Bundle of Rays

23/09/2026

There's a photo doing the rounds of maths teachers protesting calculators. Easy to laugh at. But they weren't fools. They were defending something they believed made a student competent.

In the past month I've been in conversations with senior nursing teams where AI came up, and the first response was caution. Is this a gimmick? That's a fair instinct. Scepticism has protected patients for a long time.

But calculators automated a task. AI reshapes the reasoning around it. That's a different kind of change.
We tell junior nurses that nursing is patient-centred care. Then they spend their shift filling in forms that exist to audit the hospital. We already let someone else decide what our time is for.

Australia needs around 75000 more nurses by 2035. We won't get there by recruiting harder. We get there by making this a profession people want to stay in, where a new graduate is nurtured rather than processed.

The role isn't fixed. It's changed enormously in thirty years, and nurses changed it.
So the question I'd rather ask is, what kind of profession are we inviting them into?

Because that's still ours to decide.

23/09/2026

No one would publish a drug trial that said "we gave participants a medication." You'd state the drug, the dose, the route, the frequency, the duration. Without that, no one can replicate it, prescribe from it, or pool it with anything else.

Now look at how technology gets reported in nursing research. "Participants used ChatGPT." Which version? Single question and answer to a LLM, or a multi-turn agent with guardrails? "Participants used VR." A 360 video they looked around from a chair and simply clicked or a room-scale simulation where they walked, reached and handled equipment to conduct a skill?

And here's the one that should worry us most. Ask a large language model the same question twice and you can get two different answers. That variability is a property of the technology, not a glitch. Yet across hundreds of nursing AI papers I've read, almost none of them account for it. There's no record of whether outputs varied, or what participants actually saw. If every participant potentially received a different intervention, what exactly is being generalised?

The cost isn't academic. Internal validity suffers, because you can't tell what actually produced the effect. External validity suffers, because no one can tell whether it would work in their setting. And we now have databases full of studies that can't be translated into practice, for no reason other than the intervention was never properly described.

So I'm kicking off DEFINE (Describing Essential Features of Interventions in Nursing Evidence). It's twin reporting checklists, one for AI and one for XR. DEFINE is now registered with the EQUATOR Network and follows their four-phase process: structured review, multidisciplinary Delphi, consensus meeting, then piloting on real papers.

Report the technology, not the label.

Early days, long road ahead through consensus. More to come.

17/03/2026

Wrapping up a few days in Townsville with the team at James Cook University's School of Medicine.

This trip has been packed with quality meetings across leadership teams and wider faculty, exploring how XR, AI agents, and immersive content can transform healthcare education — from occupational violence training through to curriculum-integrated simulation.

A huge shout out to Nikki Harvey, whose research project is at the heart of why we're here — investigating how immersive content can be embedded into the distributed medical undergraduate curriculum. This is our core research collaboration, and the potential to improve outcomes for students training across regional and remote sites is exactly what gets us out of bed in the morning 😉👍

The conversations have been sharp, the energy has been real, and I'm leaving with a full notebook and a lot of momentum.

Townsville, you've been unreal. 🌴

16/02/2026

Extended Reality in Higher Education:

- What 295 Studies Tell Us -

A 2025 systematic review published in TechTrends analysed 295 peer-reviewed studies (2020–2024) examining the use of Extended Reality (XR) in higher education.

The findings are worth pausing on.
🔹 VR dominates (58%)
🔹 81% of research focuses on undergraduate learners
🔹 Studies span 48 countries across 6 continents

XR is no longer emerging. It is established.
But here’s what’s more interesting.
The research clusters around three learning domains:
• Knowledge acquisition (51%)
• Psychomotor skill development (35%)
• Affective outcomes (14%)
Much of the public conversation frames XR as an “engagement tool”.

The evidence suggests something deeper.
In healthcare and other high-stakes disciplines, XR is being used to develop procedural competence in safe environments — tasks that are high-risk, rare, or difficult to scale consistently.

The barriers are also clear:
– Cognitive overload
– Authenticity concerns
– Cost and maintenance
– Connectivity and classroom management

The technology itself is not the limiting factor.
Implementation strategy is.

The next phase of XR in higher education will not be defined by better headsets.

It will be defined by:
• Faculty capability
• Deployment models
• Curriculum alignment
• Sustainable commercial frameworks

With data like this, I honestly feel we are now past the experimentation phase.

The question is no longer “Does XR work?”

The question is:

How do we scale it responsibly?

Burke, D., Crompton, H., & Nickel, C. (2025). The Use of Extended Reality (XR) in Higher Education: A Systematic Review. TechTrends.

15/02/2026

I keep hearing the same line:
“AI can assist nursing, but it can’t replace it.”

And I understand it. Nursing is human. It’s presence, judgement, sitting with someone when they’re scared. No machine replicates that.

But I wonder whether “can’t replace” is a truth — or a comfort line. History shows us that structural shifts don’t ask permission.

Horses didn’t negotiate with cars.
Blockbuster didn't outpace Netflix

They were replaced because the system evolved.

So the real question isn’t what AI is capable of in 2025.

It’s what it may be capable of by 2035 — and whether nursing leadership is preparing for that horizon.

AI one day triages more accurately than an overstretched ED team…
it conducts consistent first-pass telehealth assessments…
it detects deterioration earlier through pattern recognition…
it provides meaningful companionship in moments of isolation…

Do we pre-emptively limit it to “assistant”?

