Four For Friday | March 14, 2025
LF164 | National Grid of AI Innovation, digital triplets, David Brooks on the failures of the elite, Discovery's new personal health pathways + MIT's AI Risk Repository.
Welcome to this week’s Four For Friday. Four things that have piqued my interest this week, together with a bonus: AI Tip of The Week.
A national grid of civic learning systems
This paper from the UK’s Northern Health Science Alliance proposes creating a "national grid of civic learning systems". Essentially this is looking to connect the UK’s disparate, devolved regions together in a trusted way that allows them to deploy tailored services locally but al learn from each other. This aims to boost the UK economy while reducing regional inequalities.
“Compared to the South, people in the North of England die earlier, spend less of their lives in good health, and have just a third of the median household wealth. This results in a productivity gap that costs the UK around £13.2bn annually.”
The model (above) envisions networked local systems where public services continuously improve through AI-assisted learning loops at population, provider and person levels.
The So What? This framework takes the concept of Civic AI further to suggest we can thread the needle of local trust and scaled insights by harnessing AI-powered collective intelligence.
Mastered digital twins? Get ready for digital triplets
This overview of digital triplets suggests adding AI to the ‘digital twins’ to create another dimension.
Digital twins are virtual replicas of physical assets, whereas a “digital triplet" adds generative and explainable AI.

So the triplet looks like:
The physical entity
The digital twin
Agentic AI-based intelligence (using both generative AI (GenAI) and explainable AI (XAI)).
Digital triplets combine generative AI for scenario creation, explainable AI for transparent communication, and multi-agent systems for data validation—enabling informed decision-making through natural language interaction with digital twins.
For organizations already committed do digital twins, adding AI seems like a good way to returns on digital spend.
The So What? Whether or not the extra dimension warrants new branding, the idea that AI can be added on to digital twins is powerful.
David Brooks on how the educated elites have failed us
In this thought provoking talk, David Brooks (who describes his role as the conservative commentator in the New York Times as being the Chief Rabbi at Mecca) makes some powerful arguments about the “failures of the educated elite"“:
Economic & Social Inequality – The meritocratic system was rigged to benefit the elite, creating a ‘caste system’. Wealthy children are 77x more likely to attend Ivy League schools, and those without a college degree die 9 years earlier.
Moral & Social Fragmentation – By ‘privatizing morality’, the elite dismantled shared values, leading to a decline in trust and rising despair. Mental health issues have surged—45% of high school students report persistent hopelessness, and social isolation has quadrupled since 2000.
Institutional Breakdown – Instead of building, modern elites focus on destroying left-leaning institutions, creating political chaos. Many figures labeled as "populist" are actually anti-left, not pro-conservative, leading to incompetence, instability, and erosion of governance.
And since we’re talking about David Brooks, his latest piece on the trashing of America’s brand among allies is worth a read, especially by those of us bullish long-term, on the country. It’s summarised by this phrase: “we have to think of a new global architecture.”
The So What? It’s worth searching for root causes and understanding of the seismic shifts underway, and identifying implicit assumptions about how the world works. Then trash them.
Discovery’s Adrian Gore on Personal Health Pathways
South African insurance company Discovery Health is well-known for their Vitality program that rewards customers for healthy choices. They’ve now collected over 80m person-years of data about people’s lifestyles, and their health outcomes, to power a new Personal Health Pathways (PHP) AI platform.
This delivers individualised health recommendations to over 2m members. It analyses its large clinical and behavioural datasets to generate unique action sequences / pathways for each individual. I’ve been advocating for a while the concept of “personalised purpose pathways” to help people reach their goals, and this is a big step in that direction.
Demonstrating its effectiveness, in tests, six "digital twins" of healthy 60-year-old males with similar demographic profiles each received entirely different recommendation sequences.
The technology calculates recommendations based on three key metrics: current health status, completion probability, and potential long-term health impact. The system, accessible via Discovery's mobile app, represents a significant advancement in precision preventive healthcare.
The So What? Making recommendations personalised and actionable is going to help with behaviour change. But… I’d love to know how explainable their recommendations, or is it a black box?
Bonus: AI Resource of the Week
MIT Risk Repository. MIT manages this AI Risk Repository that looks at over 1000 risks, extracted from 56 existing frameworks. It looks at how, when, and why these risks occur, and classifies them into 7 domains (e.g., “Misinformation”) and 23 subdomains (e.g., “False or misleading information”).
That’s all for this week. As always, feedback welcome. Feel free to share insights or links of interest.
- Stephen
do we need other meritocracy? or other system, which one?