Mass
~1.35 kg (3 lb)
Technical Datasheet · Rev. B
You're overwhelmed. You can't get a human on the phone anymore. And the AI everyone bolted onto everything answers confidently — and wrong.
You don't have time to think.
I do.
Not a chatbot. Not a junior you'll catch scrolling their phone. A single human mind that will sit with your hardest problem for hours — and reason all the way to the bottom of it.
By design, since 2015. I built deliberate metacognitive space for exactly this work. Over the years multiple clients have hired me for a full day to do nothing but think about their project — and somewhere in there I realized this is the thing the world is actually missing. Not more IT skill (mine runs deep); more thinking. When I lock onto one problem for an entire day — cross-referencing with AI as an instrument, never a crutch — that is a genuinely powerful process.
Organic Compute Module — Human Specs
UNIT: 1 × Homo sapiens
Throughputon the order of one exaflop
~1018 FLOPS[1]
Power draworganic substrate, at rest
~20 W[1]
Coolant4–5 tall glasses of water
~1 L / day[8]
These are your specifications — the machine reading this page is hosted in the skull behind your eyes.
The problem nobody will say out loud
A real, serious gap in the marketplace — dressed up as a datasheet.
The largest companies on earth have removed the human from the loop. You call, and you reach a script, a queue, or a model answering with total confidence and zero understanding. When the problem is genuinely hard — the kind that needs sustained, structured reasoning — there is increasingly no one home.
Meanwhile you're buried. The calendar is full, the inbox is louder, and the one thing you cannot manufacture is uninterrupted time to think well about the thing that actually matters.
So here is the offer, stated plainly: you bring the problem; I bring the hours, the focus, and 2,500 years of logic standing behind the method. Silicon executes. The organic substrate decides what is worth executing. You are not buying raw operations — you're buying judgment under ambiguity, the one spec no machine below can quote.
Who you're actually hiring
No theatrics. Just an honest account of the instrument.
I am, as near as humanity gets, an organic computer — and I mean that literally.
Architecture
Wired for depth. Autistic, with Tourette's. My mind locks onto a single problem and does not drift. What reads as unusual is exactly the trait that makes the work possible.
Duty cycle
No scrolling. No earbuds. No hangover. I'm not the assistant you catch half-listening to a podcast. I don't drink, I don't party, I sleep every night, I drink my water — and I work in silence. My favorite song is silence. The work gets a rested, sober, undistracted mind — every time.
Clean signal
Information from the source — in engineered quiet. No doom-feed or cable-news narrative polluting my inputs, and no music competing for the channel. I'm not sound-averse — the opposite: so acoustically precise that music can't stay background, it becomes a signal I'm compelled to decode. So I think in a room built for it — steady fan, white noise, concrete walls. Garbage priors produce garbage conclusions; I go to primary evidence and reason from there.
Operating discipline
AI as instrument, not author. I use AI to cross-reference and stress-test — with human logic always driving. That's the difference between a tool and a crutch.
The honest deal: you'll never catch me phoning it in, because this isn't a side hustle — it's how I'm built. I don't have a feed to check or a hangover to nurse. When I take your problem, you get the whole apparatus — rested, hydrated, undistracted — pointed at it for as long as the problem takes.
Who I am
The origin of the discipline — and the turn that made it.
I'm Carey Balboa, founder of IT Help San Diego. I've been solving technology problems for 27 years — for entertainment, medical, and legal clients, and for the PhDs who needed someone who could actually go to the bottom of a problem.
It started in a closet in Memphis around 1980: my father John — a Navy veteran and career technologist — and me, building a computer from Radio Shack parts. The fascination never stopped. Two decades of deep IT work in Nashville followed, then an offensive-security era spent learning how systems actually fail, not how they're advertised to work.
Then, around 2015, the question changed. It stopped being “how do you break in?” and became “why does this keep failing — and why won't anyone slow down enough to think about it?” I made a deliberate turn: less social media, less noise, fewer wrong sources polluting a clear mind — and more sleep, more water, more uninterrupted thought. I gave up the things that fracture attention and built my life around protecting it instead.
There's an older root to this. As a kid I loved math — the real kind, the kind that makes things happen — but I was handed “just memorize it, just trust us” instead of “here's why it works.” I wouldn't accept answers I couldn't see the foundations of, and that refusal cost me in school. It also became my method. To this day I don't memorize what changes — I look it up, every time, in front of the client, from the highest-authority source — because the honest move is to verify, not to trust a memory, including my own. I can quote you Cat 8 at 2000 MHz from recall, then look it up anyway, and cut the run shorter than the spec allows because I know what copper actually does. That's not insecurity; it's where I choose to put the logic.
