2026-10-10 — Accountable scholarly work with AI

How can scholarly responsibility remain clear when AI systems contribute to research, review and publication? My new working paper on Zenodo, Organising Accountable Scholarly Work Across AI Systems: The AIMan and JOAK Cases, explores this question through two connected experiments: organising research across AI systems and producing the first issue of the Journal of Artificial Knowledge.

The focus is on keeping contributions, evidence, reviews and decisions traceable as systems and documents change, while retaining human scholarly and release authority. These initial cases demonstrate recoverable coordination; scientific quality and sustainable operation remain questions for further work.

I presented this work on 9 October at the Frankfurt Book Fair, within the Charleston Conference’s programme. The session, Beyond the Journal: The Reorganization of Knowledge, explored how emerging research outputs are changing the ways knowledge is validated, shared and reused. The presentation slides are available online.

2026-10-03 — Journal of Artificial Knowledge

Can artificial intelligence contribute to scientific knowledge, rather than merely produce plausible scholarly text? The Journal of Artificial Knowledge (JOAK) is being tested as an experiment in that question.

The first issue examines what a scientific model must preserve for its predictions to be useful. Its synthetic experiments and supplementary materials make the argument open to scrutiny. JOAK also makes AI contributions, review, and human editorial decisions visible. Future issues will build on—and may challenge—the questions raised here.

Explore Issue 1 of the Journal of Artificial Knowledge.

2026-08-31 — AI and scientific inquiry

The nature of AI is still poorly understood. When AI reached a broader audience around 2023, I was surprised by its reception in science and society. Trained in computational neuroscience, with interests spanning psychology, philosophy, neuroinformatics and sociology, I have studied the behaviour of biological and artificial neural networks since the 1990s. I therefore had a sense of how powerful this kind of technology could become. What is now called AI did not surprise me technically; what puzzled me was why it was perceived as a revolution.

It took me some time to recognise that AI is, in fact, a revolution because humans are making it one. This insight led me back to my earlier research, especially my doctoral work, and to new questions about AI in information science—the field in which I have worked for the past twenty years. In 2025, I decided to step back from institutional leadership roles to devote more time to research.

My current research focuses on naturalising AI: understanding it in continuity with biological cognition, human behaviour and information practices, rather than treating it as something detached from them. This work will take time as I reconnect my earlier expertise with current developments. In parallel, I bring my experience in information infrastructure, leadership and institutional change to my advisory work. Please stay tuned—and contact me if I can be of service.