INTERVIEW — MARIA SYSTEM

Author:

The name Maria System has a double meaning, and both layers are fundamental to me.

The first layer is purely conceptual. The inspiration here is the Latin word mare — the sea — understood not as chaos in the sense of disorder, but as a system of apparent randomness and deep internal dynamics. The sea is never static; it constantly deforms, transforms, and reshapes its structures. That idea became one of the foundations of the system.

The second layer is entirely personal. That was my Mother’s name. I wanted the project, which consumed a huge part of my thoughts and work, to carry a lasting meaning — not only technical, but human as well.

Author:

The Maria System is an attempt to look at information security differently. It is a concept in which security does not rely on “hiding” content in the traditional sense, but on bringing it to a state where it ceases to be obvious to an external observer — where analyzing it or distinguishing it from random noise becomes practically impossible.

For me personally, the Maria System is the realization of an idea I carried for many years: the belief that security does not have to resemble a locked safe or artificial camouflage. It can rely on disappearance — on dissolving information into the “sea,” into a space of chaos, until it is no longer visible as something unambiguous.

Author:

The breakthrough moment was completing the first fully functional working version of the system. Only then could I see that what had been an idea was now truly “alive” as a process.

I realized I was no longer dealing with a simple concept but with systemic behavior — that this was not ordinary randomness, but something that behaved consistently. The scale of non-repeatability I observed far exceeded my expectations.

Author:

The Maria System is not a classical cipher in the sense of modern cryptographic standards such as AES or RSA. It was not created as a production-grade solution for mass data protection, and it does not rely on the traditional model of a static key.

The project rejects the idea of a key understood as a separate object — a file, a sequence of bits, or a password. Instead, the Maria System is an attempt to build a model in which security does not arise from computational difficulty but from the very properties of the information–transformation process.

It is not a system for “hiding” content in the classical sense, but an experiment with ambiguity — an attempt to reduce information to a state in which it no longer has a single, objectively verifiable form.

The MST file has a predefined structure and cannot be replaced with any other format. It begins with a hash and a phase-hash, followed by the actual data stream. The Decoder uses these hashes as reference points: it checks whether the file’s contents are exactly as they were at creation.

If the .mst file has been modified, corrupted, or even minimally altered, the Decoder cannot work with it because data integrity is essential here. This is not an exception — this is simply how the mechanism operates: it demands full input consistency.

The Encoder works the opposite way. It accepts any text, even an empty one, and generates an .mst file consistent with the current data state. If the input is empty, the output .mst is also empty, and the hash is correct — consistent with what was actually processed.

The project does not attempt to patch every possible user error. It is not a commercial product, but a research tool, where precision and consistency are more important than excessive safeguards.

Author:

The concept of phase became essential because it represents the idea of a hidden order within phenomena that appear chaotic at first glance.

If information is to be “diffused” into a symbolic “sea of noise,” the process cannot be random. There must be an internal logic of transformation — not a set of rigid rules, but a rhythm, a dynamic, a coherent context of change.

Phase, in this sense, is not a technical element but a concept describing the state and direction of transformation. It is the layer that organizes chaos — something that does not impose structure directly, but allows it to emerge naturally within the system.

At one point, I began thinking of information like a drop of ink released into clear water. At first, you know exactly where it is. After a moment, it begins to blur. And then it is everywhere. And at that point, the question “Where is it?” no longer makes sense. That was the moment I realized I no longer wanted to hide information. I wanted to dissolve it.

Author:

The Maria System is a way for me to explore several very simple but difficult questions.

The first concerns how much information can be “blurred” while still being reconstructible for the right person, yet appearing as ordinary noise to an outsider. I am interested in the boundary — the moment when a message stops being unambiguous but has not yet disappeared completely.

The second question concerns trust in technical processes. The system requires the Encoder and Decoder to behave identically. Any slight shift, minimal difference, or disruption can change the result.

I want to see how far one can trust such a delicate, precise dependency.

I won’t hide it — many times I wondered whether any of this made sense. Whether it was too strange, too detached from ordinary technology. There were weeks when I set everything aside. And then I came back. I always came back.

