SHADE Newsletter 23rd January 2025

Welcome to the thirtieth 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 two weeks, bringing you a selection of the latest news, upcoming events, academic publications and podcasts in the SHADE space.

In this newsletter we have two highlights: First, two examples of digital health that could benefit low resource areas and second, two examples of the advantages of co-benefits approaches. We also take a sweep through the latest on AI’s healthcare promise and environmental impacts, check in on critical minerals, meningitis, misinformation, carbon credits and the LA wildfires, call out Amazon, take a look air pollution, mammals, microbes and spreadsheets and much more.  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].   

Highlight on digital health for low resource areas and the advantages of co-benefits approaches.

• The potential of wastewater metagenomics in Africa to give early warning on disease outbreaks and contribute to environmental impact assessments is outlined in this paper in PLOS Global Public Health. The paper also highlights how advances in data infrastructure make the technique ever more feasible in low resource areas without advanced labs. For another example of digital health for low resource areas, hear about AI for neglected tropical diseases in ‘what we’re listening to’ below.

• A data analysis in Sustainable Cities and Society makes use of the Cities Mission dataset - in addition to datasets on air pollution and sectoral contributions to air pollution - to address the question Are cities ready to synergise climate neutrality and air quality efforts? The study advocates for a co-benefits approach in urban policy-making to effectively address climate change and air quality challenges, and it emphasises the need for transdisciplinary research and governance to optimise outcomes and reduce trade-offs.

• Pathogens and planetary change, from Nature Reviews Biodiversity, looks at shared solutions for biodiversity and health.

News

• AI’s promise for healthcare is repeatedly invoked as President Trump, together with Big Tech and their financiers, unveil Stargate, ‘the largest AI infrastructure project in history’. Sam Altman of OpenAI says ‘we will see diseases get cured at an unprecedented rate’ whilst Larry Ellison of Oracle looks forward to AI enabling customised cancer vaccines with a 48 hour turnaround. AI’s environmental impact is not mentioned.

• Picking up the slack on AI and its environmental impacts, two articles, both of which also reference the huge potential of generative AI applications in the health sector: First, Data Center Dynamics thinks the jury is still out when it comes to AI and the green data centre paradox: Can we really power progress sustainably? Second, from MIT News, a Q&A on the climate impact of generative AI reveals how researchers and consumers can help mitigate the environmental impact of generative AI. One approach trialled in academic computing platforms supporting generative AI has been the introduction of a power cap. This resulted in minimal impact on performance and had the additional benefit of reducing hardware operating temperatures and consequently the need for cooling and maintenance.

• The FT reports on the next commodities supercycle in China: Fuelled by the demands of AI and the energy transition, attention is shifting from steel and iron (which fed China’s now slowing construction boom) to critical minerals such as copper, lithium, nickel and cobalt. Meanwhile, the Windhoek Observer reports on the toxic legacy left by copper smelting in Namibia.

• Mongabay highlights the link between climate change and meningitis outbreaks and VaccinesWork reports on how a new vaccine and DHIS2, the health information management system adopted in more than 70 countries, helped Niger beat a major meningitis outbreak in 2024.

• As misinformation overtakes climate and conflict as the biggest global risk, 2025 must be the year we join together to fight health misinformation says the director of the London School of Hygiene and Tropical Medicine.

• The FT looks at the predictions, data and politics around the Atlantic Meridional Overturning Circulation (Amoc) and its potential shutdown.

• The first comprehensive study on health and climate change in Small Island Developing States is out.

• The FT highlights Amazon’s challenges in bringing AI to Alexa, its voice powered digital assistant. Along with the enduring challenges of ‘legacy’ systems, the story focusses on the challenges of addressing AI issues such as hallucinations and emphasises Amazon’s intention to first and foremost, “deliver customer value and impact, which in this era of generative AI is becoming more important than ever because customers want to see a return on investment”. The environmental impact of allowing 500 million consumer devices worldwide access to generative AI is not mentioned.

What we’re listening to

• The World Health Organisation reported a year ago on snake and human distribution modelling under current and projected climate conditions. The results indicated changing patterns of human snake interactions which would put additional pressure on health services already struggling with insufficient supplies of effective antivenom. This Nature podcast reports on how AI-designed antivenoms could help treat lethal snakebites and highlights how such methods could substantially reduce costs and resource requirements for the development of therapies for neglected tropical diseases more generally..

• As Earth breaches the 1.5C limit for the first time, the FT reports on Big Tech’s latest efforts to curb their AI emissions with nature based carbon credits. Planet A exposes how such solutions have caused problems in Kenya.

