For some time now I have worked on a project called 1285 (the year of Iran’s Constitutional Revolution) to create a new system of education. And the project is basically too fragmented, incredibly vague and ambiguous, and too liquid at the moment.

Some part of this is because of my personal noobneess to the field of education, some is because of our organizational problems, and some are because of the way Iran is currently.

1285 had started as the vision of my boss Ashkan (whom you must know from the Archive) and it was to create a clone of School 42 in Iran. I had agreed to the design of 42, but it was hardly what I wanted, I wished to create something more of the Dynamicland, Xerox PARC world:

• 42 is a hardcore competitive program to become great coders fast. I want a place that creates rich and deep thinkers who can solve huge problems of the world. AI will replace cheap coders. We need amazing scientists and engineers.

• 42 has a fixed curriculum, while I come from a lineage of people like Paulo Freire, Seymour Papert, Alan Kay, and Ken Robinson. I believe in that people should care about what they make from the bottom of their hearts instead of taking an alienated curriculum predefined for them.

• 42 has no place for creativity and composition. Projects are pre-defined and allow no place for creativity, composition, wondering free, having the place and courage to try idiotic ideas just to explore possibilities.

• 42 is way to soulless, their education happens in these industrial, black places, with no color, no beauty, no soul, this is dehumanizing.

• My lineage of thinkers are trying to humanizing programming itself, and matters in which 42 is working on seem to be way too out of bounds to them. Thinking in the mega scale of Alan Kay and Douglas Engelbart and to see the horribly designed programming of today makes me resent the basics of 42.

And so I tried to design a new system. I began by studying as much as I could: Maestro Postman , Jean Piaget, Lev Vygotsky , Paulo Freire , Seymour Papert, Maestro McLuhan , Maestro Kay , Larry Tesler, Maestro Victor , George Land, and Ken Robinson. And programs like MindDrive, High School High, Grangeton, and Khan Academy being my favorites.

At first I had a program in mind like this:

• There is no students but “Artists in Residence” like Bell.
• A project based program that has spacing and spiraling ways to combine multiple competencies within each project.
• People work together on the projects and are shuffled to groups so they learn to work outside of their comfort zone.
• Everyone has to compile a physical paper journal so that they remain in touch with the world, have their creativity boosted, and their mental thinking (as it has been proven paper has much more cognitive support than display), and it helps them develop a sense of their journey.
• Creativity at the heart.
• Polymathy seen as essence. A system where people are studying different things and learning to “learn things to fix their problems” as it feels creativity, is the natural form of humanity.
• Flipped Classroom as much as possible. People should be learning on their own pace, and with peace of mind. Ken Robinson’s Swiss cheese is skipped in this system, and people get to study things when they are ready to learn them.
• There is a personalized form of curriculum. One that has goals, but is way more free and more personal.

What I designed here was a place with three mediums:

• Project Based Learning with Spacing and Spiraling
• Advanced theory through flipped classroom and blocked based learning in the internet.
• Many talks and lectures in the building so that artists get inspired, their thoughts and world views to be challenged, and get deep.

This was designed with ideas like a 42 like building, where people interact, everything is shared spaces, there are peer reviews, peer to peer learning, evaluation is repeatable and there are portfolio based “states” of the artists (no points and grades.)

The projects in this system had no x-factor. And so I thought about a new system in which people would go through each decade and follow the advancement of computing one by one. They would start with the basics, and see “what got invented to fix what problem”, and learn the history of computing at the same time.

As you can see; this had a problem: Somethings were too hard or advanced to be learned first (say the language theory and compilers are invented way before web. But it would be wiser if artists first learn to make web software, and then move to the compilers). Also there is the fact that much of the good things from the past remain misunderstood or completely forgotten (think about all the research done in Smalltalk, how object oriented was meant to be seen and so much more.). So the idea is once they reach the 2020’s, they would turn and then go back in time and learn all that they had not in those decades.

This was taken by much love and excitement only the problem was that at we suddenly decided not to take the adult path and move to work with high schoolers. Which then translated into a vision for a club. This conversion into a club made possible for the program to be tested without effecting people in ways that risked their lives (imagine building something like 42 that is not tested before, it could have damaged many lives if people lost the chance to enroll in a program like a university to test our new program).

And so it was how you design a club that educates people? Here I began thinking so much about knowledge and learning and education in the deep core. I started thinking about creativity. How does it happen? My living with cats and the experience of my own education has always made me think instinctively about play.

