I Hate Entropy — Reflections and Conclusions on Parkinson's Disease
The Place I Arrived at While Trying to Stop Parkinson's
Prologue
This essay is not a medical exposition on Parkinson's disease. It is a record of a non-specialist's attempt to stop Parkinson's, the trajectory of that attempt, and the place that trajectory finally arrived at.
When I first heard of entropy, I thought it was nonsense. The very claim that there is a law dictating the universe must move toward disorder sounded to me like a preemptive justification of human defeat. I still do not accept this law at face value. I only acknowledge that it is a frame that has not yet been refuted at the current level of human knowledge. This essay stands on the word 'yet.'
Parkinson's disease was a subject I concentrated on for a few hours. Having worked my way through Alzheimer's, ADHD, depression, hikikomori, cancer, and aging, I had accumulated a methodology — reframing existing discourse, gathering historical and cross-cultural evidence, proposing mechanisms, suggesting actionable next steps — that had worked to some degree for each disease. I expected the same method to work for Parkinson's.
It did not. This essay is a record of that failure and the one point I arrived at at the end of it.
Part 1 — The Problem of Parkinson's Disease
What happens
Parkinson's disease is a condition in which dopamine-producing cells in a deep brain region called the substantia nigra slowly die. Dopamine acts as a kind of lubricant that makes movement smooth. Every motion — lifting an arm, taking a step, forming a facial expression, writing — rides on this dopamine signal. When the cells die, dopamine supply is cut off, and movement becomes stiff.
Symptoms only appear after 60 to 80 percent of these cells have died. By the time of diagnosis, the disease has already been progressing for 10 to 20 years. This delay is the most fundamental difficulty in treating Parkinson's.
What actually happens inside the cell is this. A protein called alpha-synuclein, whose normal role is to assist dopamine release, is a diligent worker in the healthy state. For reasons not fully understood, some of these proteins fold into the wrong shape and, once misfolded, clump together into sticky aggregates. These aggregates are called Lewy bodies. A Lewy body occupies space inside the cell, paralyzes the cleanup system, damages the mitochondria, and eventually kills the cell.
What is remarkable is that this misfolded protein spreads from one cell to another through synapses. An abnormal protein produced in one cell migrates to a neighboring cell and induces the same misfolding in the normal proteins there. The disease spreads along circuits. This is the shared mechanism of the protein aggregation disease family, which also includes tau in Alzheimer's, TDP-43 in ALS, and huntingtin in Huntington's disease.
What we do not know
From here, I need to be honest. The list of what modern medicine does not know about Parkinson's must be stated plainly, or the rest of the essay cannot stand.
Why does alpha-synuclein begin to misfold in the first place. There has been no answer for sixty years.
Why are the dopamine neurons of the substantia nigra uniquely vulnerable. The body has dopamine neurons in several locations, and the ones in the ventral tegmental area are relatively spared in Parkinson's. Why this differential death occurs is unanswered.
Why does one person develop Parkinson's and another does not, given the same environmental toxin exposure. Unknown.
Why do dead dopamine neurons not return. Adult neurogenesis occurs in certain brain regions but almost not at all in the substantia nigra. The reason for this asymmetry is unknown.
In sixty years, with trillions of dollars in pharmaceutical capital and tens of thousands of researchers, not a single drug has been developed that stops the progression of Parkinson's. In February 2025 the GLP-1 agonist exenatide failed its phase 3 trial. Before that, two first-generation alpha-synuclein antibodies both failed. The development timeline for LRRK2 inhibitors has been pushed to 2031. Attempt after attempt, failure.
I began to suspect that these repeated failures were not accidental but structural. That suspicion is the starting point of this essay.
Part 2 — The Angles I Tried
Attempt 1. Force dopamine in
I started with the simplest idea. If dopamine is in short supply, why not put it in.
This was the point Swedish scientist Arvid Carlsson reached in the 1960s. He won the Nobel Prize for this logic, and the drug he proposed, levodopa, has been the standard of Parkinson's treatment for sixty years. The fact that I independently arrived at the same place is evidence that the direction is natural, but it is also evidence that this direction has already been taken to its end.
Levodopa crosses the blood-brain barrier. It is converted to dopamine inside the brain. It dramatically improves symptoms. But it does not stop the cells from dying. After five to ten years, drug efficacy begins to fluctuate, involuntary movements appear at peak levels, and rigidity returns at trough levels. Eventually a stage is reached where the drug no longer works.
Putting dopamine in is masking symptoms, not stopping the disease.