• Catch Sasha Luccioni talking about AI and the environment on the New York Times Hard Fork podcast (it comes after a lengthy item on TikTok). In the intro the podcast hosts highlight the volume of social media linking the LA fires with AI. Calling on the tech industry to be more transparent about AI’s energy consumption, Luccioni concludes ‘let us use google in a way that is coherent with our values’.

What we’re reading

• Tabular data gets its own AI foundation model which can interpret any spreadsheet below a certain size. The model, TabPFN, overcomes the relative lack of tabular data (in comparison with the documents used to train large language models) by synthesising its own training data. It could be transformative in areas such as healthcare. Its environmental impacts are not quantified, although its computational requirements are described as ‘modest’, .

• Wireless micro electronic devices that are fuelled by light could massively accelerate drug development and reduce its environmental impact, reports Nature.

• The Lawrence Berkeley Lab evaluates the increase in electricity demand from data centers, finding that they are expected to consume between 6.7 and 12% of total U.S. electricity by 2028. Meanwhile, data (finally) on the efficiency of the major cloud providers - Adrian Cockcroft takes a look.

• From the ECO: DIGIT project, a Life Cycle Assessment methodology for assessing the environmental impact of digital services.

• With the LA wildfires following two years of record breaking rainfall, the FT reports on ‘hydroclimate whiplash’, noting that LMICs are most at risk and highlighting the dangers to public health systems designed for 20th century extremes. The story is based on this paper in Nature which calls for large scale climate modelling incorporating land-ocean-atmosphere feedbacks and emerging machine learning methods to better predict future volatility scenarios and inform responses that don’t risk exacerbating the problem. Meanwhile, Marketplace gives a firefighter based perspective as it highlights how, from satellites to AI, tech has a role to play in battling blazes and the Verge tells the story of Watch Duty, a wildfire tracking app that became a crucial lifeline for LA. The app relies on publicly available data.

• This Nature feature highlights the growth in evidence of the adverse effects of air pollution on the brain. Meanwhile a paper in Environmental Modelling and Software undertakes a systematic review of the application of AI in air pollution monitoring and forecasting.

• From The Lancet Regional Health, Small mammals and associated infections in China: a systematic review and spatial modelling analysis. The paper exposes the extensive intersection between small mammals and microbes that pose a pathogenic threat to humans. The paper generates zoonotic risk maps and highlights their climactic and geographical drivers.

• The challenges around the volumes of e-waste generated by the tech industry are well documented. But with Big Tech moving towards nuclear as an energy source, another sort of tech industry waste is looming - nuclear waste. Where Will All of Big Tech’s Nuclear Waste Go? asks Gizmodo, noting that in nuclear “when things go bad they go catastrophically bad’.

• From Environmental Science and Ecotechnology, A social-environmental impact perspective of generative artificial intelligence. The paper identifies key focus areas for discussing the social and environmental impacts of generative AI and highlights the challenges involved in identifying and quantifying these.

Events

• Tackling Climate Change with Machine Learning is an International Conference on Learning Representations (ICLR) workshop happening between April 24th and 28th in Singapore. Details on attending can be found here. The submissions deadline for papers and proposals is January 31st and health is a topic of interest.

• The next Oxford Global Health & Bioethics International Conference will take place on the 8th and 9th July 2025. The call for abstracts is now open and the deadline for submissions is February 5th.

• Hype, hallucination, hope – what might AI mean for our health? asks Hetan Shah, Chief Executive of the British Academy, at a lecture on February 13th. The hybrid event will take place at in London.

• The 2nd Annual CAFE Climate and Health conference is happening virtually March 3rd to 6th. Pre conference workshops will happen on March 3rd - many have a digital/data component.

• Do we know enough about the risks and problems of using digital technology in health and care? Register for the Downsides of Digital on Monday January 27th. This free lunchtime event comes from the Strategy Unit in collaboration with the Health Foundation.

Opportunities

• The European Space Agency has opened applications for funding for projects with a focus on climate change and health. The aim is to explore the role of Earth observation data in identifying, monitoring and reducing climate-related health risks. Submission deadline is February 14th.

• There is less than a month left to submit an application under Wellcome’s Climate and Health Award: Advancing climate mitigation solutions with health co-benefits in low- and middle-income countries.

• For anyone working with DHIS2, the deadline for submitting an abstract for an in person presentation or an online poster at this year’s DHIS2 Annual Conference is February 2nd.

And finally, this News Feature from Nature asks Can quantum computing crack the biggest challenges in health?

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