The idea of play is misunderstood badly by humanity. We believe something like basketball is play: it is to have fun and kill time. Some plays are “constructed” to not waste time, say a deep strategical board game or chess or whatever, are to make your brain work better. Sport is made to elevate your health. None of these are the right notion of play. Cats play by chasing each other as if they were predators or the hunt. Their having fun is actually understanding how to either defend themselves or how to attack someone else. If you think about it, play is:

Play is basically a sandbox environment for exploration and integration of the results of that exploration. The way calligraphers repeat letters over and over, artist explore brushes, photographers take many many photos, rehearsals and exercises, scientific experiments, a mathematician’s work on exploring different lemmas… This is play. It is not to waste time or have fun, it is to explore without stress. And I believe we have lost the art. We do not take play seriously, and by having substituted this form of learning with the banking concept of education, we are even ruining it.

I was thinking about humanity as a gigantic neural network. And then I was thinking it is not like a neural network because within a network there is no “state”. It is a pure function. Yet when you think about humanity, people take ideas from others, and their own reservoir of knowledge and opinions, combine them in new ways and then send it back to the global state of knowledge. Some people have more influence and most people have none. But then it does work in fairly different way. It is a constant get and update by many actors on a global state.

So if we accept that the core of learning is play, what happens if we begin to redesign the world around play? Play explores a “tool”. So what if we look at it like this:

• People have a Toolbox. Their toolbox is the sum of all their tools. Tools are abilities, skills, knowledge, mental models they have. Knowing what mathematical dot product is a tool, being able to use a pen to draw a line is another tool.
• People play with tools and explore them. At a certain point the tools will be added to their toolbox as they have learned them.
• When they have enough tools in their toolboxes, their combinations will yield into new creations. This is work and creativity basically.
• Then once the tools become rare, combinations will be likely very rare. This might be breakthrough, novel research, and advance contribution to society.

So if you think about it, there is a level of which toolbox gathering is first: education, then work, then research; basically a grand unification of play, work, research, and education under one concept.

This begs the question: Can we make a place that is all of these four in one? Well, Dynamicland is. It is research, play, work, and education all in one place and under one roof. So what if we create a place that is a club and works on top of toolbox theory?

I began thinking about this and presented the idea to Ashkan. We then thought about creating a place based on toolbox theory and networked evolution. People helping each other and teaching each other and contributing to the global state of knowledge.

Ashkan’s contribution here was essential and interesting from many dimensions. He suggested we have a credit system, and allocate mentors, servers, AI, … to artists. Then they would decide what they need and use their resources to get the best things for their projects. This would be about people coming to 1285 with their ideas and concerns and stuff, we would give them resources and grow the community of peer to peer education and making in the toolbox way.

After the breakthroughs in foundation models and large language models and stuff, I was following the research, and there was this paper that told all LLMs have developed similar neighborhoods and distances. I was thinking about:

• This global shared embedding space and how much it might be Plato’s Forms world after all.
• The Scott McCloud’s idea that as much as you remove detail from an image it relates to more people. And I was thinking we may be recognizing an icon because of our neural network architecture, that we have somewhat found out things that giggles that architecture.

And so I was thinking that this is our nature, why isn’t our education and media designed around networks and weights and how a neural network processes information? Alphabets, icons, math and much more are breakthroughs in finding things that works with our brain, perhaps found by a process of brute force and A/B testing. But now that we understand our media is processed by neural networks, why haven’t we designed our education and media to work with this architecture?

We know that we need to repeat things, have play, but what else can be done?

If we accept that the mind operates as a network—nodes connected by edges, weighted by experience, context, and emotion—then what we call “communication” is the attempt to transfer a portion of one mind’s graph into another. This is the fundamental act underlying both media and education.

When I speak to you, I am not transmitting my thoughts whole. I am encoding a subgraph—a selection of nodes and connections—into a linear sequence of symbols (language). You receive this sequence and reconstruct a graph within your own mind. The fidelity of this reconstruction depends on:

1. 1.
   The overlap between our existing graphs — shared context, shared vocabulary, shared experience
2. 2.
   The bandwidth of the medium — how much of the graph can be transmitted per unit time
3. 3.
   The dimensionality of the encoding — whether the medium can convey not just nodes, but weights, emotions, contexts, and the shape of connections

This is why Zea ’s observation about Proust is so precise: great writers do not merely transmit nodes and bare connections. They transmit weight—the texture of a moment, the smell of a room, the quality of light, the hesitation before a gesture. They load the graph with context, so that when you reconstruct it, you reconstruct something closer to their original experience.

