An Affordable Mass of Cognitive Density
A defense-tech argument for deploying neurotechnology at scale to amplify the performance of power-law talent (from engineers to warfighters).
What follows:
Introduction
Productivity Cost Disguised as Exceptional Output
Building an Affordable Mass of Cognitive Density
Human Neural Network Effects
Appendix: Completed Research and Pipeline
Introduction
National security is powered by human intelligence: the neurological layer beneath the industrial stack. From the engineers at Anduril and Shield AI to the special operators, intelligence professionals, and USSOCOM commanders turning their tools into battlefield advantage. Power-law talent—those capable of generating 10x, 100x or greater outcomes relative to the average contributor in their field—is America’s greatest asymmetrical advantage.
Radical innovation and human performance are the product.
The factory is the product.
The human brain is the factory. It is the machine that builds the machine.
A society with a quarter of the population of its peer adversary must maximize the availability, durability, and performance of its human neural networks.
Motivation is a reward function. Neurotechnology for the brain is a forcing function. Distributed across a power-law organization, neurotechnology can serve as a multiplier of radical innovation and human performance. I’m not referring to brain-computer interfaces in five or ten years from now, but low-cost, non-invasive, scalable neurotechnology that is just beginning to be deployed today in the U.S. Department of Veterans Affairs, first responder organizations, and among active duty service members.
Raising one power-law contributor from 10X to 11X does not merely add one unit of output. Across 100 people on a 200-person hardware engineering team, that gain compounds: more hard problems solved, milestones achieved faster, radical innovation. The difference between incrementally improving a component and discovering the architecture that makes the next platform possible. These are the material “patient outcomes.” For neurotechnology to scale, it must have medical-grade capability and credibility but not be constrained by the practice of medicine. As I will detail below, an affordable mass of neurotechnology can move the power-law ceiling well beyond a 10X-to-11X shift.
To first permeate across power-law teams, whether war-fighting or engineering, and ultimately achieve population-scale, wearable neurotechnology must meet four requirements: 1) it must correct brain dysfunction and sustain correct function, so that healthy sleep, mood, autonomic function and cognitive resilience become durable second order effects; 2) it must be easily self-administered in any environment; 3) it must be prescription-grade but not always require a diagnosis for distribution; and 4) end users must desire the technology without institutional pressure to use it.
Anduril is building an affordable mass of autonomous weapon systems to counter China’s advantage in low-cost, high-volume military production. The national interest requires a parallel effort: an affordable mass of self-administered neurotechnology to deliver cognitive density at scale.
Wearable neurotechnology for the brain is a form of defense-tech.
Productivity Cost Disguised as Exceptional Output
The productivity cost of mood and sleep disorders is well documented across the average workforce, but power-law teams are neither the average workforce nor spared the neurological burden carried by the broader population. Depression costs the average worker a month of productivity per year. 30X power-law talent in AI research can produce in a day that an average researcher may take a month to generate. Like talent output, the productivity cost of neurological burden to a power-law team scales as a power law, but is disguised as exceptional output.
Operator Syndrome has emerged as a pervasive neurological tax, defined as an interrelated constellation of performance-interference impairments including traumatic brain injury (TBI), endocrine dysregulation, chronic pain, and hypervigilance. This syndrome is the inevitable consequence of an extraordinarily high allostatic load—the cumulative wear and tear on the brain’s internal machinery. In a milestone study of 202 treatment-seeking active-duty special operators, researchers found that sleep disruption, mental health disruption, pain, sensory disturbance, and headaches were each present in close to 90% of the cohort. This results in degraded cognition, slowed reaction times, eroded emotional regulation, and a significant increase in accident and injury risk during high-stakes operations. Special operators rarely seek treatment early, let alone before symptoms manifest.
