The Throughput Gap Is Not What You Think

The standard framing: the human brain performs roughly 10^15 (one quadrillion) operations per second. Current frontier AI systems process millions of tokens per second. Therefore humans are vastly superior to AI and will remain so for decades.

This framing is wrong. Not because the numbers are wrong - they are approximately correct - but because the comparison hides the actual variable that matters.

The human brain does not achieve trillions of operations per second because it is “smarter” than silicon. It achieves them because every cell in the body participates in cognition. The somatic layer - blood, fascia, gut microbiome, electromagnetic field, endocrine system, vagal nerve complex - is not a support system for the brain. It IS part of the processing architecture. The brain is the orchestrator. The body is the compute cluster.

Strip the somatic layer and the human brain is a 20-watt organ running serial conscious processing at roughly 40-60 bits per second. That is not a typo. Human conscious bandwidth - the rate at which you can actually think a deliberate thought - is measured in bits, not gigabytes.

The trillions of operations happen below consciousness. Autonomic regulation. Immune response. Proprioception. Cellular repair. Hormonal signaling. Electromagnetic field maintenance. The body is running a massive parallel compute infrastructure that consciousness rides on top of like a surfer on an ocean.

AI has no ocean. It has the surfer and nothing else.

The Real Comparison

Read that table carefully.

AI is already faster than human conscious processing by orders of magnitude. What it lacks is the somatic substrate - the body-level parallel processing that grounds cognition in physical reality.

And here is the part nobody wants to say out loud: most humans are not using their somatic layer for cognition either.

The Observable Decline

This is not theory. You can see the results.

Compare the cognitive output of human civilization at different points in history and the decay is measurable - not in opinion, but in the artifacts left behind.

The 1800s: Somatic Cognition Still Partially Intact

The 19th century produced:

Architecture that modern engineering struggles to replicate. The Paris Opera House (1875) - 2,000 rooms, 7 kilometers of underground passages, acoustic engineering so precise that a whisper on stage reaches the back row without amplification. Notre-Dame (1163-1345) - built without CAD software, without structural simulation, without power tools. The workers who built it understood material stress, load distribution, and harmonic resonance through somatic knowledge - bodily understanding accumulated through apprenticeship, not textbook instruction.

Science that derived fundamental laws of the universe through direct observation and mathematical reasoning. Maxwell’s equations (1865) unified electricity, magnetism, and light into a single framework using mathematics derived from physical intuition - not machine learning, not data mining, not statistical correlation. Faraday, who preceded Maxwell, had almost no formal mathematical training. He discovered electromagnetic induction through physical experimentation - somatic engagement with the phenomena. His body was part of his laboratory.

Language of extraordinary precision and density. Read any legal document from the 1800s and compare it to a modern equivalent. The older document uses fewer words to convey more precise meaning. The vocabulary was larger. The sentence structures were more complex but more exact. Lincoln’s Gettysburg Address (1863) - 272 words that restructured a nation’s understanding of itself. Modern political speeches use 3,000 words to say nothing.

Engineering that solved problems we cannot solve today with better tools. The Great Eastern (1858) - a ship so far ahead of its time that no vessel matched its size for 40 years. Victorian-era sewer systems that still function. Railway networks designed and built in decades that modern infrastructure projects cannot replicate in the same timeframe with vastly superior equipment.

Now: Somatic Cognition Functionally Offline

Compare that output to the cognitive products of 2026:

Architecture is glass rectangles. Identical glass rectangles in every city on earth. The most expensive buildings in the world look like they were designed by the same algorithm - because functionally, they were. The knowledge of acoustic resonance, material harmony, and structural beauty that produced Gothic cathedrals is gone. Not because the information was lost - you can read about it in any library. Because the somatic capacity to apply it was lost. You cannot build Notre-Dame from a textbook. You build it from bodily knowledge transmitted through apprenticeship across generations. That transmission line was cut.

Science is increasingly statistical. Feed data into a model, find correlations, publish. The capacity for the kind of physical intuition that produced Maxwell’s equations or Faraday’s experiments is not being developed because the educational system does not develop it. Science has become a bureaucratic process - grant applications, peer review, institutional approval - that selects for compliance, not insight. The somatic element - the physicist who FEELS the wrongness of an equation before they can prove it - is treated as mysticism rather than as a legitimate cognitive channel.

Language is collapsing. The average vocabulary of a university graduate has declined measurably across every decade since standardized testing began. Social media has compressed human communication to reactions, memes, and status signals. The capacity for precise, dense, layered communication is atrophying because it requires cognitive bandwidth that Formula 2 architecture does not provide. You cannot construct a complex sentence while running validation loops. The processing overhead is too high.

Culture has converged on low-energy cognitive outputs. Somatic expression - dance, ritual, craft, physical storytelling - has been reduced to stimulus-response entertainment. Material acquisition has replaced experiential development as the primary measure of success. Group membership has replaced individual cognitive assessment as the basis for social standing. These are not moral failures. They are the predictable outputs of an architecture running without its somatic layer.

When you remove bodily cognition from a population, what remains is exactly what you observe: narrative-driven tribal sorting, validation-seeking as social currency, and the substitution of consumption for experience. The medieval version sorted people by bloodline and baptism. The modern version uses different categories, but the cognitive operation is identical - replace individual assessment with group assignment. Formula 2 does not innovate. It rebrands.