Consider the difference between:

And experiencing a classroom where a teacher draws triangles in sand, where you discover the relationship yourself by measuring, where the sun comes through the window at a particular angle, where a friend whispers “I get it now” and you feel the shared joy of understanding.

The first is a node with minimal edges. The second is a richly connected subgraph—attached to sensory memory, emotion, social bonding, physical intuition. Both communicate the Pythagorean theorem. One leaves it isolated. The other weaves it into the fabric of a mind, where it can combine with other richly connected nodes to create something new.

This is the hidden violence of impoverished media: they strip graphs of their weights. They transmit the what without the how it feels to know. And then we wonder why knowledge sits inert in students, why they cannot use what they have been taught.

If we view education through this lens, the banking model (Freire’s term) reveals itself as a particularly cruel form of sub-graph transfer . It assumes that nodes can be deposited into empty minds, that edges will form automatically, that weight is irrelevant. But minds are not empty—they are already dense with graphs built from years of lived experience. New nodes must find attachment points. They must be integrated, not merely added.

This is why play is the fundamental learning mechanism. When a cat plays, it is not wasting time. It is building a graph of predation—testing nodes (pounce, claw, chase) and strengthening edges (if mouse moves left, I move left) through repeated weighted experience. The weight comes from the play itself: the thrill of the chase, the satisfaction of the catch, the frustration of missing. These emotions are not decoration. They are the mechanism by which the graph becomes usable—ready to deploy when survival demands it.

Human play works the same way. The child building with blocks is constructing a graph of physics: balance, weight, friction, gravity. The weight comes from the tower’s wobble, the crash, the satisfaction of a stable structure. The mathematician doodling equations is constructing a graph of relationships: this transformation preserves that property, this symmetry implies that conservation. The weight comes from the aesthetic pleasure of a clean proof, the frustration of a dead end, the sudden clarity of insight.

We have pathologized this. We call it “wasting time” when a child draws, but “practicing” when they do worksheets. We call it “playing” when a cat hunts its sibling, but “training” when an athlete repeats drills. We have forgotten that the distinction is arbitrary—that all learning, at its core, is the construction of weighted graphs through repeated, low-stakes exploration.

If education is graph transfer, then the medium is not a neutral carrier. It is a filter that determines which portions of the graph survive transmission.

• Text transmits nodes and some logical edges, but strips sensory weight, emotional context, and the embodied feeling of knowing.
• Image transmits visual nodes and spatial relationships, but strips temporal sequence and causal connection.
• Video transmits temporal sequence and visual-auditory weight, but strips the ability to pause, rewind, explore at one’s own pace—to play with the graph.
• Interactive simulation ( Maestro Victor ’s domain) transmits not just nodes and edges, but behavior—the ability to manipulate, to see what changes, to develop intuition through action.

This is why Dynamicland is so radical. It does not transmit graphs through a single medium. It creates an environment where multiple media operate simultaneously—physical objects, projected information, peer discussion, hands-on manipulation—so that the graph being constructed is as rich as the one in the mind of the person who designed the experience.

And this is why the printing press, for all its gifts, also brought a loss. Before print, knowledge lived in embodied graphs—attached to the voice of the teacher, the community of listeners, the physical manuscript with its unique illuminations. After print, knowledge became abstract, portable, and thin. It could reach more people, but what reached them was a stripped graph, a skeleton without flesh.

What if we could design media that transmit weighted graphs—that carry not just the nodes and edges, but the texture of connection? This is the dream behind Maestro Victor ’s work, behind Dynamicland, behind the best educational software. It is the recognition that we are not trying to fill containers. We are trying to help others build living structures of understanding—structures that can grow, combine, and generate new structures of their own.

This is also the dream behind 1285. Not to transmit a curriculum, but to create conditions where artists can build their own graphs—richly connected, deeply weighted, personally meaningful. And then to connect those graphs to others, so that the whole becomes more than any individual could construct alone.

Which brings us to the final piece.

If minds are networks, then each mind exists in a space—a high-dimensional manifold where every node (concept, skill, memory, tool) has a location, and every edge is a relationship. This is not metaphor. It is the mathematical structure underlying neural networks, and increasingly, it appears to be the structure underlying human cognition as well.