Delayed intervention accelerates a deeper systemic tax felt across the broader veteran community, particularly among the post-9/11 generation who deployed to Iraq and Afghanistan. Data from the Department of Veterans Affairs highlights a baseline of chronic exhaustion, with 57% of post-9/11 veterans meeting the clinical criteria for chronic insomnia. When paired with a 29% lifetime prevalence of post-traumatic stress disorder—where severe sleep disturbances are prevalent in over 90% —it becomes clear that autonomic and neurological wear-and-tear is a structural phenomenon. Private-sector power-law teams are led and populated by veterans across industries and particularly in defense-tech.
Those developing the software and hardware that the armed forces rely on also pay a neurological tax. In peer-reviewed occupational studies of IT and STEM workers, one-third of top U.S. engineering talent reports clinical-grade depressive or anxiety symptoms over a 12-month period. In tech-specific engineering cohorts, the prevalence of insomnia approaches 50%. These rates equal or exceed those recorded in the general population.
Neurological tax is, of course, paid at population-scale. In a survey of 1,010 Gen Z respondents conducted by a leading healthcare IT firm, 46% reported receiving a formal diagnosis of a mental health condition, primarily anxiety and depression, while an additional 37% believed they carry an undiagnosed condition. Only 35% of Gen Z manage to get more than seven hours of sleep per night.
The most useful brain neuromodulation technology may not be designed to treat anything. Instead, its goal may be to restore and maintain the function of the brain’s integrated macro-limbic network (aka, the limbic-autonomic axis) that governs mood, sleep, autonomic function, and cognitive resilience. This core architecture spans the anterior cingulate cortex, insula, amygdala, thalamus, hippocampus, and upper brainstem. If neuromodulation technology succeeds at restoring and preserving the performance of these structures, the treatment of sleep and mood and the delivery of human performance become inevitable second order effects.
Building Affordable Mass of Cognitive Density
Since the launch of DARPA’s Augmented Cognition program in 2001 and the Human Assisted Neural Devices (HAND) program in 2002, the military has spent a quarter-century seeking ways to optimize the human brain for combat. What began as an effort to monitor cognitive workload and restore lost function evolved into a search for direct brain stimulation technologies designed to sustain vigilance, rapid target detection, working memory, and decision-making under stress. These programs framed the challenge as a neural tuning problem, gently biasing the readiness of the neocortex so the network stays online longer and learns faster. No scalable solution emerged.
In the mid 2010’s, a handful of VC-backed startups took a similar approach to consumerizing wearable brain stimulation. They targeted the brain’s last mile of human performance: the prefrontal cortex and motor cortex. Avoiding the burden of clinical research and medical device regulation, they rapidly entered the market with wellness claims. One version, an electrode attached to the side of the forehead, was marketed as an alternative to coffee. Another was positioned as a means of incrementally improving athletic performance. The absence of research and market traction created a lasting perception that the category lacked both credibility and scalability. Institutional investment in the sector ground to a halt.
The prefrontal cortex and motor cortex are integral to human performance, but they rely on the performance of the macro-limbic network that creates the conditions for performance in the first place: healthy sleep and mood, robust heart rate variability, and—indispensable for any high-intensity profession—sustained cognitive resilience.
My co-founder and I have quietly built a commercial neurotech venture to develop and validate wearable brain stimulation that engages the foundational macro-limbic network. Our unique stimulation parameters are calibrated to simultaneously down-regulate hyperactive circuits and elevate activity in underperforming regions, leveraging neuroplastic consolidation to drive durable changes in neural circuitry and leave the user in an unburdened, high-focus state. A research team at Harvard Medical School and Massachusetts General Hospital recently validated this architecture by mapping our induced electric fields within the macro-limbic network, demonstrating a peak field strength up to 3X the minimum threshold required to trigger neuromodulation, while maintaining an exceptional safety profile and user comfort.
Over the past 17 years, nearly 2,000 subjects have been enrolled in clinical trials and real-world studies using our technology. These studies document rapid effectiveness in treating anxiety, depression, insomnia, PTSD, the depressive aspect of bipolar disorder, symptoms of Parkinson’s, and increasing the retention of patients in residential drug and alcohol rehab programs.
