A few months ago, I wrote about the growing graveyard of “smart” home devices, expensive bits of plastic and silicon that turned into bricks the moment a corporate server in Virginia or San Francisco was switched off. It’s a personal annoyance when your $300 security hub stops talking to your lightbulbs. Scale that logic up to national defence and critical infrastructure, and the stakes shift from a darkened living room to a crippled nation.

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.

In 2020, journalists asked Australian police forces whether they were using Clearview AI, the American company that scraped three billion social media photos without consent to build a facial recognition database. The answer, from several state forces and the AFP, was no.

Then Clearview suffered a data breach. The stolen customer list included Australian law enforcement agencies. At that point, the denials stopped.

That sequence (quiet adoption, public denial, disclosure only under external pressure) is the pattern. Last week I wrote about the cultural conditions that make it possible: the institutional trust, the “she’ll be right” pragmatism, the absence of organised civil liberties infrastructure that might have generated friction. This post is the inventory. Here is what Australia has actually built, layer by layer, and what you can do about it.

If you have spent any time in the off-grid radio scene over the last few years, you know the frequency divisions. You either ran on the sub-GHz bands (915 MHz in Australia and the Americas, 868 MHz in Europe) for long-range, bush-penetrating reliability, or you accepted the high-congestion limits of local Wi-Fi. It was a trade-off we took for granted. If you wanted to send a message across 10 km of dense stringybark, you needed the long waves. If you wanted global hardware standardisation, you looked elsewhere.

Every few months, Australia’s business press discovers a new emergency. Right now it’s the capital gains tax. According to the usual commentators, founders, VCs, and their aligned media, Labor’s move to replace the 50% CGT discount with inflation-adjusted indexation is an act of vandalism against Australian ingenuity. Entrepreneurs will flee. Talent will dry up. The unicorns won’t come.

I’ve been working in and around Australian agtech and startups for the better part of two decades. I’ve watched the same arguments recycled through every policy debate: the R&D tax credit, the ESVCLP scheme, the startup visa. The answer is always the same: give us more upside, or we’ll take our toys elsewhere.

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.

I’ve spent 30 years watching tech cycles come and go, from the first dial-up modems in rural Austria to the mesh networks I’m currently stringing across the Australian bush. Each time a “next big thing” arrives, we see the same pattern: a frantic rush to centralise, followed by a slow, painful enclosure of what should have been a common resource.

The current noise around AI in open source feels different. It feels heavier. There’s a justified fear that AI-generated code is hollowing out our commons. Maintainers are being buried under a drift of unvetted, mediocre pull requests, while a handful of platform monopolies strip-mine decades of community work to feed their proprietary black boxes.

I’m tired. Not the kind of tired that a good night’s sleep or a weekend off the grid can fix. It’s a deeper, more pervasive exhaustion, the fatigue of living in a culture of relentless, performative positivity. In the tech world, we’re told to “crush it,” to “move fast,” and to embrace every new “game-changer” with uncritical enthusiasm. If you’re not a believer, you’re a “naysayer” or, worse, a “blocker” of progress.

Over the past few weeks, I’ve pulled apart four different LPWAN mesh technologies. Now it’s time to bring those findings together and look at which tool fits which job on the property or in the community.

There is no “perfect” protocol. What we have is a set of tools with different trade-offs. I’ve evaluated all four across five parameters to help cut through the marketing noise and get to the technical reality.

The ClusterDuck Protocol (CDP) was where my mesh networking journey truly began. The story behind Project OWL (Organisation, Whereabouts, and Logistics), students building emergency communication networks after Hurricane Maria, resonated deeply, highlighting a technology designed not for hobbyists or industry, but for saving lives when infrastructure fails. While I found its concepts “much better thought through” from the outset, the project’s slow pace and patchy hardware support meant my personal involvement never truly moved beyond some initial tinkering.

After years of experimenting with various LPWAN mesh networking technologies, I’ve settled on Reticulum as my primary LoRa mesh platform. It emerged as the clear frontrunner not because it’s simpler than the rudimentary Meshtastic (it isn’t), nor because it’s overtly more feature-rich than the structured MeshCore, but because its design philosophy fundamentally aligns with what matters most: privacy, resilience, and true decentralisation. It supports multiple bearer protocols, making it a powerful tool for building a future decentralised network.

While Meshtastic serves as a solid introduction to LPWAN mesh networking, MeshCore represents a move toward more structured networks, particularly when the limits of ad-hoc flooding become a bottleneck. It addresses the “airtime” congestion common in simpler protocols, offering a far more robust path for community-scale infrastructure where a “best effort” approach isn’t enough. MeshCore is built for managed deployments and regional sensor networks where reliability and structured routing are the priority.