Applied Research

AI data centres account for roughly 0.14% of global freshwater withdrawals — far less than farming or leaking pipes. The real risk is where they are being built: more than half of the world’s data-centre hubs sit in high water-stress regions, and their location fixes their water source for decades. This paper maps the evidence, examines the Nile basin as a case study, and proposes three tracks to reduce lock-in.

A public-facing summary of the above. Written for a general audience, it argues that AI is not draining the world’s water — the problem is siting, and the tools to fix it already exist. Uses Egypt and the Nile as an illustrative case and outlines three practical measures.

Parliament’s minimum-age law checks a birthdate, not a person. This paper uses complaint data, investigative reporting, and Australia’s own evaluation of a comparable ban to show why it fails — and what would actually close the loop: prosecution of adults and verified-identity accountability rather than age gates that any child can bypass.

Moving a recipe from a 1 kg bench batch to a 112.5 kg kettle batch changes the physics more than the arithmetic. This handbook carries one demonstration recipe through both steps: Part A covers scale-factor mathematics, heating-surface geometry, evaporation, cooling time and cost per portion; Part B applies five commercialisation gates, a full HACCP plan, and the shelf-life rules that decide which heat process a sauce needs.

A practical kitchen team guide written for multilingual brigade use. Covers handwashing, temperature control, cross-contamination, storage hierarchy, and personal hygiene — written plainly enough for a new commis and precisely enough to meet HACCP standards. Free to download and adapt.

Cooking is usually explained by what it gave hominins — more energy, smaller guts, larger brains. Why hominins cooked at all receives less attention. This paper reviews the ape preference evidence, makes the case for deliberate repetition of an accidental discovery, and sets out an experiment to settle what current data leaves open.