---

The publication sequence now provides a useful view of that progression:

October — Mastering AWS Cloud
Established the cloud and computational infrastructure layer of the work, with AWS, SageMaker, and related technologies providing the initial implementation environment.

January — Mastering NLP with Hugging Face
Extended the work into transformers, diffusion models, reinforcement learning, and the language and generative systems required by increasingly autonomous architectures.

2023–2024 — Implantable AI research
Following invitations from MedHealth Outlook in the United States and Brains Magazine in Sweden, I began publishing elements of the research publicly. A subsequent series of more technical papers used cardiac implantable devices as a concrete environment in which to develop and test propositions concerning AI autonomy, control, monitoring, and decision-making.

Those papers eventually began traveling independently of the original research context.

Work from this period was subsequently referenced in European health-policy research concerning AI-enabled medical devices and independently entered research on artificial organs and implantable bioengineering. Other researchers extended the cardiac-device proposition into control engineering and, more recently, clinical cardiac research.

2025–2026 — From application-specific research to a general architecture
The work progressively moved beyond cardiac devices. The objective became to formalize the underlying mathematical, computational, governance, and cognitive architecture rather than continue treating each application domain separately.

That development became Thinking Cybersecurity: AI Automation through Diffused Intelligence , which is presently in the final publishing process. Prepublication material from the project has also circulated through academic repositories, including individual research components hosted through Authorea.

March 2026 — An unexpected mathematical crossover
Benedict Irwin's Mellin-Laplace-Fourier Transform and the Connection to Riemann Zeta Zeroes  (DOI 10.22541/au.177490745.56701615/v1) references material from the forthcoming work in connection with the mathematical development of his proposed sequential integral-transform approach.

I mention this one because it illustrates something I had not originally anticipated: research developed for an autonomous-systems architecture can become relevant considerably outside the application domain that motivated it.

2026 — Latent cognition
The research has since expanded into what I call the Latent Cognition  program: work concerned with moving beyond AI systems that retrieve, predict, or externally orchestrate knowledge toward architectures in which retrieval, symbolic formation, memory, and cognitive control become properties of the model itself.

Parts of this work are already available to the academic and scientific community through research repositories while I determine the appropriate publication route for the complete series.

Next — portability and operationalization
One immediate direction is to demonstrate that the architecture is not dependent upon the AWS implementation environment in which parts of it were originally developed. I am therefore working on its translation into a Google Cloud implementation.

A further volume will take the canonical architecture into a complete autonomous cybersecurity environment — not merely applying an LLM to security workflows, but treating cybersecurity as the operational domain in which the complete cognitive architecture can be exercised.

In parallel, I am developing a separate research paper from an early cloud-era cybersecurity incident I encountered while at Microsoft. That work is still at an early stage and will follow the appropriate internal review and publication process before I make any substantive claims from it.

Looking backward, what interests me most is not the number of publications. It is the trajectory: an initial attempt to model greater autonomy in implantable medical devices developed into a broader architectural research program, while parts of the earlier work independently began appearing in health-policy analysis, control engineering, clinical research, artificial-organ research, and mathematical work.

Some of those developments occurred years after the original propositions were published. That makes the chronology itself useful: it provides a record of which ideas were proposed, when they were proposed, and where they subsequently appeared or were taken further by others.

That is also why I am now consolidating the research into a more formal canon. The individual books and papers are increasingly better understood as stages of the same program rather than separate publications.