Attempt 2. Use dopamine sparingly
The next idea was this. If a resource is scarce, use less of it.
This idea soon hit a wall. Dopamine is not a resource that can be saved. Cells have short-term reservoirs called synaptic vesicles, but these are only buffers on the order of seconds to minutes. It is not a structure that accumulates over days the way a reservoir does; it is closer to the relationship between a faucet and a small tank.
More fundamentally, dopamine consumption in substantia nigra neurons is not under voluntary control. It is continuously expended on unconscious micro-adjustments like breathing, standing, swallowing saliva, blinking. There is no way to not use it.
Still, this attempt did not close entirely. If the problem is shifted to a different level, it comes back to life. Substantia nigra neurons fire autonomously throughout life, from birth. Like a pacemaker, they maintain 2 to 10 firings per second even during sleep. Over 80 years of life, a single neuron fires more than one billion times. This baseline firing consumes calcium, produces oxidative stress, and burdens the mitochondria.
In that case the real form of 'use less' becomes 'reduce the cell's own autonomous baseline firing.' This idea was actually tested in a phase 3 trial using the L-type calcium channel blocker isradipine. It was concluded as a failure in 2020. But whether that failure refutes the hypothesis as a whole is under debate. It is possible that the drug had no effect because it was given to patients who had already been diagnosed (with over 60 percent of neurons already dead). The approach of administering it to prodromal patients decades in advance has not yet been attempted. The diagnostic tool for prodromal Parkinson's (alpha-synuclein seed amplification assay) has only become practical around 2023.
Attempt 3. Transfer the method from Alzheimer's
In my Alzheimer's essay I proposed a single approach. Instead of fighting the impossible battle of preserving the patient's entire cognitive ability, target only the narrow circuits that constitute the patient's identity. On top of the neuroscience finding that heavily-used circuits are preserved to the end, I layered five traditional Korean learning methods to arrive at a concrete protocol.
What happens if I apply the same logic to Parkinson's. The hypothesis becomes 'heavily-used motor circuits are more protected against pathology.' If this were true, a lifelong right-hander with Parkinson's would show better preservation of the left substantia nigra (which controls the right hand) than the right. This is a hypothesis verifiable from autopsy data.
There is partial support for this direction. The LSVT BIG protocol in Parkinson's rehabilitation, which concentrates on 'large repeated movements,' meaningfully slows symptom progression. But this is 'maximize the remaining function' rather than an answer to 'does use protect that neuron from pathology.' The latter has not yet been systematically verified.
And there is a structural difference between Alzheimer's and Parkinson's. In Alzheimer's, damage progresses dispersed across synapses, so a strategy of concentrated reinforcement of core circuits is feasible. In Parkinson's, damage concentrates at a single point — the substantia nigra. It is not dispersion but single-point wear. The logic of preserving heavily-used circuits has little room to operate.
The same methodology arrived at different results in different diseases. This difference matters. This attempt taught me that the tools I have do not work with equal depth on all neurodegenerative diseases.
Attempt 4. Refold misfolded protein
In my aging essay I proposed the principle 'repurpose instead of discard.' It was a question directed at the current discourse, which treats misfolded proteins exclusively as objects for disposal. The cystic fibrosis drugs ivacaftor and lumacaftor restored function by refolding misfolded CFTR protein. Does a misfold necessarily require discarding.
Looking at the state of Parkinson's, this principle felt unusually persuasive. Every current alpha-synuclein-targeted drug points in the 'disposal' direction. They use antibodies so immune cells clean up the aggregates, activate the autophagy pathway, or reduce production at the genetic level. The refolding direction has been almost untouched.
And the disposal strategies have failed in succession. This permits the interpretation 'the target is right but the direction is wrong.' If the cause of misfolding keeps operating while the aggregates are being cleaned, new aggregates keep being made. It is like mopping water off the floor while the leak continues.
Refolding drug development lies beyond the current horizon of technology. Designing a chaperone-inducing molecule that returns misfolded alpha-synuclein to its original shape is still at the laboratory level. But the direction is valid. What one can contribute to this direction is not drug design but an essay-level intervention asking 'why has the current discourse concentrated in one direction only.'
Attempt 5. Vaccines and immunotherapy
The idea 'prevent it with a vaccine' arises naturally. Therapeutic vaccines targeting alpha-synuclein (PD01A, UB312, and others) are actually in clinical trials. But all of them are administered to patients who have already been diagnosed. At this stage, even if antibodies clear aggregates, dead neurons do not return.