Or do we rigorously shape it to improve care? Full view, open mind, utter industry change?

Australia will need tens of thousands (75,000) more nurses in the coming decade. Demand and complexity are increasing. Burnout is real.

I feel leadership isn’t about defending current boundaries. It’s about designing the future responsibly.

“Should AI replace nurses?” is the wrong question. It immediately frames a fight.

A better one could be:
Can AI improve patient care?

If the answer is yes — even partially — then our responsibility is not fear or protectionism.

It is preparation.

The 2035 conversation starts now.

For the comments -

Are we designing nursing for 2035 — or defending it in 2025?

AI HealthcareInnovation

12/02/2026

I just finished reading a recent systematic review on the plane on the way home looking at AI in teaching and teacher professional development.

It analysed 95 studies over the past decade.
What struck me wasn’t what it said about tools. It was what it revealed about imbalance.

Most of the research focuses on how AI is being used in classrooms — chatbots, automated feedback, lesson planning support, assessment generation.
Far fewer studies look at how educators themselves are being supported to develop real capability with AI.

And that gap feels even more significant in nursing.

In our world, AI isn’t just helping draft lesson plans. It’s creeping into documentation systems, triage processes, decision support tools, predictive analytics, simulation design.
We’re talking about systems that can influence clinical thinking.

Yet professional development often still looks like:
“Here’s how to use ChatGPT.” “Here’s a quick demo.” “Here’s what not to do.”
That’s not development. That’s orientation.

The review also points out that most AI-related professional development is short-term. Intervention-based. Measured immediately afterwards. Rarely longitudinal.
But judgment — especially clinical judgment — doesn’t develop in a workshop.

In nursing education and hospital systems, the real question isn’t whether students will use AI.
They 100% will.

The question is whether nurse educators and clinical educators feel confident enough to:
Know when to trust AI.
Know when to question it.
Know how to teach students to override it.
Know how to discuss bias, error, and algorithmic limitation without fear or defensiveness.

AI literacy in healthcare cannot be about prompts. It has to be about professional responsibility. If AI is going to shape the future of care, then educator development has to come first — not as an afterthought once the software is installed.

Nursing has always adapted to complexity.

This is just the next layer.

Reference:
Tan, X., Cheng, G., & Ling, M. H. (2025). Artificial intelligence in teaching and teacher professional development: A systematic review. Computers and Education: Artificial Intelligence, 8, Article 100355. https://lnkd.in/gFzZKMuy

11/02/2026

India, you’ve been extraordinary. 🇮🇳

Wrapping up what has been one of the most energising and productive trips we’ve had.
From healthcare leaders to education innovators to government stakeholders — the appetite for technology-enabled transformation is real. And more importantly, it’s practical. It’s action-focused. It’s moving.

A huge shoutout to NetbookFlix and TimesPro — our newest partners for Bundle of Rays across India.
Together, we’ll be working across education and government-focused initiatives to accelerate immersive learning, AI-enabled education, and workforce readiness at scale.

This trip has secured four national MoUs across multiple sectors.

That’s not small.
But MoUs are only the beginning.
Now it’s time to build.
To deliver.

To execute with discipline and intent.

What stood out most on this trip wasn’t just opportunity — it was openness. A willingness to pilot. To move fast. To collaborate across sectors. To focus on impact rather than optics.
Healthcare across India is ready for innovation.
Education is ready for immersive, simulation-driven learning.

And technology — when built with practitioners — can genuinely lift standards at scale.

Proud of how far we’ve come.
Grateful for the trust.
But now it’s time to take it further.
2026 is going to be big.

Simulation NursingLeadership GlobalPartnerships

09/02/2026

08/02/2026

I’ve just come from a conference focused on clinical readiness, and it reinforced a question I keep coming back to.

We talk a lot about new graduate nurses lacking industry readiness. The literature is full of it. But the harder question is:
are universities and clinical health services actually giving graduates and developing experts the foundations required for the complexity of modern nursing practice?

A recent paper by Suriano et al. (2025) is useful here—not because it claims AI is good or bad, but because it refocuses attention on the endpoint. The study shows that how students engage predicts critical thinking performance more strongly than what they simply know. For nursing, that maps directly to clinical decision-making and clinical judgement, not abstract reasoning.

We’ve made strong progress with simulation and OSCE-style assessment, particularly for procedural skills. But we need to ask honestly:

are we truly assessing how nurses make decisions under uncertainty, or are we still rewarding task completion more than judgement formation?

Because this isn’t 1995.

New graduates are entering environments defined by large-scale digital systems, ageing populations, multi-morbidity, constant cognitive load, and rapid escalation expectations. Clinical judgement now has to develop earlier, faster, and more deliberately—before nurses are placed in high-risk settings.

I feel strongly that this is where educators and nursing leadership need to pause—not to debate AI (technology) adoption, but to rethink what industry readiness actually means. How do we teach clinical judgement explicitly? How do we assess it in ways that reflect modern practice?

AI didn’t create this problem—but it has removed our ability to ignore it. Polished answers are no longer reliable proxies for judgement.

Change is never easy, particularly in healthcare. But this feels like a moment where progression isn’t optional—it’s necessary.

Article for rumination here - https://lnkd.in/gMDRJYwn

07/02/2026

How do you travel to your meetings?