While I was a grad student at MIT, I was impressed by a number of things, including the role that technical solutions play in the local worldview. I’ve been thinking –in a broad and probably naive way– about how and when technical solutions can work. I have two anecdotes.
During a symposium about microbiome science, I heard some MIT folks present very impressive genetic engineering work whose avowed purpose was to help treat cholera. (I really don’t want to denigrate this work at all; take my thoughts on this as a broader cogitation rather than as a critique on anyone’s particular work.) The idea was that this engineering would produce a diagnostic for cholera infection, and the team mused that this would be helpful for treating the disease.
After the presentation, a doctor who practices in Nepal stood up and said that this work wasn’t a useful solution. (I beg forgiveness if I misrepresent his words here.) The onset of cholera is very sudden (you go from perfectly healthy to really sick on the order of hours), it’s very easy to diagnose (look for “rice water” stool), and it’s very easy to treat (oral rehydration therapy). The trick is that you need to get treatment quickly: death can follow the onset of symptoms also in a few hours. So the problem isn’t about getting the diagnosis earlier; it’s about being sure that the person, once they have symptoms, has access to care.
After some response from the speaker, another doctor stood up –this one who had practiced in India– and basically repeated the points of the first doctor.
So what I saw here was that it’s really tempting for engineers and scientists to solve a technical problem, which is easier to formulate in an academic environment than finding a technical solution to the “real” problem.
My transition from MIT to a school of public health made another deep impression. Broadly speaking, public health practitioners are interested in using what tools we have to solve the problems we have. Often the existing tools work really well, but we have a hard time implementing them.
Take infectious disease. Antibiotics are really incredible drugs, but most of the reduction in infectious disease that the developed world has experienced is due to improved sanitation. People in the developed world mostly don’t die of cholera not because of antibiotics or better diagnostics or better treatment, but just because their water sources are not contaminated by human feces. Most of the improvement in health was due to infrastructure, not to science or technology (excepting the fascinating science and engineering of water treatment plants, reservoirs, pumps, pipes, etc.).
Now take obesity. We have really effective tools: reducing caloric intake, improving nutrition, and exercise. We could treat obesity as a problem of infrastructure. Just as poor management of sewage was seen as a public problem requiring a public solution, we could see the ease of taking in many calories, having poor nutrition, and not exercising as a public problem requiring a public solution. (The idea of “food deserts”, which I don’t know if the evidence really supports as being meaningful for the obesity crisis, is nevertheless an example of this kind of infrastructure thinking.) But sewage stinks, and infectious disease is contagious, while it’s actually pleasant to eat, and obesity is not, strictly speaking, contagious. (We have a history of treating obesity as a problem resulting from personal choice, perhaps in the same way that cholera epidemics among immigrants to America was in some cases seen as a problem of their personal hygiene rather than as a problem with the plumbing system in tenements.)
I worry about technical solutions to obesity because we haven’t used the infrastructure solution first. Will we produce a drug that somehow cures obesity, and we end up continuing our unhealthy food infrastructure, and then piling on everyone needing to take an obesity-prevention drug? (And just as opioid use causes constipation that means you need another drug to deal with the constipation, will the obesity-prevention drug cause some side effect that you need yet another drug to treat?)
The interesting thing here is that the cycle of technical solutions makes money, while the infrastructure solution saves money. As Adam Smith pointed out, a solution that makes money doesn’t require political will, it just happens. So our capitalist system of organizing society could very conceivably lead us along locally optimal paths –from one profitable technical solution to another– while leaving us in a globally awful place.