What would really matter is a prodromal vaccine. Administer it to RBD-positive, alpha-synuclein seed amplification assay positive, or genetically high-risk individuals 10 to 20 years before symptom onset, to prevent the pathology from starting at all. This would be the strategy closest to 'stopping.'
But such a vaccine does not yet exist. Three walls: prodromal diagnostic tools only recently emerged so subject identification has just become possible; the safety bar for immune stimulation in healthy people is extremely high; effect measurement takes decades, which the pharmaceutical industry structure can barely sustain.
The place where these three walls intersect is the real open seat in Parkinson's immunotherapy. Who will fill it, and when, I do not know.
Part 3 — The Wall Called Entropy
Why did all of them fail
Looking back at the five attempts, there was a common thread. Each overlapped with an open seat in current research discourse, but none arrived at 'stopping the disease.' All remained at the level of slowing or bypassing.
The reason became clear. Parkinson's is not an infectious disease. Not a genetic disease. Not an autoimmune disease. Not a cancer. It is a disease in which cells wear out as time passes. There is no external invader, no single gene fault, no immune malfunction, no runaway proliferation. They simply wear out.
This is a qualitatively different category from other diseases.
For infectious diseases you kill or block the invader. For genetic diseases you fix the gene or supplement the product of the mutation. For autoimmune diseases you modulate immunity. For cancer you stop proliferation. Each category has its matching methodology.
Neurodegenerative diseases have no such methodology. 'Stopping wear' does not correspond to any traditional medical tool. For sixty years the pharmaceutical industry has approached Parkinson's the way it approaches infection. Find the 'enemy' — alpha-synuclein — make a 'drug' that attacks this enemy, and run clinical trials. It failed every time.
The real reason for failure is that there is no enemy. The protein aggregates identified as the cause of the disease are more likely to be consequence than cause. The real cause is that the cell operates in a state of lifetime overload and wears out, and this is not an enemy but a condition. It is the kind of problem you cannot kill.
Time, aging, and entropy are one face
When I grasped this structure, something else became clear at the same time. Why does time flow, why do humans age, why does Parkinson's progress — these three questions are actually one.
In physics, the only law that defines the direction of time is the law of increasing entropy. Energy spreads uniformly; order moves toward disorder. The sensation that time is flowing is the way we experience the direction in which the universe's overall disorder is increasing. Asking why time flows and asking why entropy increases are the same question.
Aging is the biological expression of this law. As a cell runs for decades, damage accumulates in its DNA, proteins misfold, mitochondria grow old, waste piles up uncleared. This is not a specific disease but a byproduct of being alive. To stop it, one would have to stop being alive. That is the paradox.
Parkinson's is a locally accelerated expression of this aging. Because substantia nigra dopamine neurons are structurally a high-risk group, they reach their limit earlier than other regions. Alzheimer's is the same thing happening in the hippocampus and cortex. ALS, in motor neurons. Each neurodegenerative disease can be reinterpreted as organ-specific accelerated aging.
If this perspective is correct, the reason Parkinson's research has repeatedly failed becomes sharper. It has been trying to solve the problem at the wrong scale. Parkinson's is not solved by 'Parkinson's drugs.' It is solved when the biology of aging as a whole is solved. The part has to wait for the whole.
And solving the biology of aging means locally reversing the entropy law with respect to living systems. The four open seats I proposed in the aging essay — designing improved mitochondria, brain-directed whole-body repair, functional repurposing of proteins, standardization of intercellular communication protocols — were all proposals of concrete pathways for such reversal. Can they be filled? I think they can, but they will not be filled by my hands.
I hate entropy
At this point I need to be honest. I hate the concept of entropy.
'Hate' here is a matter of cognition, not emotion. The law of entropy is treated as the final form of the physics understanding humanity has arrived at. But looking back at the history of science, laws once considered 'final' have been repeatedly broken by the next generation's breakthroughs. That the Earth is the center of the universe. That species do not change. That time is absolute. Every era's list of impossibilities has become the next era's list of possibilities.
There is no reason entropy should be the exception to this history. Humans are the only lineage that rewrites the set value evolution has assigned to other species, and eventually they will make the same attempt at the set value the universe has assigned.
The probability that this attempt will be made in my generation is low. I do not know how many generations it will take. But the direction exists. My feeling that I hate entropy is not groundless emotion but an intuition that this direction exists.