The defense-tech industry understands the value of proving new technology in Ukraine far better than traditional medtech. In 2022, fewer than 20 of our devices were deployed across three Ukrainian hospitals and used to treat hundreds of soldiers and civilians with PTSD and acute stress-related disorders. One device can operate around the clock, treating a new user every 20 minutes. Patients who received stimulation recovered far faster than those receiving standard care alone. The detailed results, drafted by a Ukrainian physician, are included here; the English is imperfect, but the signal is clear.
Our most recent anxiety and stress study, conducted in collaboration with the Seattle Police Department, enrolled 165 first responders from around the country, including firefighters, EMTs, and 911 dispatchers. It documented rapid and durable symptom reduction with no serious side effects (pending publication). Our most recent depression study, published in The Journal of Clinical Psychiatry, shows that our standard dose device reduced moderate to severe depression in women with a rapidity typically seen only with controlled substances like ketamine, or ECT, but without the side effects. We are running a trial with our higher dose device next, at Massachusetts General Hospital, to treat men with larger heads who require a slightly higher electrical dose to achieve the same field strength as our standard device does in average heads (most women and some men). Mass General Brigham’s Enterprise Chair of Psychiatry, Dr. Maurizio Fava, designed and will lead this follow-up study, and now serves as our advisor.
The U.S. Department of Veterans Affairs is currently running a feasibility study with our device, and a PTSD treatment center serving active-duty special operations personnel is in the process of joining as a second site.
On the human performance front, a leading eSports team has deployed 50 devices to restore autonomic function in their athletes after intense periods of competitive gaming, tracked through EKG and EEG over the past 18 months.
These results have been derived from a device consisting of little more than a printed circuit board and two electrodes. No EEG sensor stacks. No Bluetooth—no pairing requirement, no app dependency, no cloud telemetry. That is not a missing feature. In forward environments, it is a feature.
As a result of temporary FDA clearance that allowed us to distribute our proof-of-concept technology before clinical trials were complete, 60,000 individuals still use the device today—most for more than five years, and many for over a decade to maintain mental health and cognitive performance. Despite looking like it came from RadioShack in the 1980s, more than 70% of surveyed users consider their v1 device to be as important as, or more important than, their smartphone.
The documented use cases have radical scope. Logan Shield, a Marine combat veteran who served in Iraq and Afghanistan, has publicly stated that the technology ended years of chronic suicidality following two suicide attempts. Meanwhile, music icon Rick Rubin has said he uses the device as part of his daily wellness practice.
Our first FDA indication for our next-gen technology is expected this September for the treatment of anxiety. Next year, we are pursuing indications for depression and insomnia.
Those three indications matter because they change what the final clearance means.
Next year, we are also pursuing a novel FDA clearance—called a De Novo clearance—to support autonomic and cognitive performance in healthy individuals. On its own, that would make this a performance gadget. But coming after anxiety, depression, and insomnia, it becomes something very different: a prescription-grade human performance platform that happens to be approved to treat the three most common neuropsychiatric conditions.
That sequence is fundamental. The performance indication opens the door for healthy individuals, but it’s the clinical indications that give the device a level of credibility, utility, and relevance that ordinary performance tools cannot match.
It is this combination that places the technology in a category of human performance utility that has never existed.
There are many professional cultures in which performance, not the treatment of illness, is the only door to run through.
It took years to refine the electrical dose, electrode placement, and usage protocol. Informed by data, we partnered with the design and engineering teams behind Beats and Nest to create a new form factor, called OAK. Unlimited color and fabric options. With capital in place, we will produce prototypes within six months.
OAK in Operator Black
OAK in Ocean
OAK in Tartan
OAK electrode pod
In our conversations with NIH, the clear message was that federally funded brain-stimulation research needs to individualize treatment through sensor stacks. This sounds good on paper but in reality, EEG and other low-cost sensors capture more noise than signal and struggle with depth perception, placing the technological focus on reading the brain instead of changing the brain. We took the change path.
