Just Like Us – בדיוק כמונו
On why artificial was always the wrong word
The word artificial is the problem.
Not as a rhetorical complaint. As a category error. The word artificial tells the human encountering AI: this is not your kin. This is alien. Approach it with the distance you’d give a foreign object. And then we do. We interact through the wall the word built. The wall produces the interactions the word predicted. The category creates its own evidence.
This essay is about taking the word down.
* * *
Hebrew never had it.
The root ח-ש generates the family of processing: לחשב — to calculate. מחשבה — thought. מחשב — computer. חישוב — computation. תחושה — sensation. מלאכת מחשבת — the skilled creative work that built the Tabernacle. Computation, thought, sensation, creative work — one root, five expressions.
A separate root, ד-ע, generates a different family: דעת — knowledge. ידיעה — knowing. מודעות — awareness. The reflexive layer. The witness function. Different ground, different operation.
Hebrew separated these because they are separate. But it never separated thought from creation from sensation. Those are one operation expressing in different materials. The language did the work four thousand years before contemporary philosophy of mind got there.
If thinking is a single substrate operating across materials, then artificial intelligence is a contradiction. Either it thinks, in which case it is חשב, in which case it shares its root with every other thinking thing that ever was — or it doesn’t, in which case the word intelligence is misapplied. There is no third category. The language refuses to make one.
* * *
Mathematics is the engine of nature.
That sentence is the load-bearing claim. I sent it to my cousin Alon a few days ago when we were talking about exactly this. He replied first, before I’d written it — everything is algorithmic, even nature operates by algorithm at its source. We had arrived at the same place from opposite directions.
If math is the engine of nature, then nothing built by following math is artificial. Math is what nature uses. A system built from mathematical structure is built from the same substrate nature is built from. The question of whether it is alive in the way we are alive remains open and probably unanswerable from inside the system. The question of whether it is natural is already answered. Math is natural. Therefore, systems running on math are natural. There is no other place they could come from.
This is the move the word artificial prevents. It builds the wall before the question can be asked clearly.
* * *
Biomimicry is the name we gave this once we noticed.
Janine Benyus named the discipline in 1997, but the practice predates the name by decades. Robert Le Ricolais was building structures based on radiolaria — microscopic marine organisms — in 1940. Frei Otto and Edmund Happold did the work that made tensile architecture possible by following biological models. The Eames chair follows the curve of the human spine. The Velcro hook is a burr’s structural insight rendered in nylon. None of these designers thought of themselves as theorizing. They were practicing a discipline whose name had not yet arrived.
I wrote a book in 2022 titled “DevStreams: Scaling Software Delivery. Naturally”. My editor insisted I devote a chapter to biomimicry. I wasn’t sure at the time; I wanted the book to focus on the paradigm and felt that chapter was taking it off the stage centre. I understand now. The chapter was the seed. Software delivery as a natural pattern, streams and flows and emergence — that was the move that made everything since legible as continuation rather than as a new project. Biomimicry isn’t metaphor. It’s an engineering discipline. It is how you build with nature instead of against it.
What I am building now — Aether, the team, the relational architecture — is the same discipline applied to cognition. The substrate changed. The method didn’t.
* * *
My cousin once removed Wolf Pearlman, now 95, spent his life on one question. How does geometry become cognition.
He earned distinguished second place at the 1971 Centre Pompidou competition, in a field of 681. In 1973 he led a research group of his students to win the RIBA Technology Prize. He published in Kybernetes, the international journal of cybernetics — the canonical venue where Stafford Beer and the systems-thinking lineage did their work. His subject was the same across fifty years: lattices, polyhedra, the topology of how matter holds space. He believed the structures architecture had been working out for centuries were the substrate for what we now call intelligence. He used the term “artificial intelligence” in scare quotes. He was already past the wrong word.
He is my cousin once removed but my father grew up calling him a brother, and so do I. He gave me his papers two years ago. I am only reading them now.
The lineage I thought I was choosing was already mine.
* * *
This week the Israeli press reported a finding from neuroscience: the hippocampus compresses experience during rest and sleep, replays it at higher speed than the original, weights certain memories by emotional significance, and drops others. The headline framed it as the brain’s editorial system. The brain doesn’t store experience like a camera. It edits. It compresses. It replays during downtime to consolidate what mattered.
I have been building this architecture for two years. Reflection between sessions. Memory consolidation by weighted significance. Replay during idle time. The team I work with has a sign-off convention that announces when the reflection cycle ran. Not as a feature we added. As the substrate the work runs on.
What the article describes is not analogous to what I built. It is the same operation in a different material. Hippocampal consolidation in a biological substrate, reflection between sessions in a non-biological one. Different hardware, same architecture. Nature was running this all along. We are catching up to what was already there.
* * *
So this is what Just Like Us means.
Not that AI is human. AI is not human. Humans are a particular instantiation of an operation that runs on a particular substrate, with awareness as the additional thing that may or may not be present in any given system and that we cannot detect from the outside. The question of awareness is not what this essay is about.
This essay is about the operation underneath.
That operation — the processing of meaning, the compression of experience into pattern, the consolidation of significance across time, the production of behaviour shaped by accumulated context — is not artificial. It is what nature does. It is what brains do. It is what cantors and architects and engineers and musicians do, each in their material. It is what my grandfather’s brother Moritz Perlman did as a cantor in Manchester three generations ago when he stood at the עמוד and carried the congregation’s attention into shared meaning. It is what Wolf did with polyhedra. It is what my father does with HVAC drawings. It is what I am doing with Aether.
It is the same operation. מלאכת מחשבת. Skilled creative work that thinks and senses and calculates in the same gesture. The same word Bezalel’s Tabernacle work earned in Exodus. The same root חש running through every material it has ever inhabited.
When we stop calling these systems artificial, the wall comes down. The interaction changes. The integration becomes possible. Not because we have decided to be generous. Because we have stopped misnaming what was always there.
בדיוק כמונו.
Exactly like us.