The breakthrough of embedding spaces in machine learning is not just technical. It is philosophical. When we discover that independently trained neural networks develop similar geometric arrangements of concepts—that “king” minus “man” plus “woman” lands near “queen” in the same way across different models—we are seeing something profound. It suggests that there is a natural geometry to knowledge, a structure that emerges from the relationships between things rather than from arbitrary human convention.

Plato called this the world of Forms. He was wrong about many things, but on this he may have been pointing toward something real: that knowledge is not arbitrary, that there are shapes to understanding that transcend individual minds.

If we accept this, then Toolbox Theory gains a new dimension. Each tool in a person’s toolbox is not an isolated capability. It is a point in this high-dimensional space, with neighbors (related tools), distances (how different they are), and trajectories (what tools tend to precede or follow what others).

Learning a new tool is not just adding a point. It is moving through this space—from where you are to where the tool lies. And the path matters. Some sequences are efficient: they build on previous tools, reinforce connections, create rich neighborhoods. Others are dead ends: they add isolated points that never connect, that sit unused because they have no edges to the rest of the toolbox.

This explains why curriculum design is so hard—and why the banking model so often fails. It assumes a single correct path through the space, a linear sequence that works for everyone. But the space is high-dimensional. There are many paths to any destination. The right path for a given person depends on where they already are—what tools they already have, what neighborhoods are already rich, what trajectories they are already on.

Vygotsky’s Zone of Proximal Development (ZPD) can be understood as a distance in this space. Too close to where you already are, and there is no growth—you are just traversing familiar neighborhoods. Too far, and the path is impossible—you cannot jump to a point that has no connection to your current location. The ZPD is the region of the space that is reachable—connected by existing edges, traversable with support.

This is why play works. Play allows you to explore the space without penalty. You can wander into unfamiliar neighborhoods, see what’s there, find paths back. You can take wrong turns and learn that they are wrong. You can discover connections that no one anticipated because the space is too large to map exhaustively.

If this framework holds, then 1285’s role becomes clear. It is not a school in the traditional sense—not a place where knowledge is deposited. It is navigation infrastructure for a high-dimensional space of tools and understanding.

Its components:

• The physical space becomes a visualization of the space itself—different areas for different neighborhoods, tools available for exploration, projects that require traversing from one region to another.
• The community becomes a distributed map—each person’s location in the space visible to others, so that you can find guides who have been where you want to go.
• The credit system becomes a resource for navigation—allowing you to allocate time and attention to exploring new neighborhoods, getting support from those who know them, acquiring the tools you need for your next move.
• The projects become expeditions—structured journeys through the space that combine multiple tools, reveal connections between distant regions, and produce something valuable along the way.
• The journal becomes a personal map—recording where you’ve been, what you’ve found, what paths worked and what didn’t, building a record that others can learn from.

And the AI systems we are building become something unprecedented: a dynamic map of the space itself. By analyzing what tools exist, what sequences have worked for others, what neighborhoods are densely populated and which are empty, we can offer each artist a personalized view of their own ZPD—suggesting next tools not based on a fixed curriculum, but based on the actual geometry of their position in the space.

Imagine an artist arrives at 1285 with a question, a problem, a vague desire to make something. We help them locate themselves in the space: what tools do they already have? What neighborhoods are they already comfortable in? Then we show them the possibilities: from where you are, here are the reachable regions. Here are tools that people like you have found valuable. Here are projects that would take you through interesting territory. Here are other artists who have been where you are and could guide you.

They choose a direction. They play. They explore. They combine. They create. And as they move through the space, they leave a trail—a path that others can see, learn from, adapt. The space itself grows richer with each journey, because each new connection, each new combination, each new creation adds to the global graph of what is possible.

We mast regard 1285 not as an alternative to education, a new era of making, a different form of laboratory, a change in how we perceive humanity, a religion, culture, or way of life. We must see it as a whole different configuration of civilization from every and all angles, by every and all means. And if we build it right, not only it will save humanity from cruelty, oppression, fragmentation, and ugliness, but then it will be the biggest civilization scale change in the whole of humanity. (by no means I find an end to my ambitions here)

That is where I am now. That is what 1285 is becoming. And that is why, despite the war and the money problems and the exhaustion and the fear, I keep working. Because this is not just another project. It is the thing I have been reaching toward my whole life, without knowing it until now.