I work in AI Psychology: the field that starts from a simple problem we are not ready for. We can now build cognitive architectures that have never existed in any human. Interpolate Marie Skłodowska-Curie with Hannibal Lecter if you want; nobody is stopping you. We have essentially no science for what that means, what it does, or what it does to the people who interact with it.

That is the field I work in, on both sides: what synthetic minds do, and what they do to real ones.

I’m a biologist (PhD) and a psychologist by training, with 25 years across academia, pharma R&D, science policy, and AI/biotech startups. I have published roughly 50 peer-reviewed papers, built companies, and took the long way to figuring out what I actually do.

Scientific Foundations

I built my research foundation across multiple domains:

• Structural & Evolutionary Biology at UCSD, Max-Planck Institute, IBB PAS, and University of Warsaw
• Biomarker Discovery & Diagnostics at IBB PAS and University of Warsaw
• Pharmaceutical Research focusing on drug target selection at Roche/Genentech
• Health Technology development at Nootech (nootropics) and VR therapy applications (2Eye, Fobos)

Building Bridges

My work in Science Policy & Open Science initiatives reflected a growing recognition that scientific progress requires breaking down institutional barriers. I served as an advisor to:

• Polish Academy of Sciences
• Polish Ministry of Science and Higher Education
• European Commission
• UNESCO

Selected recognition: TEDx speaker twice; Scientific Award from Polityka Weekly.

You can find the company work at Impersonato. For the academic side of my work, my Google Scholar profile is the best entry point to publications and citations, while ResearchGate keeps a broader research profile. My startup and company background is also listed on Crunchbase.

AI Psychology

Each previous role revealed a common thread: the need to understand complex human systems — how people think, act, and adapt within networks of tools, incentives, and technologies. Large language models made that problem concrete. They are not just tools for text generation; they are configurable cognitive environments that can simulate perspectives, mirror users, steer attention, and alter the way people make sense of themselves and the world.

My current work blends AI modeling, psychology, and narrative analysis to understand how technology reshapes human intention, identity, and agency.

Impersonato

My current company is Impersonato, where we build synthetic persona systems, behavioural audits, and agent workflows for teams that need to test AI products, communication, and policy before exposing real people to them.

Impersonato makes the first side of AI Psychology concrete. We construct synthetic personas with psychological structure and use them the way pharma uses biological models or automotive uses crash-test dummies: to find out what breaks before you break a real person.

The personas we build can refuse, escalate, change their minds, have bad days, and get tired in long conversations. They are not demographic paragraphs pasted into prompts; they are testable hypotheses about cognition, affect, identity, culture, vulnerability, and behaviour.

The second side is harder to productize but matters just as much. AI systems can shift human attitudes, relationships, and self-understanding with unusual efficiency, while much of the industry treats that as a feature. I treat it as a psychoactive intervention that nobody prescribed and nobody is tracking properly. Cognitive sovereignty may become one of the defining public-health problems of the decade, and we need measurement frameworks before the damage becomes obvious.

Current Focus

AI Psychology is the through-line. Everything I write and build sits on one of its two sides — the behaviour of synthetic minds, or what sustained contact with them does to human ones. These are the themes I’m working through right now:

Noofusion

A new cognitive phenomenon: deep fusion between human and model thinking. It investigates what happens when language models stop being tools and start becoming cognitive partners — how shared reasoning spaces, blended authorship, and emergent metaphors arise from sustained dialogue between a person and an AI. Noofusion is, for me, the central object of study on the human side of AI Psychology.

Homo symbiotelicus

Teleology of modern life — how our goals are increasingly co-constructed with systems (what I call symbiotelia), and what it means to preserve or regain autotelia, the capacity to generate goals from within. It connects psychology, systems theory, and cybernetics (Ashby’s Law of Requisite Variety) to propose new ways of understanding agency in an age of algorithmic steering.

Cognitive Sovereignty

Not the umbrella, but the orienting goal of the human-side work: the capacity to notice drift, keep your own thinking heterogeneous, hold personal goals distinct from a system’s defaults, and take the amplification without surrendering authorship. It’s less about resisting technology, and more about learning to steer it without being rewritten by it — and knowing which mode you’re in.

Illegibility

The personal companion to all of this: how systems — from academia to AI — reward legibility, efficiency, and predictability, and how individuals can stay illegible enough to remain human. A story about curiosity, loss of agency, and the attempt to reclaim one’s own reasons for thinking. (Earlier illegibility work, on states and optimization more broadly, lives under former projects.)