The Brain-Computer Debate: Is the Human Brain Like a Computer?
Daniella Lorman
Illustrations by Ava Sclafani
In the late 1940s, pioneering computer scientist Alan Turing asked what it would take for a machine, or a computer, to think for itself. Turing predicted that, by the year 2000, “one will be able to speak of machines thinking without expecting to be contradicted” [1]. Turing’s comment was admittedly a bit optimistic. Now, over seventy years later, there is debate amongst philosophers, neuroscientists, and computer scientists regarding the similarities between the brain and a digital computer [2, 3]. If machines could theoretically ‘think' and process information like the human brain, would this mean that human brains process information like machines? To begin answering this question, let’s delve into the architectural differences between a typical digital computer and the human brain.
Transistors vs. Neurons: How do Computers and Brains Compare?
Our first instinct may be to think of a computer as a consumer device such as a Mac, a PC, or a smartphone. However, a plethora of different computer types exist, ranging from the most primitive analog computers to massively complex quantum computers. According to some neuroscientists, a computer is generally any system that transmits signals, integrates information, and converts inputs into outputs [3]. While the brain and the computer both process information, defining a computer as something that receives information, manipulates it, and determines outputs does not account for the embodied nature of a human brain, nor the impacts of external processes and environmental factors on cognition. Modern computers are largely based on the Von Neumann architecture, first developed in 1945 [4]. A Von Neumann based computer uses a single central processor (CPU), which loads and executes instructions, as well as a memory unit that holds instructions and data. There are also peripherals, or output devices that send data to human users or other computers. The CPU, memory unit, and peripherals are all connected via communication systems known as ‘buses.’ A computer constructed around the Von Neumann architecture processes information sequentially, following the specific order of instructions encoded in the memory unit [4].
In contrast to Von Neumann based computers, the human brain engages in parallel processing, or the processing of multiple tasks continuously and all-at-once [5]. Most sensory and motor system neural pathways, for example, intercept several brain regions. Even within each brain region, there are multiple neural networks that operate at the same time. In this way, the human brain’s architecture differs from the more rigid Von Neumann model, in which information is processed in discrete, sequential intervals. Notably, however, computer systems that process in parallel do exist, such as those with more than one processing unit. In fact, modern technologies like the iPhone implement parallel processing.
Before taking a position in the debate, it’s imperative to consider prevailing arguments considering the structure of the human brain relative to a computer. Some academics have argued that the biological brain functions as a massive parallel computer [6]. For example, computer scientist Marvin Minsky argues that the human brain is formed by several ‘agents’ — ak in to the transistors in a computer — each of which is mindless [7, 8]. Together, these ‘agents’ create the mind by processing information simultaneously [7]. Others broadly define a computer, arguing that both the brain and the computer essentially complete the same task: information processing [3]. In addition to the many philosophical arguments conceptualizing the human brain, the inability to isolate what in particular contributes to parallel processing in the human brain poses an obstacle in drawing comparisons. Several computational models have been employed to study or attempt to simulate the parallel processing characteristic of the human brain, yielding results suggesting that parallel processing of various tasks in the human brain occurs at the level of neural regions rather than individual neurons themselves [5]. However, many questions remain unanswered regarding the precise role of individual neurons, neuronal connections, and structural regions that contribute to parallel processing in the human brain. For example, we have not yet been able to determine just how many processes the human brain is capable of processing at the same time [5]. Until the precise mechanism of parallel processing within the human brain is elucidated, equating the human brain to a computer may be an apples to oranges comparison.
The Impact of External Processes on Cognition
A fundamental difference between the human brain and a device — such as a digital computer — is that the brain’s functioning is inherently dependent on its attachment to and integration with the body. This theory, known as embodied cognition, posits that cognition is shaped by the physiology of the entire organism [9]. Considering the brain as a computer may obfuscate the fact that a brain is still a wholly active biological organ, part of a larger body that interacts with the surrounding environment. Consequently, the brain and its cognitive abilities are subject to a plethora of diseases and environmental factors, both of which impact the human body. When considering the structure of the human brain as a computer, philosophers often focus solely on the role of neurons in cognition and information processing, commonly comparing neurons to transistors, or the electronic devices within a computer that regulate electric current [3, 7, 10]. But, the neuroscientific reality is that there is much more to the functioning of a human brain than the contribution of neurons.
