In the Wimmera Mallee town of St Arnaud in 2023, around 300 residents joined a tractor rally through the streets to protest renewable energy infrastructure. At a community meeting about the proposed VNI West transmission project, government representatives were heckled until Victoria Police advised the meeting be adjourned. More than 80 landholders signed a declaration refusing survey access to the company responsible for building the transmission line. Many of them had first learned their properties were in the proposed corridor through a QR code in a local newspaper.

Six months ago I pulled a local language model onto my laptop. Took about 12 minutes with Ollama. No account, no API key, no data leaving the machine. It felt like a small act of sovereignty, exactly the kind of local-first approach I’d been arguing for.

Then I started using it. And I noticed something.

The model’s cultural centre of gravity was somewhere around San Francisco, circa 2022. Ask it about food systems and it defaulted to commodity agriculture and supermarket supply chains. Ask it about community governance and it reached for American municipal frameworks. Ask it about traditional land management and it gave me a careful, earnest summary that read like it had been assembled from university anthropology papers, not from anyone who had actually grown anything, or sat with the country long enough to know it.

A few days ago I was listening to an episode of It Could Happen Here, Cooper Quintin and Colonel Panic from the EFF walking through the American surveillance state. Flock cameras on every corner. Cell site simulators at protests. Facial recognition with no accountability, built on databases scraped from your social media without asking. PenLink buying location data harvested from your phone’s apps and selling it to law enforcement, no warrant required, because it came from advertising networks instead of a phone carrier.