Neurohacker, systems scientist, and philosopher Daniel Schmachtenberger discusses the questions at the heart of modern science which will allow us to unlock the secrets of the most complex instrument in the known universe: the human brain, and offers avenues to begin quantifying and hacking it.
Episode Summary
On a recent episode of my Head First podcast, I sat down with a systems scientist and co-founder of a neurohacking research group to talk through the hard problems at the edge of brain optimization. The conversation ran wide, from the philosophy of consciousness to the chemistry of acetylcholine. I want to pull out the parts that change how you should actually approach your own brain.
What does "neurohacking" actually mean?
My guest defined neurohacking as anything that optimizes the mind-brain interface and the function it produces. That covers cognition, mood, psychomotor skill, and general well-being. The framing I found useful was "sovereignty optimization," your capacity to make good adaptive choices.
That capacity has two halves. There is the input side, your sensory intake and information processing. There is the output side, your ability to act, which runs on impulse control, executive function, and emotional resilience. When difficulty hits, a resilient nervous system processes it and keeps learning. A depleted one shuts down.
The hardware and software are not separate systems. Your physiology shapes your subjective experience, and what you focus your attention on changes your physiology back. Both keep changing each other through experience-dependent plasticity. If you want a deeper read on this loop, I cover it in Biohacking Plasticity.
Why does the same symptom have completely different causes?
Take anxiety. When the research describes anxiety, the presenting complaint is the same word for very different brains. My guest laid out the causal possibilities clearly, and they match what I see in the brain maps.
Anxiety can be physiogenic. A gut microbiome disrupted by a parasite picked up traveling can drop GABA and serotonin production in the gut and drive inflammation along the enteric nervous system. A minor head injury that a CT or MRI misses can still show up on a QEEG as a measurable shift in the EEG. Mold exposure, methylation differences, and genetic predispositions all sit on this side.
Anxiety can also be psychogenic. Early childhood attachment patterns produce a different neurological and psychological picture than acute adult trauma. The approaches that help are different too.
For one real person, anxiety is often a confluence, a causal cascade where several of these factors stacked over time. This is why anxiety is better understood as a cluster of symptoms with multiple possible roots than as a single disease with a single fix. I break down the circuits involved in Biohacking Anxiety, and the QEEG side in Neurofeedback for Anxiety: What the Research Shows.
Is personalized brain medicine even computable?
I work with brain maps every day, and I am humbled by how little any of us fully understand. My QEEG gives me one slice. It captures less than half the EEG, only the dipoles oriented to the skull, averaged across several minutes. That means it emphasizes traits and largely misses states. I look at roughly 30 to 40 features and read those. It is a powerful tool with hard limits.
The reason no single tool solves the whole problem is mathematical. The total information in a human system exceeds conscious working memory by a wide margin, which is itself the tightest bottleneck in human performance. The genome operates through combinatorial dynamics across tens or hundreds of thousands of interactions, not single SNPs. Layer on the epigenome, the transcriptome, the proteome, and the microbiome, most of which is non-human DNA, and you reach what computer science calls an NP-complete problem. The system is also chaotic, so any model projected forward in time diverges from reality.
The practical takeaway is not despair. Brute-force big data alone will not crack this. You have to understand the dynamics of causation in a complex adaptive system well enough to model what actually matters. That is the modeling work worth doing.
What is "homeo-dynamics" and why does aging fit it?
My guest preferred "homeo-dynamics" over homeostasis, and the distinction is sharp. Life is not stasis. You hold a dynamic range across many axes, and health is the capacity and resilience to stay within those ranges when stressors push you out.
Insulin makes this concrete. Insulin rises when you eat sugar, cells take up glucose, insulin falls. Push that system to the top of its range chronically and it stops varying. On the way there you get insulin resistance, fasting insulin climbs, and blood sugar swings further out of balance, which cascades into every system that depends on it. The same pattern shows up in the brain as neuronal insulin resistance, sometimes called type 3 diabetes, where brain cells cannot use glucose well despite adequate blood levels (de la Monte & Wands, 2008).
Here is the reframe worth holding. Aging is not disease. Disease is when a metric leaves its effective range and a pathophysiologic cascade follows. Aging is the decreased capacity to absorb a stressor and stay in range, which raises susceptibility to disease. Someone can have every marker inside the healthy band and still be fragile, less able to sit in the healthy basin of the phase space when stress arrives. For more on when this fragility accelerates, see The Critical Aging Window.
Can the brain regenerate what it lost?
The old dogma said you do not make new neurons and that neurogenesis happens in only a few regions. That is wrong. We now find neurogenesis in the adult hippocampus (Eriksson et al., 1998), and we are learning how to support pluripotent stem cell production, differentiation, neuroprotection, and synaptogenesis.
A self-generating system that built its own tissue should be able to rebuild it. The capacity often drops or switches off for various reasons, and those switches are likely modifiable. The same logic applies to clearing senescent structures and protecting cells from becoming senescent in the first place.
There is a catch the gerontologists name antagonistic pleiotropy (Williams, 1957). A mechanism that protects you when young can harm you when old. Telomere shortening blocks runaway replication and cancer in youth, then in age it stops fibroblasts from clearing themselves and creates a pro-cancer environment. You cannot simply reach in and switch one thing on. Crank telomerase and you risk cancer.
Why does targeting one variable rarely work?