That discipline produced the evidence-first confidence engine behind my DNS Tool — and it's the same discipline I'm offering directly here. Over the years multiple clients have hired me to do nothing but think about their problem for a day. Somewhere in there I realized: this — not more tooling — is what the world is actually starving for. Organic Computer is that realization, made available.
The organic compute module
One unit. Carbon substrate. Here is its datasheet.
| Parameter | Value | Basis |
|---|---|---|
| Peak throughput | ~1×10¹⁸ FLOPS (spiking-network level) | Processing to emulate the brain at a spiking-neural-network level of detail.[2] NIST independently frames the brain as "the equivalent of an exaflop."[1] |
| Functional range | ~9×10¹⁵ – 3.4×10¹⁷ FLOPS | Derived from estimated brain TEPS at a computer-like FLOPS:TEPS ratio.[3] Conservative bound < 1 PFLOPS.[4] |
| Power envelope organic | ~20 W | Whole-organ draw at rest — roughly a dim lightbulb.[1] |
| Energy efficiency | ~5×10¹⁶ FLOPS / J | Derived: ~1 EFLOPS ÷ 20 W. ~10⁶× better than today's best silicon at exaflop scale.[1][9] |
| Decision latency | ~100 ms | Perception-to-action loop in the visual system — sub-second response.[1] |
| Architecture | Massively parallel, sparse, event-driven | ~10¹¹ units concurrent; activity encoded in spike timing, not clock cycles.[1][2] |
| Fault tolerance | Graceful degradation | Continuous self-re-weighting via synaptic plasticity; no single point of failure.[6] |
| Coolant intake organic | ~1 L / day (4–5 tall glasses) | Beverage portion of daily hydration; food supplies the rest.[8] |
The numbers, drawn to scale
Because a spec sheet should show its work.
Emulation cost ladder
// FLOPS required to simulate the brain at increasing fidelity — log₁₀ scale — Sandberg & Bostrom 2008[2]
what it functionally does
spiking-network emulation
full biochemical detail
Power to reach exaflop scale
// organic substrate vs. Frontier-class supercomputer — the real gap is six orders of magnitude
Attention span, drawn to scale
// sustained focus on a single screen — the market is shrinking; the module runs the other way
In under two decades, the average attention span on any screen fell from 2.5 minutes to 47 seconds — a 97% collapse.[10] Under a genuine tech crisis, when someone can't even reach their own email, it falls further still. Honest note on the scale: the two human bars are drawn to scale with each other (47 s is ~16% of 2.5 min). The module bar is not — and deliberately so. A single sustained focus block of ~12 hours is roughly 290× the 2004 bar and ~900× today's 47 seconds; at true scale it would run off the page by a factor of nine hundred, so the bar breaks its axis rather than fake a tidy ratio. This module is built to hold a single problem for hours — sustained partly by the same water on the spec sheet: even mild dehydration measurably degrades vigilance and working memory.[11]
Buy the silicon, or engage the organic one
Matched on throughput. Not matched on anything else.
Carbon · organic
Organic Compute Module
Throughput~10¹⁸ FLOPS
Power~20 W
Footprintone chair
Coolinga glass of water
Understands the problemyes
"The answer to a question lives in the foundations of the question. To understand those foundations — that is logic."
Sticker price of the equivalent machine
$600,000,000
What the U.S. Department of Energy paid Oak Ridge / HPE Cray to build Frontier — the closest silicon match to one organic compute module.[9]
Or engage the organic one for an afternoon.
2,500 years of standing behind the method
Organic compute is not new. It is the oldest computer there is.
Logic came before the abacus. Reason came before the transistor. Every machine you can buy is a recent footnote to a much longer line of human thought — and that lineage is the warranty on this work.
c. 450 BCE
The Socratic method
Knowledge pursued by disciplined questioning — the original debugging protocol for a human mind.
c. 300 BCE
Euclid & Aristotle
Formal proof and the syllogism: reasoning becomes a system with rules, not just opinion.
1847
Boole's algebra of thought
Logic reduced to 0 and 1 — the bridge that would one day let silicon imitate a sliver of what minds do.[1]
1943
McCulloch & Pitts
The first mathematical model of a neuron — the seed of every neural network and "AI" that followed.[1]
Today
The organic module, on purpose
The whole 2,500-year stack, running on one disciplined mind — using the machines as instruments, not replacements.
The Verification Principle
Understanding a question's foundations — stated as math.
The maxim above isn't a slogan; it's a theorem. A question's foundation is the belief you start with — the prior. And there's a hard rule about priors: if you begin in absolute certainty, no evidence on earth can move you. Honest reasoning lives strictly between the extremes, and updates with the evidence. This is the same engine that scores findings on the DNS Tool — the principle is domain-agnostic.
Dogmatic prior · trust
Blind trust
Accept it at face value. If it parses, call it fine. Ignore context, ignore drift.
