Last week I facilitated a full-day generative AI workshop for community aged care leaders in Melbourne, hosted by Moonee Valley Connect. The session was booked out, and the room was strong, with managers and coordinators from across community aged care services; most of whom arrived with questions and left with working tools.

I want to be specific about what happened, because the generic version of this story — "we explored AI and it was great" — doesn't do it justice.

The session covered Claude, ChatGPT, CoPilot, NotebookLM, and several other tools. By mid-afternoon, some participants had built custom GPT workflows. Others set up structured projects inside Claude for ongoing use. Others had developed step-by-step AI-assisted processes for tasks they'd been doing manually. Someone emailed me over the weekend, happy that they worked through a full data analysis for their organisation.

We're now in conversation about running similar sessions in other locations. If your organisation or network would benefit from a hands-on workshop like this (designed for aged care leaders, delivered in person), reply to this email and we'll work something out.

A study published this month by researchers at Harvard Medical School, MIT, and the University of Rochester pretty much states that immigrant workers are keeping older people alive.

The research, released as a National Bureau of Economic Research working paper, analysed data from 18 million Medicare recipients in the US. The central finding is that adding 1,000 immigrants to a metropolitan area results in approximately 142 additional healthcare workers — including 88 aides, nurses, and doctors. The effect is additive: foreign-born workers don't displace domestic workers and don't depress wages. They expand the pool.

The mortality impact is measurable. The researchers estimate that a 25% increase in the steady flow of immigrants to the US would prevent around 5,000 deaths among older adults nationwide. Put simply, more home-care workers means fewer people end up in nursing homes, and we know that people consistently do better ageing at home.

The US numbers are striking, but the pattern translates. In Australia, roughly one in three aged care workers was born overseas. The sector currently employs around 456,000 workers and needs to reach over 850,000 by 2050. The gap between those numbers is going to be the main challenge of the next thirty years, and migration is one of the few levers with the scale to meaningfully close it. (Another one is technology.)

I found this research quite interesting for another reason. Usually, policy debates about immigration rarely account for downstream health outcomes. Rather, the framing tends to be economic, jobs, wages, fiscal impact, and then inevitably, it turns political. This study reframes immigration as a public health intervention, one that directly determines whether older people live or die.

The Harvard study was covered by the Harvard Gazette last week. The full working paper is available online.

(By the way, the subject connects directly to last week's conversation with Robert Bean on cultural intelligence, and why Australia's culturally diverse workforce is an asset many providers haven't learned to fully leverage.)

Pew Research published survey data earlier this month showing that 33% of people caring for a parent said it had negatively affected their physical health, and 39% said the same about their emotional wellbeing. At the same time, most said caregiving had positively affected their relationship with that parent.

The data captures a tension that anyone who has cared for an ageing family member will recognise. The work is simultaneously enriching and depleting, and there is not a lot of support for the people doing it. Worth reading alongside any discussion about workforce shortages, because the unpaid family workforce is under pressure too.

An essay on Aeon revisits Beauvoir's 1970 book La Vieillesse and her argument that society's neglect of older people is driven by bad faith, a wilful refusal to acknowledge that old age is our own future. Beauvoir's claim is that we treat older people as "the Other," a foreign species "outside humanity," while preaching an official ethic of respect that our actions contradict daily. The essay is one of the more challenging things I've read on how we talk about ageing versus how we actually treat the people experiencing it. It's long, but it’s worth the time.

AARP Research surveyed over 500 solo agers (people aged 50 and over living alone without a spouse or partner). Solo agers tend to be healthier and more socially connected than the stereotype suggests, but they also have fewer safety nets when something goes wrong. More than a third said freedom and autonomy are the best parts of living alone. Their top sources of joy are friends, hobbies, family time, and physical activity. However, 68% are concerned about becoming reliant on others, and 62% worry about physical decline.

The tension between present happiness and future vulnerability is the real story here, as the number of people ageing without a live-in support network continues to grow.

This week I reviewed Anthropic's documentation on reducing hallucinations.These are the three system prompt instructions I recommend. They've certainly changed the quality of what I get back.

"You are allowed to say I don't know." Without this, Claude fills knowledge gaps with plausible fiction. The default behaviour is to always produce an answer, even when it shouldn't. Adding this instruction means you actually get "I don't have enough information to answer that" — which sounds simple, but the difference in output reliability is significant.

"Verify with citations." Tell Claude that every claim needs a source, and if it can't find one, it should retract the claim.

"Use direct quotes for factual grounding." This forces Claude to extract word-for-word quotes from your documents before analysing them. It prevents what I'd call paraphrase-drift, where the model subtly changes meaning while summarising.

There's a trade-off, however. Research suggests that citation constraints can reduce creative output. So I don't run these for everything. For research and compliance tasks, all three go on. For brainstorming and drafting, I let Claude think without the guardrails.

ARIIA's Partnership Program matches aged care providers with technology innovators and researchers for co-design, piloting, and implementation projects. They actively facilitate introductions and are currently listing several open opportunities.

Three examples currently on their noticeboard:

👉 If any of these are relevant, or if you want to browse what's available, visit the ARIIA Partnership Opportunities page. Providers and technology partners can both list and respond to opportunities.

Remember Press helps people turn rough notes, scattered memories, and half-finished drafts into published books. They've been doing it hands-on for years — working closely with people through the whole process, from first conversation to finished copy.

They're now in the final stages of building an app that guides people from first idea to finished memoir: structured prompts, expert guidance, a clear path through. No finished manuscript required. Rough ideas, old journals, fragments — that's what it's designed for.

They're looking for three beta testers before launch. It's free, and your feedback will directly shape the final product.

👉 If you've been sitting on a story you keep meaning to write down, get in touch with Christine Green by [email]

Novabridge, an Australian ServiceNow Elite partner with existing work in health and aged care, is looking for two or three providers to help shape a platform they're developing for the sector.

The problem they're starting from is familiar: staff credentials tracked in spreadsheets, complaints logged in systems disconnected from care records, compliance reporting assembled manually every time an auditor arrives. What they're exploring is whether a single hosted platform (purpose-configured for aged care, no enterprise IT team required) can bring those workflows together for providers operating five homes or more.

This is a co-design arrangement. Providers contribute operational knowledge and help shape what gets built; in return they get early access and direct input into the product's direction. (If the conversation moves toward a commercial arrangement down the track, Novabridge will name a price range; no obligation assumed.)

If you know ServiceNow from a previous life in health systems or government, this is that, built for aged care.

👉 If this is relevant to where your organisation is headed, contact Simon Mackey by [email].

Each week, I review developments in ageing and aged care and what they mean in practice. If this was useful, forward it to someone in the sector who'd appreciate it.

George Gouzounis