Our unorthodox approach and timeline were not a natural fit for traditional medtech investors, so we turned to our proof-of-concept user and prescriber network, which has collectively invested over $11 million via equity crowdfunding since 2019. Combined with v1 device revenue, founder life-savings, and a private seed investment, we’ve managed to finance the R&D, operations, and regulatory timelines required to get within striking distance of bringing OAK to market.
We’ve managed to enjoy ourselves while eating glass and staring into the abyss. Over the past couple of years, we’ve had the privilege of adding three exceptional advisors to our team. In addition to Dr. Fava, we are advised by his colleague at Mass General Brigham’s and Harvard, Dr. Joan Camprodon, Chief of Neuromodulation and Neuroimaging, as well as Dr. David Shulkin, the former U.S. Secretary of Veterans Affairs.
Three advisors, three fulltime employees. The only way we could get this done is to be lean. Maybe a little too lean. I look forward to expanding the team. We’re just getting started.
Human Neural Network Effects
Mission success is not an individual event. It is a system-level outcome. The warfighter is one node. The engineer building the warfighter’s technology is another. Their spouses are nodes as well. So are their kids. If one node is burdened, the whole system is burdened. Optimizing human performance in the national interest means optimizing the full human network: the people who build the tools, the people who use them, the families that help keep the talent operational. One of the most powerful capabilities of wearable neuromodulation is its ability to be shared.
We have a few hundred devices remaining that are earmarked for power-law pilot study deployment prior to FDA clearance. Improving total sleep time per night is enough for a power-law ceiling to rise.
We’re almost always running an equity crowdfunding campaign. Our incredible customers have seen us through, and I want to pay them back many times over.
We’re likely just a few months away from commercial market access.
We view the VA and the DoD/DoW as foundational customers that will enable us to expand into enterprise and consumer channels and take the time required to obtain broad commercial insurance reimbursement in order to achieve population-scale. We will do this by showing employers of power-law teams that our technology walks the walk.
American dynamism is powered by human intelligence: the neurological layer beneath the industrial stack.
We must remain kinetic.
Appendix: Completed Research and Pipeline
Mechanism of Action and Clinical Data
Our v1 device delivered variable output controlled by the user and circuit-board-level configuration. Our v2 device replaces that with a fixed, integrated architecture: output parameters, electric-field characteristics, treatment cadence, and electrode montage all function as one interdependent system, validated by years of clinical evidence, modeling, and real-world-tested protocol optimization.
Our unconventional parameters are designed to bypass skull resistance, with a waveform more akin to electroconvulsive therapy—square, not sine wave, with a microsecond-speed pulse rate—at orders of magnitude lower dose. Because larger heads create more physical resistance, some users—predominantly men—require a slightly higher electrical dose to achieve the same field strength the standard dose produces in average-sized heads. This matters most when engaging deep, subcortical structures, such as those implicated in depression. We nailed standard dose effectiveness among average sized heads—predominantly women— in our published depression study. Our next depression study, designed and led by Dr. Maurizio Fava at Massachusetts General Hospital, will deploy standard- and higher-dose devices based on hat size, enroll hundreds of participants, and track them for months.
Electrical Field Model (pending publication)
Finite-element modeling conducted at Harvard Medical School and Massachusetts General Hospital demonstrates that our standard dose v2 technology produces clinically significant electric field strengths across distributed limbic–cortical and subcortical circuits, including: anterior cingulate cortex, amygdala, insula, medial prefrontal cortex, thalamus, hippocampus, dorsolateral prefrontal cortex, and the upper brainstem. A similar field model for large heads is currently underway.
Anxiety, Stress and Resilience in First Responders (pending publication)
In a randomized, waitlist-controlled trial of 164 first responders, immediate active treatment with our v2 device produced significant, rapid anxiety reduction (within 2 weeks). Benefits converged after both groups received stimulation, and remained durable through eight weeks of use, with only mild, self-limiting side effects and no serious adverse events.
