Additionally, the use of ML systems in medical practices raises concerns about perpetuating existing inequalities, overestimating their capabilities and displacing human authority. While this is an ...open area of research,4 many advances have and continue to be made in ML interpretability and explainability, including in the medical domain.5 This type of formalisation can help to mitigate the ‘black box’ critique by requiring that models be built using well-defined and transparent methods, with clear inputs, outputs and decision rules. ML systems are commonly trained using large data sets, which often contain more information than any human could hope to remember or recall.
When AIDS was first recognized in 1981, most experts believed that
it was a plague, a virulent unexpected disease. They thought AIDS,
as a plague, would resemble the great epidemics of the past: it
...would be devastating but would soon subside, perhaps never to
return. By the middle 1980s, however, it became increasingly clear
that AIDS was a chronic infection, not a classic plague. In this
follow-up to AIDS: The Burdens of History , editors
Elizabeth Fee and Daniel M. Fox present essays that describe how
AIDS has come to be regarded as a chronic disease. Representing
diverse fields and professions, the twenty-three contributors to
this work use historical methods to analyze politics and public
policy, human rights issues, and the changing populations with HIV
infection. They examine the federal government's testing of drugs
for cancer and HIV, and show how the policy makers' choice of a
specific historical model (chronic disease versus plague) affected
their decisions. A powerful photo essay reveals the strengths of
women from various backgrounds and lifestyles who are coping with
HIV. A sensitive account of the complex relationships of the gay
community to AIDS is included. Finally, several contributors
provide a sampling of international perspectives on the impact of
AIDS in other nations.