Another major cell type in the nervous system occupies over half the volume of the human brain and is far more numerous than neurons: glial cells. Glial cells are non-neuronal cells that, unlike neurons, do not produce electrical impulses. A variety of glial cell subtypes exist — including astrocytes, oligodendrocytes, and microglia — and the functions of each glial cell type are similarly diverse. Astrocytes remove waste material and interact with the brain’s extensive vasculature to provide cells with nutrients. Further, astrocytes maintain brain homeostasis by modulating the chemical environments in the brain that are critical for communication, or signaling, between neurons [11]. For example, astrocytes modulate the recycling of the primary excitatory neurotransmitter in the brain, glutamate. Glutamate is crucial in regulating learning and memory processes, but dysregulation in glutamate processing is widely correlated with the progression of neurodegenerative diseases [12, 13]. In this way, astrocytes are directly involved in the facilitation of learning and memory processes in the brain [12, 13]. Oligodendrocytes are responsible for creating and maintaining the myelin sheath in neurons, a ‘sleeve’ of insulating material surrounding neurons that allows for quicker electrical signaling or communication between neurons. Similar to astrocytes, when oligodendrocytes are compromised and unable to produce myelin sheath, they contribute to the pathology of neurodegenerative disease [14]. Finally, microglia maintain neural networks by responding to injury, inflammation and pathogens. As such, microglia are commonly referred to as the ‘immune cells’ of the brain.
In addition to glial cells, abnormal cell types may arise in the human brain, such as in instances of neurodegenerative disease [15, 16, 17]. For example, individuals with Alzheimer’s disease experience an abnormal buildup of the tau protein, which conventionally helps stabilize the internal skeleton of neurons in the brain. These abnormal tau proteins form tangles by clinging to other proteins in the brain, and may even build up within astrocytes. Another hallmark of Alzheimer’s disease is the accumulation of aggregates or plaques of a protein called beta-amyloid. Together, tau tangles and amyloid plaques disrupt neuron function and contribute to cell damage and death [15, 16, 17]. A common thread across neurodegenerative conditions is that their development has little to do with a single neuron [18]. Rather, while a dysfunctional non-neuronal cell or protein results in damage to the neuron itself, pathways of neuronal connections are required to propagate electrical impulses. As a result, there is limited use in investigating the role of solely one neuron in modulating cognitive processes [18].
At first glance, some may consider glial cells and tau proteins to be internal factors, or components of the human brain’s broader information processing system. After all, these structures exist in the brain and influence neuronal functioning. In the ‘brain as a computer’ metaphor, neurons are often equated with the transistors that make up the computer’s combined information processing components (the central processing unit and memory). However, in the computer, there is no substance or mechanism that is immediately analogous to that of a glial cell, tau tangle, or beta-amyloid plaque. Using this foil, we can see that glial cells and tau proteins can impact cognition by impacting neurons, but they appear to function externally to the brain’s system of information processing. This marks one of the major differences between a biological brain and a computer; cognition in the human brain is impacted by biological factors outside of the immediate information processing system. In the inorganic computer, these factors are simply not present.
The Brain is a Biological Organ. A Computer is Not.
Let’s consider a seemingly obvious statement: the brain is a biological organ, while the computer is a digital system. Conceptualizing the human brain in relation to a digital system is difficult [2, 3]. As a biological organ, the brain has a natural evolutionary history that has forged its structure and function, in tandem with the body. Computers, on the other hand, have advanced through developments created by human scientists and software developers; as it stands, computers cannot yet fully program themselves or evolve via a system comparable to Darwinian evolution [19]. For instance, Microsoft Windows cannot identify, troubleshoot, or repair errors or shortcomings in its own code, nor can it organically ‘evolve’ into the next, more ‘fit’ version of Windows.
Diseases and disorders occurring outside of the brain can impact its functioning and cognition [20]. Syphilis, a disease stemming from a bacterial infection, commonly begins with a small lesion. If left untreated, syphilis can transform into a systemic, long-term disease with devastating physiological consequences. In some instances, neurosyphilis, or the spread of the disease to the central nervous system, may develop. Patients with neurosyphilis commonly experience dementia, cognitive impairment, mania, and psychosis as the disease ravages their central nervous system [20]. Syphilis is a disease with biological origins outside of the brain that causes the degeneration of neurons. Importantly, syphilis is not caused by the brain; rather, the brain becomes involved as a consequence of its embodied nature.
Computers, like a human being, can also contract viruses. A computer virus works by replicating itself to modify and damage the functioning of computer programs. Oftentimes, viruses cause critical damage to the functioning of the computer’s operating system or cause specific programs to cease working. Computer viruses are vastly different from the pathogens that infect living beings. For instance, neurosyphilis impairs cognition by physically damaging the structure of the brain and the CNS [20]. Computer viruses rarely impact the physical architecture of the computer itself, and instead only infect programs such as the operating system. In some rare instances, computer viruses may physically damage the computer by infecting its firmware, causing internal cooling systems to fail. However, the large majority of computer viruses do not affect the architecture of the transistors, CPU, or memory. When a pathogen impacts the human brain, its analogous information processing systems are physically damaged.