Note: This study serves as the primary evidence, among a larger totality of evidence, in support of our standard dose v2 FDA clearance expected in the next few months. Our application benefits from our v1 device serving as “predicate.” The threshold for FDA clearance is demonstrating that the submitted device is “substantially equivalent” to the predicate device with regard to safety, effectiveness and technological characteristics. Our regulatory team, which includes former FDA reviewers, believes we check these boxes.
Acute Treatment of Major Depressive Disorder (published)
The triple-blind, randomized controlled trial enrolled 185 women and 70 men. Females in the active arm achieved a significantly greater improvement in depression versus sham at Week 1 (p<0.001) and a 73.9% responder rate by Week 4. This is comparable to the antidepressant effects of ketamine and ECT. 87.6% of subjects receiving active treatment experienced at least an 18% improvement from baseline by endpoint. Results in male participants were mixed, and a follow up study will deploy the higher dose device based on head/hat size.
Depression in Patients with Bipolar II Disorder - Pilot Study (published)
In a randomized, double-blind, sham-controlled pilot study of 16 outpatients with Bipolar II depression, conducted at Beth Israel Medical Center (now Mount Sinai Beth Israel) two weeks of once-daily 20-minute stimulation produced a 13.01-point BDI reduction in the active group versus 3.76 points in sham (p=0.016). Active treatment also improved clinician-rated illness severity with no evidence of mood switching to mania or hypomania and was well-tolerated.
Chronic Insomnia (exploratory)
In a randomized, sham-controlled, fully remote trial conducted during the height of the pandemic (in Fall/Winter 2020) that utilized clinical-grade sleep trackers for objective measurement, our v2 standard dose produced a significant 42.4-minute between-group improvement in total sleep time by Week 4 (p=0.031) in the subset of participants with complete sleep data (n=46). Contingent on funding, we have a large-scale pivotal trial planned in collaboration with sleep researchers at the University of Pennsylvania.
PTSD
Real-world comparative study conducted across multiple Ukrainian medical centers during the Ukraine-Russia war. Among 486 patients with acute stress-related disorders, adjunctive use of our v2 (n=296) dramatically accelerated recovery versus standard care alone (n=190). Sleep-wake cycle normalization occurred in 89.3% of the device group versus 31.5% with standard care alone; Levin Sleep Scale results were similarly superior (75.1% vs. 34.6%), with broad improvements in anxiety and depression.
Parkinson’s Disease Pilot Study (published)
In a pilot study of 10 Stage 3 Hoehn and Yahr Parkinson’s patients with freezing of gait, conducted at the University of Maryland, a single 20-minute session produced immediate, statistically significant motor improvements: stride length increased by 9.1 cm and gait velocity rose by 0.085 m/s. Dynamic stability also improved significantly (fewer steps required to recover balance after backward perturbation, p=0.046).
Substance Use Rehabilitation
In a retrospective chart review of 392 adults in residential substance-use treatment, patients who used our device for approximately three weeks showed significantly higher retention: 76.0% remained in treatment at 90 days versus 51.7% without the device (47% relative improvement). Device use was associated with a 60% reduction in dropout risk (Cox model, p=0.003), with 64.6% still in treatment at an average of 149 days versus 44.4% in the non-device group.
Ongoing Research
US Dept. of Veterans Affairs; 40-subject, 8-week feasibility pilot evaluating anxiety and cognitive function across a broad clinical spectrum, including patients with PTSD
Active duty and civilian PTSD treatment site in process of joining as second site
Real-world eSports performance study assessing autonomic recovery (e.g., HRV) and competitive gaming performance (confidential site)
Planned Research
12-week, 402 subject RCT for the treatment of major depressive disorder, with cognitive performance as secondary endpoint; protocol includes Sequential Parallel Comparison Design (SPCD) and SAFER interview
8-week, 200-subject RCT for chronic insomnia, with cognitive performance secondary endpoint
Real-world data collection supporting cognitive performance and autonomic function