This is the daily frustration of my work. Neurofeedback is not me dialing a neurotransmitter up or down. When you perturb a system, the system reorganizes, and the response is hard to fully predict. The honest move is to raise system resilience rather than chase a single target.
So at the cellular level the general supports matter. Improving the NAD+ to NADH ratio, supporting mitochondrial biogenesis, raising ATP production efficiency, upregulating the Krebs cycle, and protecting mitochondria from senescence all help the complex work cells do go better. These are broad supports with broad effects.
This is exactly how I frame neurofeedback for people. We build more resources across the regulatory domains of sleep, stress, and attention first. That is resilience. Get someone sleeping well and thinking flexibly, and the specific complaint they came in for often sorts itself out along the way. Then, if needed, the coaching targets the remaining piece. The interconnection is also why medical specialties have struggled with complex illness. A structural knee problem can drive chronic inflammation, and those cytokines cross the blood-brain barrier to affect neural circuits, while the pain itself feeds sympathetic load into the brain. If you want the sleep foundation first, start with Biohacking Sleep and the SMR Neurofeedback protocol.
What separates a real nootropic from a smart drug?
We spent real time on definitions, because the field is sloppy with them. Three categories matter.
A brain nutrient is something you would normally get from diet that supports neural function and often sits at suboptimal levels. Tyrosine, the amino acids, omega-3 fatty acids, B vitamins, and vitamin D fall here. There is a large gap between medical deficiency, where acute pathology appears, and optimality. Plenty of people sit in that subclinical gap, including vitamin D deficiency in Southern California. Stress and dietary restriction burn through B vitamins faster, which is why a vegan or a high-stress life can need more.
A nootropic, by the original definition, is a compound that improves some aspect of cognition above baseline without meaningful negative side effects (Giurgea & Salama, 1977). That last clause is the whole point.
A smart drug is the override. This is where modafinil, Adderall, and methamphetamine live. They have real effects and real costs.
The original definition matters because it is a hill I am happy to die on. Modafinil with a significant side-effect profile is not a nootropic. A presynaptic dopamine agonist like Adderall can raise focus while simultaneously decreasing working memory, which is not the total intelligence anyone actually wants, and it brings anxiety, irritability, and depersonalization. Because it overrides an internal regulatory system, it also erodes regulatory independence and produces a comedown.
When is the risk worth it?
Western medicine runs on a risk-benefit ratio built for disease, where the risk of not treating justifies some treatment risk. Optimization above baseline flips that math. If your function is already good, why accept real side effects for a small supra-baseline gain? Trying to remedy a genuine deficit can justify more risk. Chasing a marginal boost on the chance it suits your chemistry does not.
The honest exception is citizen science. Some frontiers cannot run through formal trials for ethical reasons, and dedicated self-experimenters advance the edge. The useful response is to give those people better quantified-self tools and then aggregate what they learn. If you want the evidence-based framing of cognitive gains, I cover it in Biohacking Intelligence.
How do you build a stack that respects the whole system?
The design principle my guest described for their formulation is the opposite of single-molecule override. Map the full regulatory process for a target, address the rate-limiting steps, and aim for a broad set of positive effects without the downregulation and crash.
Acetylcholine is the clearest example. You can supply racetams and choline donors, but the donors differ. Citicoline acts more centrally, alpha-GPC more peripherally. Run the conversion path from uridine to CDP-choline to phosphatidylcholine to acetylcholine and you get different peak plasma times, so combining several smooths the curve instead of spiking it. You also have to ask whether acetyl groups are rate-limiting, whether B5 is limiting the acetylation, whether you need an acetylcholinesterase consideration, and how much NMDA upregulation the acetylcholine load can support before cellular energy becomes the bottleneck.
Then glutamate. Many racetams have AMPA-modulating, or ampakine-like, properties. Noopept gets called racetam-like and ampakine-like even though its structure is neither, since it lacks the pyrrolidone ring. It upregulates acetylcholine and glutamate uptake at the relevant complexes, acts as a precursor to cycloprolylglycine, and is neuroprotective and anti-excitotoxic by clearing excess glutamate from the synapse. The thinking here is systems-level, not single-target. Psychedelic microdosing has interesting effects but sits outside the legal frame for a formulated product, so the constraint there is regulatory rather than physiological.
What I want you to take from this
Optimization is not a matter of finding the one lever. The brain is a self-organizing regulatory system embedded in a body that is itself a web of regulatory loops. Build resilience across sleep, stress, and attention first, identify and reverse the specific causes of dysregulation where you can find them, and treat anything with a real side-effect profile as a treatment, not a daily upgrade.
If you want a starting point this week, pick one regulatory domain and measure it. Track your sleep for two weeks before you change anything, or get a QEEG brain map to see your actual traits rather than guessing at them. The map tells you where the leverage is.
References
- Monte (2008). Thctf1 transcription factor of Trichoderma harzianum is involved in 6-pentyl-2H-pyran-2-one production and antifungal activity. doi:10.1016/j.fgb.2008.10.008
- Eriksson (1998). Development of the histaminergic neurons and expression of histidine decarboxylase mRNA in the zebrafish brain in the absence of all peripheral histaminergic systems. doi:10.1046/j.1460-9568.1998.00394.x
- Williams (1957). Brucella suis infection of a child in Eire. doi:10.1016/s0140-6736(57)90165-4
- Giurgea (1977). [Epidemiological study of chronic cough and obstructive ventilatory dysfunction in a rural area]. PMID 204972