Beyond pathogens, organ dysfunction and conditions including improper blood sugar levels or hormonal imbalances can impact cognition [21, 22, 23]. Chronic kidney disease, which results in the kidneys being unable to properly process and filter toxins from the blood, contributes to cognitive decline by damaging cerebral vasculature [21]. Transplanting one’s failing kidneys improves performance in verbal and visual memory, spatial reasoning, processing speed, and general cognitive status [22]. The effects of kidney dysfunction on cognition represent yet another external, non-neuronal factor that impacts cognitive processing. In a computer, there is no function immediately analogous to the intertwined nature of the body and cognition. Installing additional computer memory and upgrading a hard drive are all ways in which we can improve a computer’s performance, but can we equate this to that of replacing a failing organ in the human body? Or treating a disease that impacts cognition? Components such as the memory unit, central processing unit, and hard drive are — according to Von Neumann architecture — essential to the computer’s ability to process information. These are not external structures, like the kidneys or liver; rather, they are internal components of the computer’s ‘brain.’ Would it, therefore, be possible for a human brain to process information if separated from the body?
The Brain in a Vat Problem
Envision a hypothetical — and highly unethical — experiment in which a scientist removes a human brain from its body and suspends it in a vat of self-sustaining fluid. The brain’s neurons are then individually wired to a supercomputer that produces electrical impulses identical to what the embodied brain would receive. This hypothetical experiment, known as the ‘brain in a vat’ (BIV) scenario, was conceptualized by philosopher Gilbert Harman to question our conceptions of knowledge, mind, and consciousness [24, 25]. In the BIV scenario, the disembodied brain would continue to have experiences analogous to those of an embodied brain, despite its detachment from the human body [24, 25]. However, if the brain is separated from the body, it cannot process stimuli that it would receive from the body. This idea that cognition arises from dynamic interactions between an organism and its environment is known as enactivism [26]. Enactivist theory posits that organisms do not passively receive information from their environments, such as the brain in the vat would; rather, that natural cognitive systems enact a world by participating directly in the generation of ‘meaning’ [26]. Therefore, in line with enactivist theory, the body is an essential component of cognition, since it is the agent that interacts with the environment.
Computers, on the other hand, are tasked with information processing. When a computer parses through a function, it does not assign meaning to that function. When our brains are tasked with processing seemingly arbitrary stimuli — such as interpreting fluctuating sound frequencies as spoken language, or the intangible social meanings we derive from interpersonal interactions — these interactions seem to translate into ‘meaning.’ Of course, there are non-biological devices, like a thermostat, that seemingly react to external stimuli. But, the action of a thermostat adjusting its temperature when it detects that a room is too hot, lacks intent. The thermostat, unlike a brain, does not care about its own survival, nor does it ascribe meaning to the factors in its environment. However, the human brain, as a component of a larger body, does not process information passively like that of a computer’s central processing unit. The human brain is also not a device, like a thermostat, that responds to environmental stimuli without intent or ascribing meaning. Rather, the brain is one component of a dynamic biological system: the human body.
“Meat” vs. Silicon
Even a parallel computer — which seemingly processes information in a manner which resembles that of a biological brain — is not embodied. One fundamental difference between the human brain and a computer lies in the brain’s function as just one component of a larger biological system; it is influenced by and dependent on stimuli that the body encounters. A computer, on the other hand, is a standalone information processing system. Unlike a brain, it is not enactive. A computer may not interact with its environment in the same way that the human brain can, nor may it be able to derive meaning from its environment. In the decades since John von Neumann first developed the fundamental architecture of the modern computer, computers have transformed from room-size IBM mainframes to highly advanced and ubiquitous elements of our day-to-day lives. So, is the human brain like a computer? There’s much to consider and several questions to ask before we may resolve the controversy.
References:
Oppy, G., & Dowe, D. (2021). The Turing Test. Stanford Encyclopedia of Philosophy.
Brette, R. (2022). Brains as computers: metaphor, analogy, theory or fact? Frontiers in Ecology and Evolution, 10. doi:10.3389/fevo.2022.878729
Falk, D. (2021). Is your brain a computer? MIT Technology Review.
Cragon, H. G. (2000). Computer Overview. Computer Architecture and Implementation (pp. 1–17). Cambridge University Press.
Pastur-Romay, L. A., Porto-Pazos, A. B., Cedron, F., & Pazos, A. (2017). Parallel Computing for Brain Simulation. Current Topics in Medicinal Chemistry, 17(14), 1646–1668. doi:10.2174/1568026617666161104105725
Chen, D., Hu, Y., Cai, C., Zeng, K., & Li, X. (2016). Brain Big Data Processing with Massively Parallel Computing Technology: Challenges and opportunities. Software: Practice and Experience, 47(3), 405–420. doi:10.1002/spe.2418
Winston, P. H. (2016). Marvin L. Minsky (1927–2016). Nature, 530(7590), 282–282. doi:10.1038/530282a
Minsky, M. L. (2007). The Society of Mind. Simon & Schuster Paperbacks.