Part 4 — The Place I Arrived At
Not being able to stop is different from having nothing to do
Not being able to resist entropy does not mean the absence of contribution. Entropy wins in the end. But when the patient loses to entropy can be pushed considerably far apart. Fifteen years of independence after diagnosis and thirty years are not the same disease. Even if the disease cannot be stopped, extending the time during which the patient can live their own life is a meaningful contribution.
In Alzheimer's I was able to take one step. Noticing the asymmetry that entropy progresses circuit by circuit but heavily-used circuits are preserved to the end, I proposed targeting only the narrow circuits that constitute the patient's identity. I adapted five traditional Korean learning methods into an applicable form. I cannot stop entropy itself, but I could propose a structure in which the direction of entropy is consciously chosen, so that the patient can remain themselves for longer.
In Parkinson's, even that did not come out. The structure of damage was different. Alzheimer's dispersed progression model did not fit Parkinson's single-point concentration model. The same methodology arrived at different results in different diseases.
So I arrived at a different place
The conclusion I reached after long thought is this. In the case of Parkinson's, the most practical contribution I can make is not to resist entropy itself but to have external devices bypass the consequences of entropy, through my cervical BCI research.
Since I cannot save the cell, I make a device outside the body do what the cell did. I read the patient's movement intent from the cervical region, and external units distributed across major joints carry out that movement smoothly. I do not touch the brain or the cell.
The concrete structure is as follows. The patient wears a U-shaped neckband. Multiple channels around the cervical region pick up EMG signals. The principle I argued in my earlier cervical BCI essay — that movement intent appears in the cervical muscles 100 to 300 milliseconds before the actual movement — is the basis. This signal is decoded in real time by the smartphone's NPU. A personalized AI model predicts 'what is this patient about to do.' Are they about to raise an arm, stand up, take a step.
The decoded intent is transmitted via Bluetooth to external units on major joints. Ankles, hips, wrists, elbows, neck. Each unit prepares the necessary assistive power 100 to 300 milliseconds in advance. When the patient actually begins to move, the external power is already smoothly joined.
From the patient's perspective, the moment they resolve to move, the body naturally follows. Tremor is stabilized before it starts. Rigidity is loosened before it sets in. Even in a state of dopamine deficiency, movement is smooth.
This is not a cure. It only helps the patient feel less of the damage entropy has caused. Entropy keeps increasing. The moment the device is turned off, the symptoms return as they were. I will not package what is not a fundamental solution as if it were.
What this place means
Even so, there are two reasons I can accept this place.
First, this approach is a new axis not present in current Parkinson's treatment discourse. The existing four axes are drug (levodopa), surgery (deep brain stimulation), rehabilitation (LSVT BIG), and cell transplant (stem cells). All of them attempt to fix the inside of the body. The fifth axis I propose is bypassing from outside. Without touching the brain or the cell, it extends patient independence regardless of disease progression. This axis does not yet exist.
Second, this direction fits the structure of the work I have been building. In ADHD Must Box I wrote 'do not fix the deficit, provide the matching environment.' In the cervical BCI essay I proposed 'reading signals from outside the brain instead of electrodes inside.' In the Alzheimer's essay I argued for 'preserving identity circuits instead of total cognition.' In the aging essay I established the principle 'repurpose instead of discard.' All of these were forms of opening a bypass path from outside, at places where the path of fixing from inside had been blocked.
The same form works in Parkinson's. This conversation has confirmed that my methodology is not a coincidence in a single disease but a consistent axis across the entire territory where entropy wins.
Scope of extension
The structure can build more than Parkinson's. The same technology applies to ALS, spinal cord injury, and post-stroke sequelae. What all these conditions share is the structure 'movement intent is alive but execution is not.' A platform that reads intent from the cervical region and has external units carry out the execution is a common bypass for this common structure.
I designate this as Phase 5 of my cervical BCI research. Phases 1 through 4 aimed at the convenience of healthy users (controlling software with thought alone). Phase 5 aims at the independence recovery of motor disorder patients. On the same technological base, a new application opens. The existing Phase 1 data (directional intent, cervical muscle pre-activation) serves directly as proof of concept for Phase 5. What remains is not additional data collection but the judgment of whether to state this extension possibility in the research plan.
Part 5 — What Remains
The problem of stopping entropy does not remain in my hands. The problem of reversing the flow of time and the problem of stopping aging stand in the same place. Since these three problems are one face, the answer will come from one place. Someone will fill it, someday. The probability that I will be that someone is low.
I already wrote this in the aging essay. Writing this piece, I confirmed it once more.
