In the film Jurassic Park, there is a moment shortly before carnivorous dinosaurs turn on the staff. One of the protagonists says something that Professor Benjamin Smart suggests should give us pause in the age of AI. The team was so caught up in whether they could clone dinosaurs that they never stopped to ask whether they should.
“To the best of my knowledge, we’re not currently on the edge of successfully cloning a dinosaur,” adds Prof Smart. “But the rapid development of AI may prove far more consequential to public health than an island of giant lizards with anger issues.”
Prof Smart is a researcher in the Department of Philosophy, in the Faculty of Humanities at the University of Johannesburg (UJ). He is also co-founder and director of the UJ Centre for Philosophy of Epidemiology, Medicine and Public Health.
He was giving his inauguration lecture at the UJ Kingsway Campus on Wednesday 5 August.
Although he accepts that AI may bring big gains in medicine, and a steep rise in what the economy can produce; regulating AI, argues Prof Smart, is not taken seriously enough by many of the people driving its development.
Lawmakers still have to ask a question that is difficult to answer. Would models that (in addition to their benefits) have the potential to replace human thinking, design dangerous germs, attack banks and control weapons of mass destruction, be good for us? These questions must be answered before these AI systems are operational, he says.
Health as a Property of Complex Systems
When a bee-hive collapses, the disorder belongs to the colony and not to a single bee. The bees may be in fine shape as individuals. But once they stop keeping the queen, the brood and the hive going, the hive as a system is sick. The same logic applies to human populations, he argues.
“The health of a population is not determined by adding up the heart rates, cholesterol counts and individual diagnoses of people. These are the capacities for health of individuals only.
Population-level capacities, he argues, determine the health of the population, and these emerge from the capacities of individuals, their environments, and their relational properties.
A new AI goal for public health
Prof Smart offers a new approach for public health policy in his lecture.
“A healthy population is not one in which nobody becomes ill, in the same way as a healthy person is not one in which no cell ever fails. Rather, it is a population with the relevant capacities to enable its members to flourish, to compensate for local failures, and to adapt without sacrificing its most vulnerable members,” he argues.
According to him, a healthy population has the capacities to provide accessible primary and preventative healthcare; clean water and sanitation; reliable electricity and transport; high-quality primary and tertiary education; and protection from crime and infectious disease. Lastly, a population with good public health has the capacity to sensibly regulate AI, he says.
“I accept that AI regulation does not, on the face of it, seem like an obvious public health goal, or even within a general domain of public health. However, when health is understood through the lens of capacities, it will soon become a significant contributor,” he says.
“Thinking in capacities also draws our attention to what might be called resilience or preparedness. An emergency stockpile can contribute to population health even when it is not currently being used, since reserves can prevent societal collapse.
“Public health policymakers must therefore ask not just how well the system performs under ordinary conditions, but also how it would perform under difference possible crisis scenarios.”
AI adds capacities – for good and bad
AI could quickly increase the capacities of what we can do together, he says. He points to early diagnoses of diseases, gene editing and a surge in economic activity. Then he encourages caution going forward.
“AI may also destroy the capacities of millions of individuals. It might lead to less free, less meaningful, less happy lives: unemployment, surveillance, misinformation, and the concentration of power in large-cap technology companies.
“The Sciences tell us what we can build and how to do it. The Humanities help us decide what is worth building,” he argues.
“Jurassic Park did not show us that scientific ambition is a huge mistake. Rather, it showed us that developing capacity without sufficient reflection is incredibly dangerous.
“Health emerges from individual and population level capacities. And so, before the next technological Tyrannosaurus rex reaches the fence, we must ensure that whatever we build expands our capacities to live healthy, happy and meaningful lives,” concludes Prof Smart.
