Page 68 - UKZN Proceedings of the Conference Report

the relationship between health and environmental matters calls for a wider viewpoint in which the effects of AI technologies on the environment are assessed. In healthcare in particular the ethical concern that comes into play, as highlighted by Moyano-Fernández (2022), is that of AI-enhancing healthcare systems decapacitating energy resources and being unsustainable. This raises a fundamental ethical question: if technology endeavours to enhance health, surely it should not come at great cost to the environment. This highlights the need for an integrative approach to technology advancement and environmental protection.
The prospects of AI in healthcare are not limited to improving clinical results; it can also help mitigate the carbon footprint of the healthcare systems. For instance, there are many benefits of AI in disease detection, diagnosis, and prognosis, as well as in individual and public health surveillance (Briganti and Le Moine 2020; Matheny et al. 2020). These facilitate quick investigations, diverse treatment options, and resource allocation, all of which leads to efficiency in health outputs and the provision of healthcare services. Ho (2022) stresses that the introduction of IT systems which support One Health strategies will maximise the advantages while making healthcare activities more environmentally friendly than before. By converting these practices into electronic health records and telemedicine, healthcare could make use of fewer physical materials, use less energy, and generate less waste. Furthermore, Purohit et al. (2021) argue that AI- based telemedicine will reduce emissions of greenhouse gases by 80%, which can be viewed as a positive effect of AI in health systems. As healthcare systems continue to incorporate AI technology, so too does the capacity to enhance health improvements alongside environmental sustainability increase.
Nonetheless, AI has risks for the environment, which makes it complex. Van Wynsberghe (2021) remarks that this is because the realisation of such technologies calls for a lot of energy and resources, which aggravate factors like climate change or extinction of certain natural resources and species. AI systems such as those deployed in healthcare and environmental management have been traced to high energy consumption, leading to adverse carbon emissions and increased effects of global warming. According to Dhar (2020) and Samuel et al. (2022), the long- term sustainability aspects related to the application of AI would potentially negate its health benefits in the short term, if not carried out in consideration of the model’s ecological impacts. Other hidden costs of the use of AI, such as waste management and environmental devastation through energy use, must be included in assessment of the advantages of AI in healthcare and environmental issues. Thankfully, these issues are being addressed, as Ligozat et
al. (2022) indicate that new methods have been developed to evaluate the environmental effects of AI systems, both in terms of their use and the production processes involved. These measured sustainability principles are helpful to the developer in the assessment of the sustainability of AI technologies being developed, making sure that such technologies are beneficial to people and the environment. To put it briefly, while AI offers several avenues for enhancement of the healthcare system and protection of the environment, implementation must be done in a way that maximises benefits while minimising environmental impact. As pointed out by Glauner et al. (2021), the use of AI applications for early diagnosis and preventive measures against potential disease threats will optimise the health of the population as a whole; however, such considerations must come with an enabling infrastructure with a low carbon footprint. It is imperative to consider the aspect of sustainability that the technology embodies. The promise of AI is not only in transforming how we provide healthcare and protect the environment, it lies in the ability to provide such services without destroying the environment.
harnessing aI for comprehensive crisis management: findings and discussion
AI holds the potential to alter the unfavourable situations South Africa has found itself in over the years, and especially in the areas of energy, crime, employment, governance, ecology, and health. Lack of resources, and inadequate infrastructure provides a great opportunity for AI to improve management and service delivery in these sectors.
Energy crisis: AI in load shedding and grid optimisation
In 1994 the African National Congress (ANC) introduced a proposal called the Reconstruction and Development Programme. This included the promise of ‘Electricity for All’, which aimed to address the energy inequalities the country was facing. At that time only 36% of the people of South Africa were receiving electricity, and the aim was that another 2.5 million households would be electrified by the year 2000. This goal appeared to be possible due to the surplus power plants that were owned by Eskom in the 1980s, when fuel was available in the form of inexpensive and high-quality coal accompanied by infrastructure that was quite efficient. However, 30 years on Eskom struggles with over R400 billion in debt, having been bailed out by the state for over R270 billion after 2008. Load shedding, which was a distant concern, has turned into a disturbing reality as homes, services, and industries are affected, leading to problems in the South African economy (Amansure 2024).
To mitigate these persistent energy deficiencies, potential solutions may be gleaned from the application of AI and machine learning in the energy management systems in use around the world today. AI has great ability to change
66 | Proceedings of the conference on Public innovation, develoPment and sustainability