P(H) = 1 ⇒ P(H | E) = 1
Evidence-weighted · verify
Understand the foundations
Start from an honest prior — bounded away from 0 and 1 — then move with the evidence. This is logic.
posterior odds=
P(E | H)P(E | ¬H)
×
P(H)P(¬H)
Dogmatic prior · accusation
Blind accusation
Flag everything as dangerous. Inflate severity. Read malice into ordinary mistakes.
P(H) = 0 ⇒ P(H | E) = 0
The core theorem — why dogma cannot learn
P(E) > 0, P(H) ∈ {0, 1} ⇒ P(H | E) = P(H)
In plain language: a belief pinned at certain or impossible is unrevisable — evidence simply bounces off. A closed mind isn't a character flaw here; it's a mathematical fact about a prior set to 0 or 1.
This is why "understand the foundations" is logic: to set an honest prior and let evidence do its work is the whole discipline. It holds for a DNS record, an argument between friends, red vs. blue politics — anywhere a mind updates on evidence. Computers are just human thinking, automated. Fix the thinking first. H = the hypothesis under test · E = the evidence · P(E) > 0 required · priors empirical, never 0 or 1
The Intellectual Resistance
Resistance is futile.
Resistance is the whole point.
In a circuit, resistance is what makes current usable — it sets the terms, drops the voltage to something that won't burn the system down. Thinking is the same. Friction, doubt, the deliberate pause before the answer: that's not the obstacle. That's the work.
Ohm's law, applied to consulting: V = I·R. Push raw input through with zero resistance and you get noise. Put a real mind in series, and you get a signal worth acting on.
Organic Computer is a project of the Intellectual Resistance — the larger body of work on logic, reason, and thinking that shows its work.
If this datasheet made you want to measure rather than trust it — that instrument exists.
Same mind. Same method. Now a local-first scientific tool that benchmarks what minds — silicon and carbon — can actually do. Blind tests, machine-verified logic, sealed evidence you can audit. Not a pitch. A proof.
See Calibration Scope →References
Every figure above is independently verifiable.
- NISTA. Madhavan, “Brain-Inspired Computing Can Help Us Create Faster, More Energy-Efficient Devices,” NIST Taking Measure, 15 Mar 2023. nist.gov ↗
- FHI OxfordA. Sandberg & N. Bostrom, Whole Brain Emulation: A Roadmap, Technical Report #2008-3, Future of Humanity Institute, Oxford University, 2008 — Table 9 (spiking ~10¹⁸, electrophysiology ~10²², metabolome ~10²⁵ FLOPS). PDF ↗ · Oxford ORA ↗
- AI ImpactsAI Impacts, “Brain performance in FLOPS” & “Brain performance in TEPS” — TEPS-derived functional range (~0.18–6.4×10¹⁴ TEPS). FLOPS ↗ · TEPS ↗
- FHI OxfordK. E. Drexler, Reframing Superintelligence: Comprehensive AI Services as General Intelligence, Technical Report #2019-1, Future of Humanity Institute, Oxford University, 2019 — brain’s basic functional capacity below ~1 PFLOPS. PDF ↗
- NIH / NINDSNational Institute of Neurological Disorders and Stroke, Brain Basics: Know Your Brain — adult human brain ~3 lb (~1.3–1.4 kg). ninds.nih.gov ↗
- Front. Hum. Neurosci.S. Herculano-Houzel, “The human brain in numbers: a linearly scaled-up primate brain,” Front. Hum. Neurosci., vol. 3, art. 31, 2009 (~86×10⁹ neurons). doi:10.3389/neuro.09.031.2009 ↗
- NeurologyD. A. Drachman, “Do we have brain to spare?,” Neurology, vol. 64(12), pp. 2004–2005, 2005 (synapse count on the order of 10¹⁴–10¹⁵). doi:10.1212/01.WNL.0000166914.38327.BB ↗
- Mayo ClinicMayo Clinic, “Water: How much should you drink every day?” (total daily fluid ~2.7–3.7 L; food supplies ~20%). mayoclinic.org ↗
- DOE / ORNL · TOP500U.S. Dept. of Energy & Oak Ridge National Laboratory — Frontier supercomputer, ~$600 M program. TOP500 Nov 2024 (No. 2): 1.353 EFLOPS Rmax, ~24.6 MW. TOP500 system ↗ · ORNL ↗
- APA · UC IrvineG. Mark, Attention Span (2023) & interview, Speaking of Psychology (American Psychological Association) — average attention on a screen fell from ~2.5 min (2004) to ~47 s. apa.org ↗
- Br. J. Nutr.M. S. Ganio et al., “Mild dehydration impairs cognitive performance and mood of men,” Br. J. Nutr., vol. 106(10), pp. 1535–1543, 2011 — mild dehydration degrades vigilance, working memory, and mood. doi:10.1017/S0007114511002005 ↗