SHADE Newsletter 5th June 2025

Welcome to the thirty eighth edition of the SHADE newsletter! 

SHADE is a research hub with a mission to explore issues at the intersection of digital technologies/AI, health and the environment. It is guided by a fundamental question: How should the balance between AI/digital enabled health and planetary health be struck in different areas of the world, and what should be the guiding principles?

The SHADE newsletter comes out every month, taking an in depth look at selected topics, as well as highlighting new resources, events and opportunities in the SHADE space.

In this newsletter we highlight the benefits of, and threats to, data sharing and the challenges posed by wealth and resource disparities, and how they can be addressed. We conclude with round ups of recent developments on three themes: One Health, how AI is changing medicine and AI’s environmental impact and how it can be addressed. We hope you enjoy it!

Please tell us what you like, what you don’t like and what you think is missing at [email protected].   

Data Sharing

• A generative AI model, Foresight, is being trained on a set of de-identified NHS data for 57 million people in England. The model learns to predict what happens next based on previous medical events. The Health Data Research Service, recently announced by the UK government (see the April 17th edition of this newsletter), is designed to facilitate more projects like Foresight going forward.

• The FT reports on how AI is creating a paradigm shift in meteorology, providing earlier warnings of extreme weather and reducing the disparity between forecasting for the global south and the global north. However, public access to data is likely to affect this and there is concern about the Trump administration’s impact on the US National Oceanic and Atmospheric Administration (NOAA). This concern is also reflected in this Nature article examining the impact of the Trump administration’s science cuts on Canada.

• Association between national action and trends in antibiotic resistance: an analysis of 73 countries from 2000 to 2023: This paper from PLOS Global Public Health uses publicly available data from the Global Database for Tracking Antimicrobial Resistance to highlight the importance of sustained national action.

• This article in Nature gives a round up of the use of AI for voice biomarkers, including attempts to create publicly available voice data sets.

• This study in Nature Scientific Reports shows that pet insurance claims can predict the occurrence of vector-borne and zoonotic disease in humans in the United States. Global climate change plays a part in the spread of Lyme disease, perhaps the best known of the diseases studied. The study made use of publicly available Centre for Disease Control and Prevention (CDC) based data on human cases for Lyme disease, giardia, and Valley Fever.

Wealth and Resource Disparities

• SHADE co-directors and members make up part of the core team working on a three year Wellcome funded project on environmentally Sustainable HeAlth REsearch (SHARE). The project sets out to move beyond ‘tool-solutionism’ to a more context-sensitive, reflexive and just health research.

• High-income groups disproportionately contribute to climate extremes worldwide. This study in Nature Climate Change quantifies the link between wealth disparities and climate impacts.

• How climate change will burden our children: data reveal a lifetime of extreme heat. This article from Nature brings the impacts of climate change into sharp focus, reporting on a study showing how the risks of experiencing extreme heat are rising with each generation, and how these risks are higher amongst more deprived populations. The article calls for climate change mitigation and adaptation now.

• From the Lancet Planetary Health, Post-growth: the science of wellbeing within planetary boundaries. This paper reviews recent advances in post-growth research and highlights outstanding questions, including those around transforming the relationship between the Global North and the Global South. Meanwhile Wellcome calls for reform of the global health system, arguing that it should not be based on charity and the global north should become ‘partners not fixers’.

• A Nature collection on AI models in the Global South examines how researchers, engineers and public institutions are building different kinds of AI systems ‘grounded in local empowerment, digital sovereignty, and the ethical collection of data’.

• A call for papers “on, for, or with DHIS2”: Building Stronger National Health Information Systems: Leveraging DHIS2 for Health Systems Performance, Policy and Research. The submission deadline is the 31st August and submissions from LMICs are particularly encouraged.

Round Ups: One Health, AI in medicine, AI’s environmental impact and how it can be addressed

• A selection of items with a One Health theme: Following genomic analysis reported in this paper in Nature Communications last year, this episode of the One Health podcast talks to one of the paper’s authors and looks at what mammal to mammal transmission of the H5N1 avian influenza virus could mean for every animal on the planet, and the approach we should take to managing this virus. Meanwhile the Conversation reports on the hidden connections of more than 100 migratory marine species revealed in interactive map. ‘This research synthesises thousands of records of more than 100 species of birds, mammals, turtles and fish that connect almost 2,000 crucial habitats. And finally, another episode of the One Health podcast asks What’s at Stake in the Cutting of US Biodefense? Highlighting the suspension of surveillance of many diseases globally, it looks at the risks around ongoing neglect of biodefense.

• Some of the latest stories on how AI is changing medicine: First, Nature reports on Google’s LLM based medical chatbot which can outdo human doctors in diagnosis using medical imagery, including electrocardiograms and PDFs of laboratory results. The article reflects the view of many that “large language models for diagnosis would be the way to go in the future”. Next up, this long read from Nature looks at the arguments for a ‘genome first’ approach to rare disease diagnosis, noting that, on average, people with rare diseases currently spend 5 years in the diagnosis journey when up front Whole Genome Sequencing (WGS) could provide much faster answers and save money. Finally this article in Nature describes an AI tool that can flag people at high risk of post-partum depression.

• A round up of recent stories on AI’s environmental impact: First up, MIT Technology Review does the math on AI’s energy footprint. Next up, AI is draining water from areas that need it most, reports Bloomberg whilst the MIT Technology Review zooms in on the data center boom in the desert, east of Reno, Nevada, and its implications for water strains as well as public health. Meanwhile, a round up of ways to address AI’s environmental impacts: Sasha Luccioni, writing in the FT, says ‘making the energy use of AI more transparent is essential’ describing the energy use of some AI models as ‘like turning on stadium floodlights to look for your keys’. Next up Reduce, Reuse, Recycle: Why Open Source is a Win for Sustainability: This community article from Hugging Face explains the environmental benefits of open source AI models from the fundamental principles of environmental conservation. This article from The New Stack provides a useful summary of the tools out there to make AI more sustainable - starting with asking whether AI is necessary in the first place - and this preprint gives the lowdown on Frugal AI, something also highlighted in the latest episode of Environment Variables. This paper from Lancaster University investigates another angle - Green Prompting for energy adaptive Large Language Models (LLMs). In From code to carbon: rewriting the rules for sustainable digital infrastructure, Professor Aoife M. Foley from the University of Manchester summarises the energy consumption challenges resulting from ‘bloated code, inefficient algorithms, and an economic model that treats data as virtually free’ and looks at how they can be addressed. Last, but not least, Jevon’s Paradox is Doomerism says Stategically Green, calling for AI platform builders and AI users to take action.

And finally, the AI Now Institute have published their 2025 Landscape Report, Artificial Power. Looking at the ways AI is built to be used not just by us but on us, the report sets out to offer concrete strategies for communities, policy makers and the public to reclaim agency over the trajectory of AI.

We hope you have enjoyed this newsletter. If it has been forwarded to you, and you would like to receive future editions, you can subscribe here