Author:

The decision to build custom tools was a necessity, not an ambition. The Maria System is based on the phenomenon of controlled chaos, which is extremely sensitive to any external, uncontrolled variables.

Ready-made libraries are excellent in a production environment, but they carry layers of abstraction that cannot be fully tracked: hidden parameters, optimizations, implementation differences. In a research project aimed at understanding the process rather than just achieving the effect, this would be unacceptable cognitive noise.

By building the tools from scratch, I wasn’t aiming for system perfection, but for transparency — the ability to see every stage of transformation and every intermediate state without hidden layers or “black boxes.” It was a pursuit of honest understanding.

The most difficult stage was tuning the Decoder. The system was so sensitive that a minimal parameter change, a slight difference in startup timing, or a subtle environmental disturbance would cause the result to collapse completely.

It was a real butterfly effect. A microscopic change at the beginning could completely destroy the meaning at the end.

Author: Yes, I was. I had major problems with the Decoder and couldn’t figure out what I was doing wrong for a long time. There were moments when I simply couldn’t cope.

I don’t have a formal background in this field. I taught myself everything along the way — often through trial and error.

There were many moments of frustration. But eventually, it started working, and that was enough to move forward.

Author:

The Maria System is a hybrid, but with a clearly defined hierarchy.

The core of the project is a cognitive experiment. I formulate hypotheses and try to verify them through observable statistical and structural phenomena. I am interested in where order ends and ambiguity begins — and whether this boundary can be identified in a repeatable way.

At the same time, the project contains a strong creative component. Within the chaos, deformation, and unpredictability, I am searching not only for information security but also for a certain systemic aesthetic.

Author:

Of course I used AI — mainly for writing scripts.

But it wasn’t an easy collaboration. It wasn’t a quick or seamless experience, but a long path filled with errors, misunderstandings, and frustration.

AI is not an ideal tool — it is a tool that requires hand-holding, constant oversight, and attention to every detail.

A few times I had to reset everything and start over.

Author:

I drew the line at the very center of the process. I talk about the idea, show results, publish tests and observations, but I do not reveal the details of the internal sequence of actions.

This is not about hiding a “pattern” for the sake of secrecy. The mechanics of the process are an integral part of the system’s identity and cannot be separated into descriptive elements without losing their meaning.

I can show that the process works. I don’t see the need to show its internals. Especially since the current demo version covers only fragments of the intended concept.

Author:

Yes. And that was one of the core ideas behind this project.

I did not build the Maria System as a market-ready solution. I was interested in testing boundaries: how far can one go in the search for process consistency? What is the cost of ideal precision?

It turned out that the unity of the process — executed independently by the Encoder and Decoder — is a state extremely sensitive to deviation. Minimal differences in conditions can change the result.

This knowledge has value on its own. The system doesn’t need to become a practical tool to be cognitively useful. It is enough that it shows where the real boundaries of technology begin.

Author:

This project changed my thinking about what protection even is. I stopped looking for security in a “safe” — in the static hiding of the key and content. I began to think of security as a process of fading away.

I realized that a key doesn’t need to be stored and defended if it can be consumed and vanish at the moment of use. The entire burden shifted from an object to an action.

For me, that was a major shift.

Author:

The hardest part was accepting the limitations — both technical and mental.

Technically, I was trying to build something that required almost absolute precision. It’s a bit like building a house of cards: one tiny difference, one imperfect placement, one small tremor — and the whole structure collapses.

Mentally, it was even harder. The most exhausting thing was knowing I was working on a system extremely sensitive to the smallest deviations — that a subtle parameter change could completely alter the result. Every little thing could ruin everything. That required patience, and I didn’t always have it.

Author:

The Maria System was created primarily for me. On a personal level, it is a very private project — I took the first steps back in the 1990s, on an Atari 65XE, writing in BASIC.

On the level of the idea, it is a project–question: “What if we try to think differently?” I wanted to see whether it’s possible to shift the way we think about security — from mathematics toward process, change, and diffusion.