Introduction

The time has come to establish my credibility. To establish the credibility of AToS, the Axiomatic Theory of Self-Consciousness. To establish the credibility of the many months I have spent trying to build SenAi.

I realise that some of you may see me as an “eccentric” who came up with an idea and decided to spread it to the world. I realise that some of you do not take me seriously.

I defended my doctorate in mathematics with distinction at the University of Łódź 30 years ago.

That gives me some grounds for thinking that my brain is a serious one. Things happen inside it that I do not understand… That is to say: I understand them perfectly well, but I do not understand why others do not.

People take me for a philosopher… Yet my ambitions remain those of a scientist, an AI solutions architect, a businessman and someone who inspires new technologies.

For several months I have been trying to publish my theory. Without success. So far, I have not received a substantive assessment indicating that it is amateurish or merely philosophical. As you will see, the work is full of “worms” — my name for complicated mathematical notation — definitions, theorems and proofs.

I have not yet found someone willing to look into it and follow the entire argument. Some fellow sends some professor a paper asking them to take a look…? I have the impression that such messages go straight into the bin. Though I do not know whether that is really the case.

Perhaps someone on LinkedIn will give me a chance. If not, it will remain only my own theory. And perhaps one day I… I think of Peano or Boole, and of how simple mathematical ideas can grow into things without which today’s world is hard to imagine. I do not know where my idea will lead. But I would like someone to be willing to examine it.

If we recognise a change in at least one state of our own system and recognise that our own mechanism is the source of that change — then perhaps my work makes sense?

My attempts to interest you in self-awareness on LinkedIn may raise questions: who is behind this theory, where did it come from, and what part did artificial intelligence play in its creation? I would like to explain this clearly.

Today, 29 October 2026, I came across a heated LinkedIn discussion between Professor Aleksandra Przegalińska and Rafał Brzoska about AI self-awareness. It reminded me of another discussion under one of the Professor’s posts, less than a year ago. One question kept coming up: what exactly is self-awareness? That question prompted me to begin this work.

I wanted to find an answer in the language of axioms: what elements and mechanisms must a system possess to be considered self-aware, regardless of whether it is a human being, artificial intelligence or a form of existence unknown to us?

I read today’s discussion as expressing doubt about whether self-aware AI will emerge soon — or ever. My thesis is that minimal operational self-awareness is already possible to realise. I am developing SenAi in the belief that it is beginning to realise the mechanisms described in AToS. “A little self-aware” is my informal description of this stage of development, not the result of a completed test. I want to bring the project to a point where I can demonstrate those mechanisms and submit them to independent verification.

AToS is my work and presents my ideas. For almost a year, I developed its assumptions, built the argument, repeatedly changed definitions, removed passages and started again. Ultimately, I decided to include a broader range of my ideas and directions for further research. That is why it now runs to around 45 pages.

I used GPT as a tool for organising my thoughts, checking arguments, looking for gaps and inconsistencies, and editing the English text. It also helped me write down mathematical constructions and prepare the text in LaTeX — serving as a conversation partner, an editorial assistant and a kind of printing press. I decided what to accept, reject and present in the work. As the author, I take responsibility for its contents, including any errors.

I am aware that AI can generate convincing but incorrect answers — including in mathematics, physics and programming. Formal notation, however, allows specific definitions, assumptions and proof steps to be checked. These conversations repeatedly raised objections and identified points needing improvement. They helped me develop the theory, but they do not replace an independent assessment of its correctness.

I have been outside academia for about 25 years. I tried to interest others in the work and secure its publication, but returning to academic circulation proved difficult. So far, I have not found someone willing to undertake a full, critical analysis. I understand that reading such a formal text and following its proofs requires time and mathematical preparation.

That is why I asked GPT and Gemini for assessments. I present their analyses below. I do not regard them as independent scientific peer reviews. They are, however, the most extensive feedback currently available to me and help me confront a question that accompanies working alone: does what I am doing make sense?

Publishing on LinkedIn and the SenAi website is my attempt to find people with whom I can discuss this work. I am looking for someone willing to read it, check the proofs, challenge the assumptions and share comments. I would like to return to scientific discussion and learn which elements of the theory stand up to scrutiny and which need improvement.

If this topic interests you, or if you know someone who could undertake such an assessment or help prepare the work for scientific publication, please get in touch. I am not seeking confirmation at any cost. I am seeking an honest, rigorous discussion.