Four For Friday | Oct 17, 2025
LF192 | University Living Labs, Rings + bloods, remote patient monitoring's growth, Nano dementia + Claude Skills
Welcome to this week’s Four For Friday - nuggets of interesting things I’ve picked up this week, plus my favorite AI tool du jour. Bit delayed as am in the Bay Area and don’t benefit from Melbourne’s time zones to get this in your inboxes for the start of the weekend. Anyway, enjoy!
1. University Living Labs
A new report by Monash University into the evolution of university-based living labs. It defines them as: ‘relational infrastructure that enables people to work together to explore challenges, experiment, prototype and test concepts in real-world contexts’, and frames them as:
Experiments: targeted, time-bound initiatives that respond to specific societal challenges and test specific approaches or solutions.
Infrastructure: relational, organisational and material systems that support cross-cutting research, education, operations, and engagement.
Ecosystems: place-based networks connecting the university with communities, government, and industry
It identifies nine roles of these Labs:
Initiator and Convenor. Develops partnerships, convenes stakeholders, and secures initial funding.
Orchestrator and Facilitator. Coordinates activities, aligns priorities, and enables co-creation processes.
Knowledge Provider. Synthesizes and translates existing research into practical insights.
Knowledge Co-producer. Conducts action research and generates new knowledge collaboratively.
Learning Documentarian. Captures, documents, and shares learnings across the ecosystem.
Skills Provider. Offers specialized technical expertise to support innovation activities.
Capability Builder. Develops training programs to strengthen participant skills and capacity.
Infrastructure Provider. Provides facilities, testbeds, and access to physical infrastructure.
Commercialization Support. Identifies impact pathways and supports scaling of innovations.
The report argues that success will lie in these Labs being able to overcome three sets of challenges: governance, connectivity, and funding.
The So What? Universities are often bureaucratic, conservative (at least in terms of exploiting innovation) and slow moving - some see this as a feature not a bug. This report suggests another way - orchestrating an ecosystem to deliver meaningful real world impacts. Those who can fix their governance, connectivity and funding will thrive.
2. Oura goes for the jugular
Your Editor’s favourite wearable has for a while has been Oura. Its Finnish origins kept it under the radar for years, but as a product it’s always been a delightful experience and just worked (functional not hype). It went mainstream during Covid when celebrities started tweeting about how it was able to identify symptoms ahead of time. It’s now introduced a $99 blood panel for US residents that tests 50 biomarkers, and connects those insights with its AI coach, Advisor.
Whoop and Ultrahuman have made similar moves recently. The winner in all this - and a ripe target for innovation - is the test labs, since they all seem to be using Quest Diagnostics, where the user experience is (believe me) ‘varied’.
This will no doubt be a source of boardroom discussions in the fast growing device-less longevity analytics services such as Function Health and Superpower.
The So What? Further proof of the fast-growing consumer longevity market, and a shot in the arm for devices, who had been shunned as slow and expensive. Now they’ve got the “last inch” - direct contact with users, and they’re determined to exploit it.
3. Diving into “remote patient monitoring”
I’ve spent the past few days holed up in San Francisco with a group of smart, enthusiastic innovators passionate about poop. And the future of the bathroom more generally. This is for one of Fordcastle’s clients, a major bathroom brand.
One of the angles we’re looking at is the future of ‘remote patient monitoring’ and it’s a fascinating journey. One definition of this is a uniquely US-specific business model that provides reimbursement codes for approved devices to track certain conditions and report back to a clinician. The model is growing - nearly 50m Americans are now using RPM devices - but not universally popular. It’s been described to me as a headache, a stigma and a scam among other things.
Another view of this is as a description of health in the home, supported by, and empowering a plethora of use cases, business models. This is where most of us would want to receive care - in the home, on our terms.
This article sets the scene, and shows the traditional model is going gangbusters - projected to double from $14 billion in 2024 to $29 billion by 2030. Drivers are aging demographics, growing chronic disease management needs, and new Medicare reimbursement codes. 81% of clincians now use RPM, despite the concerns mentioned above.
There have been some standout results in certain use cases - for example, programs have reported a 70% reduction in 30-day readmissions and significantly better disease control.
The So What? RPM is growing, but a consistent theme - relevant for the first story and this one - is that patients and doctors are hungry for more actionable insights, not just dashboards and data.
4. Nano particles beat Alzheimer’s
This feels like the future is getting closer, and dementia may no longer be as untreatable as it has for decades.
Researchers actually reversed Alzheimer’s in mice using nanoparticles called “supramolecular drugs.” These repaired the blood-brain barrier, so it recovered its its natural waste-clearing function. Within an hour, amyloid-beta levels dropped 50-60%. Treated mice equivalent to 90-year-old humans recovered normal behavior and cognition. The therapy activates a feedback mechanism that brings clearance pathways back to normal levels.
The So What? Drugs and interventions that restore the human body’s own capabilities and self-balancing capabilities, rather than playing whack-a-mole with symptoms, seem promising.
Bonus AI Tip of the Week: Claude Skills
The buzzword of the week is Claude’s new ‘Skills’. These are a bit like OpenAI’s CustomGPTs (where you can save custom instructions rather than type them out in a prompt each time) but more powerful. With Skills we’re getting closer to real agentic AI rather than automation, since Claude makes decisions on when to bring them in to help address a task.
Skills are similar to Granola’s ‘recipes’ (see last week’s LF) which allow you to give a raft of instructions for how you do things - such as applying your company’s brand guidelines, building your standard operating procedures (SOPs), automating data analysis or building a cash flow statement in Excel. But not just instructions, a Skills folder will also have specific scripts (saving you the need to burn tokens) and resources (reference material). Skills save their instructions in markdown files and Claude automatically decides which ones to use based on your request, and can stack different Skills on top of each other, unlike Custom GPTs.
That’s all for now - happy weekend everyone.
- Stephen
Regarding the University Living Labs, do you think these "ecosystems" truly integrate with local communities, or remain a bit insular? Fascinating report, always apreciate your finds.
Your comment on "last inch" reminded me of Theranos, which held out the promise of direct consumer access to lab results. I have no desire to denigrate Oura by comparison with the nefarious actions of Theranos, only that I'm reminded of how difficult that "last inch" can be with anything in healthcare.