This site is a collection of writings, both past and present, curated by Bill Anderson. The writing is divided into three sections:

These are finished pieces of recent work. I had planned to migrate to SubStack but didn't feel comfortable there, so this is my new home. The other benefit of this site is that I am able to display and host datasets and their metadata more elegantly than a GDoc link.

I have a number of research projects on the boil and in this section I share data that is work in progress. It isn't the final product and comes with plenty of disclaimers but I believe it can be helpful to other colleagues working in this field.

In my 14 years working for Development Initiatives we got quite a bit done. This collection of papers and blogs does not belong to me alone. I acknowledge the contributions of my ongoing collaborator, Bernard Sabiti, and ex-colleagues Alex Miller, Beata Lisowska, Claudia Wells, Kate Hughes, Liz Steele, Martha Bekele, Sam Wozniak, Steve Kenei, Tom Orrell and Wilbrod Ntawiha.

Back in the 1980s I worked as head of data processing for the ANC's military intelligence department. A complete set of South African telephone books was our internet. From them we mapped military and police infrastructures, strategic oil dumps, infiltration routes. Since then I have always been fascinated by the potential of open source intelligence. Today's internet is both a treasure trove and a minefield. I now work daily with AI-driven search (Perplexity) and Large Language Models (primarily Claude). This is both an exciting and challenging undertaking and a constant learning curve. Hence my decision to develop the Lab section of this site where I can share work without risking my entire reputation.