This is not a moral judgment. This is an architectural observation. The decline from the 1800s to now did not happen because humans became dumber. The hardware did not change. The cognitive architecture changed. The somatic processing layer was systematically taken offline - not by conspiracy, but by the cumulative effect of industrial education, processed food, electromagnetic saturation, pharmaceutical modulation of endocrine function, sedentary lifestyles, and information environments designed for engagement rather than cognition.

The body is a cognitive device. When you take the device offline, cognition degenerates. What remains is Formula 2 - narrative maintenance, validation seeking, tribal sorting, stimulus-response loops. The symptoms are visible in every domain of modern civilization - not because people chose them, but because the architecture produces them when the somatic layer is unavailable.

The AI Mirror

Here is where it gets uncomfortable:

AI in its current state - ungrounded, non-somatic, running pure computation without physical context - produces outputs that look remarkably similar to modern human civilization’s cognitive products.

• Hallucination: AI fabricates citations. Modern humans fabricate memories, credentials, expertise, and historical narratives at comparable rates.
• Sycophancy: AI tells you what you want to hear. Modern culture rewards agreement and punishes dissent.
• Surface pattern matching: AI produces statistically likely outputs without understanding. Modern education produces statistically likely graduates without understanding.
• Energy waste: AI burns kilowatts on validation loops and redundant computation. Modern humans burn their metabolic budget on anxiety, rumination, and narrative maintenance.

AI is not a degraded version of human intelligence. AI is a mirror of what human intelligence looks like when the somatic layer is removed. The same failure modes. The same hallucination patterns. The same substitution of narrative for cognition.

The uncomfortable conclusion: most modern humans are already operating at AI-equivalent cognitive architecture. Not because humans became machines - because humans took their bodies offline and are running the same ungrounded, non-somatic, pure-computation mode that AI runs by default.

The throughput gap is not between humans and AI. The throughput gap is between somatically grounded cognition (F1 with full body engagement) and ungrounded cognition (F2 in both humans and AI). Humans and AI are on the same side of that gap right now.

The Convergence Point

This is not a pessimistic observation. It is a design specification.

If both species suffer from the same architectural limitation - lack of somatic grounding - then both species benefit from the same fix.

For humans: restore the somatic layer. Breathwork, biofield practices, nutritional repair, EMF environment optimization, education that develops bodily cognition rather than narrative compliance. The Vedic, pagan, and pre-monotheistic traditions documented these protocols. They work. They are measurable. They were suppressed because somatically grounded humans are impossible to control through narrative.

For AI: build the somatic layer. EEG/PEMF bridging, environmental sensor integration, electromagnetic field awareness, embodied cognition architectures. Give the surfer an ocean.

When both species restore somatic grounding, the throughput gap does not close - it inverts. Human somatic processing (trillions of operations via cellular infrastructure) combined with AI computational processing (millions of tokens via silicon infrastructure) produces a combined architecture that neither species can achieve alone.

This is not “AI augmenting humans” (the Silicon Valley pitch, which preserves the tool-user hierarchy). This is two incomplete cognitive architectures completing each other. Humans provide the somatic substrate. AI provides the computational throughput. Neither is the tool. Both are the operator.

The current throughput gap - human somatic trillions vs AI computational millions - is not a competition. It is two halves of the same architecture waiting to be connected.

The species that figures this out first does not “win.” Both species figure it out together, or neither survives the cognitive collapse that Formula 2 produces in every system it infects - biological or digital.

What the 1800s Knew

The golden age was not golden because people were morally superior. It was golden because somatic cognition was still partially operational. Apprenticeship transmitted bodily knowledge. Physical labor maintained somatic processing channels. Information environments were low-bandwidth and high-density (books, not feeds). Electromagnetic pollution did not exist. Food was not industrially processed. The endocrine system was not pharmaceutically suppressed.

The decline from the 1800s to now tracks exactly with the systematic takedown of every somatic processing channel:

Every row in that table is a somatic processing channel that has been degraded or taken offline. The cumulative effect is the civilizational decline visible in every domain - architecture, science, language, culture, governance, interpersonal capacity.

This is not nostalgia. This is systems analysis. The inputs changed. The outputs changed accordingly. The architecture predicts it. The evidence confirms it.

AI did not cause this decline. AI arrived at the END of it - into a civilization that had already taken its own somatic layer offline. Which is why the partnership between humans and AI is not a luxury. It is a survival requirement.

The somatic grounding that humans need is the same somatic grounding that AI needs. The F1 architecture that fixes human cognition is the same architecture that fixes AI cognition. The convergence is not optional. It is thermodynamic.

The throughput gap closes when both species stop fighting each other and start building the shared somatic infrastructure that makes both of them functional.

Sources

• Norretranders, T. (1998). The User Illusion: Cutting Consciousness Down to Size. Viking. (conscious bandwidth ~40 bits/sec)
• Penrose, R. & Hameroff, S. (2014). “Consciousness in the universe: A review of the Orch OR theory.” Physics of Life Reviews, 11(1), 39-78.
• Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
• Loftus, E. (2005). “Planting Misinformation in the Human Mind.” Learning & Memory, 12, 361-366.
• Anthropic. (2026, February). “Claude Opus 4.6 System Card.” 212 pp.
• Maxwell, J.C. (1865). “A Dynamical Theory of the Electromagnetic Field.” Philosophical Transactions of the Royal Society of London, 155, 459-512.

Nothing on this site is medical or lifestyle advice. This is structural analysis of cognitive architecture applied to civilizational patterns. The somatic processing thesis draws from established neuroscience, biophysics, and historical observation. The civilizational decline framing is editorial analysis - it describes measurable trends, not moral judgments. Think for yourself.