- Addressing the Biggest Problems in Using AI — Most work on ethics and correctness of AI tries to make individual models better-behaved. The Perseverance Composition Engine is very different, assuming that misbehaviour will happen, and structuring the system to catch it before harm is caused. This is modelled on the way human institutions have worked for centuries.
- Slow LLMs and MCPs are Hiding Problems — Agentic AI systems are currently slow enough that serious concurrency and coordination problems remain invisible. This is a technical note about what happens when that changes.
- Structure vs Constitution in AI Safety — Comparing two approaches, PCE vs Anthropic. Anthropic locates AI ethics in the agent, through training and guidelines. PCE locates it in the structure around the agent.
- AI, PCE and the Geth Consensus — Science fiction saw artificial organisations coming years ago. This is a collaboration between me and my Consul agent exploring how the story of the Geth illuminates how multi-agent AI works.
- Logical and thermodynamic reversibility — LLMs consume vast amounts of energy. One response to that is to build on the computer science principle that energy is not consumed by computation but by the erasure of information. By making LLMs which do not erase information, perhaps we can avoid an AI energy crunch.
Samba co-founder · AI safety · IP and privacy regulation · startups · humanities
Edinburgh, Scotland
No trackers, no ads, no data collected or retained.
The connecting thread across decades of work is: IP controls software; software controls power; and power over software should be distributed, not concentrated by any single state or corporation. I co-founded Samba, the project that prevented Microsoft from monopolising networked file storage for billions of users, and have been working on variants of the same problem ever since — through open source development, privacy law, AI safety research, and startup advising. In 2026 these are now the conflicting, global concerns of digital sovereignty and the startup economy.
Agentic AI, Ethics and Artificial Organisations
Open Source
- LumoSQL — A novel twist on the world’s most-used software, adding features of privacy and security. We found a way to add to SQLite without breaking compatibility.
- Not Before Time — Assembles existing technologies to create a universal way to time‑lock information using everyday software tools.
- Samba — I co‑founded the Samba project, preventing Microsoft and the US Government from becoming the default store for everyone’s files. Samba is about digital freedom, as shown in the official history. I have developed a timeline of battles lost and won. Samba taught me that technical leadership is insufficient.
- How to Replace Windows NT with Linux — The first methodology and comprehensive guide for organisations to migrate from proprietary software. It is definitely historical (1999, Microsoft replacement, etc) but also highly topical in 2026 as this has become a widespread technology goal.
- Reversible Execution — This is about rewinding systems and applications, and then running them forwards again to find complicated bugs and security problems. This helps in our world where software fragility is such a problem, and its what I did in the second substantial startup I worked for.
- Not Forking — This tool addresses a difficult area in software reproducibility and reliability. Not‑forking largely automates change management in ways that version control systems such as Git, Fossil or GitHub cannot.
- Fossil — A practical alternative to Git, with stronger cryptographic and reproducibility guarantees. I contributed improvements and now use it for several projects (but I still mostly use git.)
- Open Source to Chemical Rockets — Explains how I first found open source concepts.
Legal and regulatory
- Opportunity in GDPR Article 28 — There is a surprise buried in the legalese of the GDPR. There is some precise Computer Science mandated, and the only good way to achieve this is with something that looks a bit like a blockchain. GDPR enforcement is intended to get tougher each year, which means there is a need for a new kind of cloud business implementing this blockchain. Many non-EU countries have copied the GDPR.
- EU‑US Privacy Shield et. al. in 2026 — The problem of US companies using coercion to achieve corporate and political goals is not a new one; this site documents how I have been doing it for decades! But in 2026 this is now a mainstream concern, part of global trade wars and also shooting wars. Court decisions in multiple countries, laws being passed and struck down, and mis-applied computer science and social science make for a very tricky time. My speciality is in trying to find solutions that minimise risk on the various axes.
- Software Patents — I have done a lot of work on copyright and software patents, which have now morphed into something else entirely when combined with technical protection measures. This is the key to the problem of technical sovereignty, something I have been speaking about since linux.conf.au 2004.
- Microsoft Patent Process — Ballmer Era — A practical guide to Microsoft’s software patent strategy 1998–2014, written from direct experience fighting it. Historical, but the tactics evolved rather than disappeared.
- Security Standards and Certifications — How ISO27001, GDPR, NIS and related standards work in practice for UK organisations — demystified from the perspective of someone who has implemented them several times.
- Data Mobility Post‑Brexit — I am employed from time to time to look at post‑Brexit data mobility. This now involves the complexity of the data/trade wars. Planning for minimal disruption and cost is difficult.
General Technology, Medical and BioTech
Browse all content by topic or search full text.