Davis, J. I., & Markman, A. B. (2012). Embodied cognition as a practical paradigm: introduction to the topic, the future of embodied cognition. Topics in Cognitive Science, 4(4), 685–691. doi:10.1111/j.1756-8765.2012.01227.x
Hawkins, J., & Blakeslee, S. (2004). On Intelligence. Owl Books.
Chen, Y., & Swanson, R. A. (2003). Astrocytes and brain injury. Journal of Cerebral Blood Flow & Metabolism, 137–149. doi:10.1097/00004647-200302000-00001
Filosa, A., Paixão, S., Honsek, S. D., Carmona, M. A., Becker, L., Feddersen, B., Gaitanos, L., Rudhard, Y., Schoepfer, R., Klopstock, T., Kullander, K., Rose, C. R., Pasquale, E. B., & Klein, R. (2009). Neuron-glia communication via EphA4/ephrin-A3 modulates LTP through glial glutamate transport. Nature Neuroscience, 12(10), 1285–1292. doi:10.1038/nn.2394
Halassa, M. M., & Haydon, P. G. (2010). Integrated brain circuits: astrocytic networks modulate neuronal activity and behavior. Annual Review of Physiology, 72, 335–355. doi:10.1146/annurev-physiol-021909-135843
Dulamea A. O. (2017). Role of Oligodendrocyte Dysfunction in Demyelination, Remyelination and Neurodegeneration in Multiple Sclerosis. Advances in Experimental Medicine and Biology, 958, 91–127. doi: 10.1007/978-3-319-47861-6_7
Alonso, A. D., Cohen, L. S., Corbo, C., Morozova, V., El-Idrissi, A., Phillips, G., & Kleiman, F. E. (2018). Hyperphosphorylation of tau associates with changes in its function beyond microtubule stability. Frontiers in Cellular Neuroscience, 12. doi:10.3389/fncel.2018.00338
Naseri, N. N., Wang, H., Guo, J., Sharma, M., & Luo, W. (2019). The complexity of tau in Alzheimer's disease. Neuroscience Letters, 705, 183–194. doi:10.1016/j.neulet.2019.04.022
Breijyeh, Z., & Karaman, R. (2020). Comprehensive Review on Alzheimer's Disease: Causes and Treatment. Molecules, 25(24), 5789. doi:10.3390/molecules25245789
Dzamba, D., Harantova, L., Butenko, O., & Anderova, M. (2016). Glial Cells - The Key Elements of Alzheimer's Disease. Current Alzheimer Research, 13(8), 894–911. doi:10.2174/1567205013666160129095924
Yampolskiy, R. (2020, December 14). Can computers evolve to program themselves without programmers? Mind Matters.
Zhong, X., Shi, H., Hou, L., Chen, B., Peng, Q., Chen, X., Wu, Z., Wang, Y., Mai, N., Huang, X., & Ning, Y. (2017). Neuropsychiatric Features of Neurosyphilis: Frequency, Relationship with the Severity of Cognitive Impairment and Comparison with Alzheimer Disease. Dementia and geriatric cognitive disorders, 43(5-6), 308–319. doi: 10.1159/000476060
Bronas, U. G., Puzantian, H., & Hannan, M. (2017). Cognitive Impairment in Chronic Kidney Disease: Vascular Milieu and the Potential Therapeutic Role of Exercise. BioMed Research International, 2017, 2726369. doi:10.1155/2017/2726369
Joshee, P., Wood, A. G., Wood, E. R., & Grunfeld, E. A. (2018). Meta-analysis of cognitive functioning in patients following kidney transplantation. Nephrology, dialysis, transplantation : official publication of the European Dialysis and Transplant Association - European Renal Association, 33(7), 1268–1277. doi: 10.1093/ndt/gfx240
Luo, M., Zeng, Q., Jiang, K., Zhao, Y., Long, Z., Du, Y., Wang, K., & He, G. (2021). Estrogen deficiency exacerbates learning and memory deficits associated with glucose metabolism disorder in APP/PS1 double transgenic female mice. Genes & Diseases, 9(5), 1315–1331. doi:10.1016/j.gendis.2021.01.007
Brueckner, A. L. (1986). Brains in a Vat. The Journal of Philosophy, 83(3), 148–167. doi:10.2307/2026572
Putnam, H. (2009). Brains in a Vat. In S. Bernecker & F. I. Dretske (Eds.), Knowledge: Readings in contemporary epistemology. essay, Oxford University Press.
Stewart, J. R., Gapenne, O., & Di Paolo, E. (2014). Enaction: Toward a new paradigm for cognitive science. MIT Press.