Essays, notes, spines, and readings — on measurement, meaning, and human oversight of AI systems.
Everything here is organised by research thread and by format. Formats are not decoration: they set expectations about length and finish, so a short note can be published without pretending to be a full argument.
ESSAY
A developed argument. Long enough to earn its conclusion, and written to be read start to finish.
NOTE
A single point, stated once. Short by design — a distinction worth having, not a full treatment.
SPINE
A historical arc of a field, maintained over time. Reference material rather than a linear read.
REVIEW
A reading of someone else’s work — a paper, a framework, a regulation — and what it changes.
All writing
- An Introduction to Counterfactual Data AugmentationIn our fast-paced world, data-driven decisions are becoming more prevalent. Artificial intelligence (AI) and machine learning (ML) models thrive on vast amounts of data to function effectively. However, not all data is…
- Generative AI and Copyright Laws: Navigating the New FrontiersThe Challenge with Generative AI Herein lies the conundrum: If a machine generates an original piece of work, who owns the copyright? One could argue that the creator of the AI should…
Spines
Spines are maintained separately because they are reference documents rather than dated posts — they get revised as a field moves. They are indexed at Spines.