subjective experience makes this hard to believe. We feel as though we're taking in every detail of our surroundings, keeping multiple trains of thought, and processing the world in full richness. But phenomena like 'inattentional blindness' where we fail to notice obvious changes in our environment and the 'cocktail party effect' where we can only truly follow one conversation in a crowded room reveal this sense of cognitive completeness is an illusion.
Daniel Kahneman breaks this down brilliantly in Thinking, Fast and Slow where he shows that our minds toggle between quick, instinctive thinking and slower, more deliberate processing. Cal Newport takes this insight and runs with it in Deep Work pointing out that we do our best work when we're fully locked in and the notifications are turned off. So, when we give ourselves permission to focus deeply on one thing, we're working in harmony with our brain's natural design.
Another puzzling outcome is that the trickle of conscious thought requires billions of neurons to implement. As the researchers note, traditional explanations that the brain needs this vast neural machinery for redundancy or noise correction don't hold up to scrutiny. One possibility is that each of our countless daily micro-tasks from reading road signs to interpreting facial expressions requires its own specialized neural subnet, with our consciousness engaging them one at a time in rapid succession. The brain may be an enormous orchestra where each section specializes in a different type of music, but only one section can play at any given moment: we need each section to play seamlessly for the orchestra to play a grand cognitive concert.
Perhaps this bottleneck is not a limitation but a necessary optimization. In a world of increasing information overload, our brain's ability to filter a billion bits down to the essential ten might not be a bug but a feature that has served us well since our ancestors developed the capacity for thought.
It also has to be said that the processing limits of our brains point to a balancing role for artificial intelligence. While our consciousness requires extreme filtering of information, AI systems can process the full bandwidth of sensory data our brains discard. For instance, an AI system might help a scientist notice subtle patterns in climate data that would take humans decades to discover, or help doctors identify early disease markers that are too subtle even for experienced clinicians.
The future might lie not in expanding our limited bandwidth of consciousness, but in developing AI systems that can process the flood of information we filter out. That would be a true brain-machine interface. The wind may be quite a bit faster than conscious thought. But perhaps that's exactly how it should be.
Anirban Mahapatra is a scientist and author, most recently of the popular science book, When the Drugs Don’t Work: The Hidden Pandemic That Could End Medicine. The views expressed are personal.