But in the category of 'humanity,' the story is different. I believe humanity will overcome this problem someday. The intuition I had when I first thought entropy was nonsense — I have not completely abandoned it. The consensus that the current law is the final law is only an intermediate consensus waiting for the next generation's breakthrough.
What remains in my hands is this. In an era where entropy wins, a technology that lets patients live through less painful time. My cervical BCI research can be one such technology. Parkinson's can be the first disease to which it is applied. ALS, spinal cord injury, and post-stroke sequelae come next.
This is the most honest point I arrived at on the subject of Parkinson's. A place where the disease could not be stopped, yet where the time during which the patient can live while feeling less of the disease can be extended. A place where entropy could not be resisted, yet where the consequences of entropy can be bypassed with an external device.
The place I arrived at while trying to stop Parkinson's was here.
If I reduce it to one sentence: why time flows, why humans age, and why Parkinson's progresses are three questions of one face called entropy, and I do not yet know how to stop it. In Alzheimer's I proposed one way to choose entropy's direction and hold on to memory, and in Parkinson's I did not touch entropy itself but arrived at the point where my BCI research bypasses the consequences of entropy through an external device. Entropy is humanity's current limit but will not be its eternal limit. That is all I can currently say.
About the Author
An Seungwon is the sole founder of Wonbrand (wonbrand.co.kr). As an external observer with no formal training in medicine, he has published a consecutive series of essays on medicine, neuroscience, AI, and social issues. This essay is the fifteenth in the series.
Contact: wonbrand.co.kr
References
Braak, H., Del Tredici, K., Rüb, U., de Vos, R. A., Jansen Steur, E. N., & Braak, E. (2003). Staging of brain pathology related to sporadic Parkinson's disease. Neurobiology of Aging, 24(2), 197-211.
Surmeier, D. J., Obeso, J. A., & Halliday, G. M. (2017). Selective neuronal vulnerability in Parkinson disease. Nature Reviews Neuroscience, 18(2), 101-113.
Poewe, W., Seppi, K., Tanner, C. M., Halliday, G. M., Brundin, P., Volkmann, J., Schrag, A. E., & Lang, A. E. (2017). Parkinson disease. Nature Reviews Disease Primers, 3, 17013.
Carlsson, A. (1959). The occurrence, distribution and physiological role of catecholamines in the nervous system. Pharmacological Reviews, 11, 490-493.
Cotzias, G. C., Van Woert, M. H., & Schiffer, L. M. (1967). Aromatic amino acids and modification of parkinsonism. New England Journal of Medicine, 276(7), 374-379.
Vijiaratnam, N., Simuni, T., Bandmann, O., Morris, H. R., & Foltynie, T. (2021). Progress towards therapies for disease modification in Parkinson's disease. Lancet Neurology, 20(7), 559-572.
Vijiaratnam, N., et al. (2025). Exenatide once a week versus placebo as a potential disease-modifying treatment for people with Parkinson's disease in the UK: a phase 3, multicentre, double-blind, parallel-group, randomised, placebo-controlled trial. Lancet, 405(10478), 627-636.
Pagano, G., et al. (2022). Trial of Prasinezumab in Early-Stage Parkinson's Disease. New England Journal of Medicine, 387(5), 421-432.
Lang, A. E., et al. (2022). Trial of Cinpanemab in Early Parkinson's Disease. New England Journal of Medicine, 387(5), 408-420.
Parkinson Study Group STEADY-PD III Investigators. (2020). Isradipine versus placebo in early Parkinson disease: a randomized trial. Annals of Internal Medicine, 172(9), 591-598.
Borghammer, P. (2021). The α-Synuclein Origin and Connectome Model (SOC Model) of Parkinson's Disease. Journal of Parkinson's Disease, 11(2), 455-474.
Fox, C. M., Ramig, L. O., Ciucci, M. R., Sapir, S., McFarland, D. H., & Farley, B. G. (2006). The science and practice of LSVT/LOUD: neural plasticity-principled approach to treating individuals with Parkinson disease and other neurological disorders. Seminars in Speech and Language, 27(4), 283-299.
Ahn, S. (2026). Four Questions About Aging That No One Has Properly Asked Yet. Wonbrand.
Ahn, S. (2026). Alzheimer's: A Person Already Remembers Far Too Much. Wonbrand.
Ahn, S. (2026). What if We Could Control Software with Thought Alone. Wonbrand.
Ahn, S. (2026). ADHD Must Box. Wonbrand.
An Seungwon / Wonbrand / https://wonbrand.co.kr