I came across a study that compared older Americans who voted with those who didn't, using decades of records from the long-running Wisconsin Longitudinal Study. Those who voted in the 2008 election were 45% less likely to die within five years, and still 29% less likely fifteen years on, after the researchers accounted for health, education and wealth. It held regardless of who they voted for. Their reading is that voting is a small altruistic act, in the family of volunteering and giving, and taking part in civic life might track with a longer life, the way exercise does.

A friend who also received the newsletter rang me over the weekend. We talked for the better part of an hour, and she kept circling one question — whether she should be open about using AI at work, or just quietly get on with it.

Later on the same day I came across an article in the Economist (the link is paywalled), titled “Will AI make you stupid?”. I think this question is the bigger version of the question I discussed earlier in the day. The worry is that if we hand our thinking to a machine, we will slowly lose the ability to do it ourselves. It sounds reasonable.

I don't think it's quite the right question, though. Whether leaning on AI costs you anything depends on what you were doing in the first place. Some tasks are only a means to an end, and the thinking in them is incidental (in our sector’s case it can be drafting rosters, formatting reports, etc). I think that handing these tasks to AI won't make you lose anything that you specifically wanted to keep.

Other tasks are the real thinking. Working out why a resident has gone quiet this week, when nothing in the notes says why. Noticing that "refusing care" is really fear, or pain, or not understanding what's being asked. These are deeply human skills that no system can reach.

So the question I'd ask instead is narrower. Which thinking do we still want people doing themselves?

That's what my friend was circling, even if neither of us put it that way at the time. Her unease about hiding it, the worry about going soft — both come from never having drawn the line. Decide which thinking is yours to keep, protect that, and hand over the rest without a second thought. There's nothing to hide once you know which is which, and not much left to worry about.

Thursday 11 June | 9am–4pm AEST | Novotel Melbourne on Collins | Free for CHSP and Support at Home providers

A full-day forum for CHSP and Support at Home-funded allied health teams. I'm speaking on AI; other sessions cover navigating change, improving referrals, dignity of risk, managing service agreements, interdisciplinary practice and change fatigue. The Department of Health, Disability and Ageing is hosting a Q&A responding to questions submitted in advance.

Hosted by SSD Connect

Monday 15 June | 2pm AEST | Online | Free

I'm running this webinar with Ageing Australia. The session covers where AI sits in the sector right now, the Three Tiers of AI in Aged Care framework, and where Tier 1 (the generic tools — ChatGPT, Claude, Copilot — already in use across the sector) is creating risk and opportunity. For executives, operational managers, governance professionals and anyone responsible for technology decisions in their organisation.

Hosted by Ageing Australia

Tuesday 16 June | 9.30am–12pm AEST | Online | Free for CHSP and Support at Home providers

A hands-on workshop for desk-based aged care staff working with AI on routine tasks — drafting, summarising, problem-solving. Covers what current tools can and can't do, prompting fundamentals, privacy and accuracy considerations, and what data should never go into a public AI tool. Participants get a dedicated resource page with prompts and templates afterwards. No technical background needed.

Hosted by SSD Connect

Tuesday 16 June | 2–3.30pm AEST | Online | Free

I'm presenting a walk-through of the refreshed Guidelines for the Responsible Use of Generative AI in Aged Care. The original Guidelines have been in use for 12 months across more than 100 aged care organisations. The update reflects what providers have done with them and how the AI landscape has shifted. The session covers what's changed, why, and how to adopt or refresh the Guidelines in your own organisation, with Q&A at the end. For providers across the sector — managers, governance and compliance staff.

Hosted by AI Adoption in Aged Care workgroup

Northwestern has spent 25 years studying "SuperAgers" — people over 80 who score on memory tests like someone three decades younger. Their brains don't follow the usual pattern. Some never build the plaques and tangles linked to Alzheimer's; others build them and stay sharp anyway. Their cortex barely thins, and one region tied to decision-making can be thicker than in much younger adults.

Exercise and diet vary across the group, but almost all of them are highly social and keep close relationships going. It lines up with the voting finding above. Read it on ScienceDaily.

After my interview with Rick Watson on capital asymmetry, Access Sydney Community Transport asked him to run a planning session with their team, and they kindly shared with me what came out of it. Watson's idea is that a lasting advantage comes from a mix of capital a competitor can't easily copy — economic, human, organisational, social, symbolic. The session turns that into an afternoon of sticky dots and an honest look at where a team actually stands. Read the case study.

As part of his ongoing research into capital asymmetry in aged care, Professor Watson is offering to facilitate strategic analysis sessions for Australian providers. (This is offered at no cost for Melbourne-based providers or covering travel costs only for providers outside Melbourne.)

More info at the bottom of this page.

To express interest: Contact Professor Watson directly by email

This is a reminder more than a link. Hallucinations are getting rarer, but the errors that slip through now are the plausible ones — clean summaries and confident answers that happen to be wrong. The trap is that the better the tool sounds, the less anyone checks it. If your team is using AI on anything that matters — case notes, reports, advice to families — remember that the human read is the part that can't be skipped.

Gemini is the tool I work with the least, so I don't write about it much. But I came across a beginners' rundown of the small, practical things it does well, and it's worth passing on if you or your team are still finding your feet.

Some simple but useful tips include:
• Paste in a dense policy, contract or service agreement and ask what's changed between two versions.
• Hand it the spreadsheet formula you can't get right and have it write or explain it for Excel or Sheets.
• Drop in a recorded session or a long video and ask for a short summary, or ask one specific question instead of watching the lot.
• Show it an error message from a system or device and ask what it means and how to fix it.

The usual caution applies — check anything that matters before you rely on it.

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