What’s artificial intelligence (AI) got to do with it—inequality and public health?

Abstract

‘We must work together for AI that bridges social, digital, and economic divides, not one that pushes us further apart.’1 In his article ‘Getting AI Right’2 James Maniyka writes: ‘While the headlines tend to feature results and demonstrations of a future to come, AI and its associated technologies are already here and pervade our daily lives more than many realize. Examples include recommendation systems, search, language translators–now covering more than one hundred languages–facial recognition, speech to text (and back), digital assistants, chatbots for customer service, fraud detection, decision support systems, energy management systems, and tools for scientific research, to name a few’. He continues that in 2006 Nick Bostrom,3 director of the Future of Humanity Institute at the University of Oxford noted, “a lot of cutting-edge AI has filtered into general applications, often without being called AI because once something becomes useful enough and common enough it’s not labelled AI anymore”. The term AI is a complex one. Marvin Minsky4 called it a ‘suitcase word’ which according to him ‘is packed variously, depending on who you ask’. I don’t claim to understand the complex world of AI. My understanding of it is as outlined in the introduction of the UNESCO Artificial Intelligence and Gender Equality report5—‘Simply put, artificial intelligence (AI) involves using computers to classify, analyze, and draw predictions from data sets, using a set of rules called algorithms. AI algorithms are trained using large datasets so that they can identify patterns, make predictions, recommend actions, and figure out what to do in unfamiliar situations, learning from new data and thus improving over time. The ability of an AI system to improve automatically through experience is known as Machine Learning (ML)’. However, AI is all pervading and if you believe some of the websites and CEOs of AI companies6,7 it is offered as a panacea for a range of issues from ending poverty to reversing climate change. There is an increasing body of research on the impact of AI on inequality. One of the issues debated is whether AI technologies and their applications may, at least over the short to medium term, increase inequality due to AI automation.8,9 Research suggests that AI, which is expected to be make objective and unbiased judgments ends up reproducing the biases and inequalities of the societies they are ‘trained’ on. ‘Fairness, accountability and transparency’ has become a major research agenda in the field of machine learning.10 Inadequate training data have resulted in inaccurate AI models for disease risk assessment, prognostic prediction, and medication usage for the data-disadvantaged populations.11–14 Research has documented the ways that ‘automated decisions are depriving people of government benefits, discriminating on the basis of sex, skin color, age and numerous other forms of difference, choosing who is surveilled, who is imprisoned, or who is targeted for economic exploitation’.14 According to David Leslie et al13 ‘The impact of covid-19 has fallen disproportionately on disadvantaged and vulnerable communities, and the use of artificial intelligence (AI) technologies to combat the pandemic risks compounding these inequities’. Fisher and Rosella15 are of the opinion that ‘AI has the potential to improve public health’s ability to promote the health of all people in all communities’. In their article ‘Using artificial intelligence to improve public health: a narrative review’ David B. Olawade et al16 contend that ‘the ability of AI to draw together various datasets allows insights to be made that would otherwise be difficult to develop with traditional methods’. They are of the opinion that ‘Algorithms can be employed in an iterative manner, with results of policies monitored, and subsequently informing and improving future policies. Leveraging these insights will allow policies to be developed that are better-targeted, impactful and timely’. They also warn that though ‘AI has a great deal to offer policymakers but, like all new technologies, trust and education in how to use it effectively and responsibly are critical to its future uptake and usefulness’. There is some evidence that AI has uses in public health from surveillance to developing targeted policy. However, ‘the considerations for public health are distinctly different from those for clinical or health care systems’.15 To quote UN’s António Guterres in his special address at the World Economic Forum17 ‘This technology has enormous potential for sustainable development – but the International Monetary Fund has just warned that it is very likely to worsen inequality. Powerful tech companies are already pursuing profits with a reckless disregard for human rights, personal privacy, and social impact. This is no secret. These two issues – climate and AI – are exhaustively discussed by governments, by the media, by leaders here in Davos. And yet, we have no effective global strategy to deal with either’. ‘Contrary to other AI applications where the goal is to identify a single model to beat a single metric and benchmark, in situations where vulnerable populations are at risk, a more sensitive approach should be taken with a focus on outliers as metrics to assess failure as well as success’.11 The CIFAR AI for Public Health Equity report18 recommends ‘collaboration of public health professionals and researchers with computer science and AI researchers, in addition to a wide range of other groups (e.g., sociologists, political scientists, engineers, civil society and citizen scientists, people with lived experience, policymakers) to help ensure health equity when using AI technologies’. Public Health has a role to play to garner the possibilities and benefits of AI while carefully considering the challenges and unintended consequences of AI working in collaboration with a range of experts and civil society to explore and exploit the potential of AI to ‘improve public health’s ability to promote the health of all people in all communities’.15

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.