Beyond the Kingdom of the Sick

In Illness as Metaphor, the American essayist Susan Sontag writes:

Illness is the night-side of life, a more onerous citizenship. Everyone who is born holds dual citizenship, in the kingdom of the well and in the kingdom of the sick. Although we all prefer to use only the good passport, sooner or later each of us is obliged, at least for a spell, to identify ourselves as citizens of that other place.

Sontag’s work analyzes how diseases acquire metaphorical meanings that reflect cultural anxieties and often burden patients with unwarranted stigma.

Writing in 1970s America, Sontag contrasts tuberculosis and cancer as emblematic Western diseases — the former, an almost allegorical epidemic of the 19th century, and the latter, the primary health challenge of the 20th.

In the Western imagination, tuberculosis was linked to the idea of an ethereal condition that consumed the body. The disease was contracted in industrial cities and was treated through bucolic retreat. The archetypal tuberculosis patient was portrayed as a poet or artist, withdrawing to sanitariums for care.

Cancer, conversely, became the dominant health challenge of the 20th century, carrying metaphors of impurity, failure, and moral breakdown.

Sontag, herself a breast cancer survivor, uncovers the aggravating effects that these metaphorical associations may have on patients. According to Sontag, cancer is seen as an a sign of weakness, an entity that colonizes the body, deforming and defeating it. In a later work, AIDS and Its Metaphors, she extended this analysis to the HIV epidemic, examining how social judgment burdened infected individuals.

As someone trained in biomedical sciences, I find Sontag’s essay enlightening. She is best known for her thinking on photography, which exposed the dangers of objectifying subjects. Illness as Metaphor has similar implications for those of us in healthcare — it provides guidance on how to best portray the patient.

The underlying message of Illness as Metaphor is that metaphors thrive with ignorance.

Scientific understanding diminishes these metaphorical associations. When Robert Koch identified tuberculosis’s bacterial cause in 1882, its romantic mystique began to fade. Similarly, advancing molecular understanding of cancer has reduced its stigma. Scientific progress allows patients to navigate illness with reduced social burden.

Personally, I prefer this view over the more futuristic one that envisions medicine as a means of augmenting ourselves. This framework also illuminates global health inequity. My research focuses on neglected diseases that affect the Global South, where diseases like tuberculosis, now rare in the West, continue to impact millions of individuals each year. In this case, the main preoccupation is equality and fair access to healthcare for all, in settings where resources are critically lacking.

Many of us are convinced that AI will have profound and transformational effects on medicine. AI will aid in the discovery of new treatments for unmet medical needs, the development of better diagnostic tools, and, overall, the improvement of life quality globally. I would have loved to see what Sontag has to say about the penetration of AI in virtually every aspect of our lives, including the post-photographic creations generated by text-to-image algorithms, and the progressive incorporation of the technology in the life sciences.

Most AI narratives, and the aesthetic and vocabulary surrounding them, are marked by the allure of acceleration, augmentation and, ultimately, the promise of liberation from our own physical and intellectual boundaries. However, the value of AI may lie less in futuristic enhancement than in addressing fundamental healthcare needs. The Ersilia Open Source Initiative exemplifies this approach, developing AI tools for biomedical research in resource-limited settings, particularly in Sub-Saharan Africa.

The overarching goal of Ersilia is to aid in discovering treatments for diseases often linked to poverty and therefore neglected by pharmaceutical companies and the global research agenda due to the low economic incentives.

Diseases such as schistosomiasis, leishmaniasis or cryptosporidiosis, to name a few, are caused by pathogens about which we know very little and for which treatments are sorely lacking. As such, these diseases carry with them stigma and dehumanizing metaphors, in Sontag’s terms, that cause unnecessary suffering and pain. Many of the conditions that affect low-income countries are, or should be, already curable or preventable.

Compared to wet-laboratory research and classical decision-making procedures in hospitals, AI can dramatically reduce costs and, ultimately, improve a global health infrastructure that has proven to be insufficient so far. Ersilia is an AI-first initiative devoted to this goal, which is not concerned with futuristic or solutionist views, but with slowly creating a sustainable research environment for the global majority, where access to medicine is commensurate with the standards that high-income countries have enjoyed for decades.

In Africa, in the midst of all the hype about AI and the so-called fourth industrial revolution, the challenges that we encounter are often related to electricity power cuts, internet connectivity, lack of training, and brain drain. In a rather crowded landscape of AI research that primarily rewards novelty, boosts in accuracy and formidable investments in technology, the speed and scale of progress is making adoption difficult in settings where the infrastructure is not available and the necessary communities are not in place for the uptake.

In healthcare, especially in healthcare applied to the Global South, we’ve found more value in deploying already-existing AI tools in places where AI would otherwise not be available, such as small biomedical research laboratories that operate under severe funding constraints. The focus on AI innovation rather than implementation is often distracting, and sometimes paralyzing.

Science should be proactive about dispelling misguided beliefs about illness. All the cultural baggage associated with diseases is captured in the current foundational AI models, which calls for careful evaluation of those models much like they are examined for gender or racial biases. Likewise, the highest standards on data privacy should be met, and open sourcing should be supported to guarantee widespread adoption, especially in less privileged settings.

Unlike AI applications focused on enhancing digital interactions, biomedical AI aims for tangible health outcomes. Many necessary tools exist but remain unknown to the broader developer community. Increased awareness and incentives could direct more talent toward global health applications. In a way, Illness as Metaphor shows that biomedical progress is inherently beneficial. This is one area where AI, if applied conscientiously, can bring undeniable good.