Opinion letter to the editor

How (not) to spend one billion dollars

We should do things not because they are hard, but because they are important

8749 opnphoto
Found in the Kendall T stop — an advertisement reads, "Using data-analytics to change car shopping? Sounds like fun."
Faraz Masroor

What are the challenges that people in the future will face? Climate change is going to be an issue, cities will be congested, diseases will still exist, nuclear energy sources need to be explored, and so on. All of these problems are in fields that MIT has great qualifications for — all of our departments, professors, students, and research are highly respected around the world.

But instead, we’re focusing on data science and artificial intelligence. These fields most definitely have applications with an obvious societal need — MIT GOV/LAB comes to mind, if you’ve heard of it — but most graduates in the field, from MIT or otherwise, do not tend to go down these paths. From my experience, they enter sectors like data mining and business analytics. At the end of the day these innovations help make businesses more efficient, which certainly contributes to our society’s productivity, but many of these challenges that MIT prides itself on trying to solve are ones that businesses alone cannot fix.

Analytics companies and others that hire data scientists and programmers often flaunt the “challenging problems” that their employees solve. I am not arguing against someone seeking such a job, because such things can be genuinely intellectually fulfilling, and sometimes the only people who can tackle such a problem are those who have been through MIT. But as an institution that seeks to better humanity through the sciences, our mission should not be to solve hard problems; it should be to solve pressing problems. Business analytics has never been, and never will be, a field with the potential to improve the human condition or a field whose lack of development severely holds back humanity from achieving something great. I stand by the belief that just about any social problem that data science can help solve (say, car accidents, or the American healthcare system) already has accessible solutions that have already been proven effective (increased public transport, and some form of universal healthcare).

If we really want to tackle important problems that have huge global impact, why aren't we putting this large sum of money into our climate change research, our urban planning department, or our economics and political science departments? Why aren't we putting our money to improve the quality of, and increase the quantity of, the policymakers, economists, and biologists that we produce? Or improve the quality and quantity of all of the research that we here at MIT produce? Why are we “selling out” to jump on the AI train, when there are things that have so much more impact on the world that we could be doing with this funding?

Or why aren’t we shifting our computing focus onto applications with a more obvious social need? Why does eecs-jobs-announce email about generic startups in need of programmers? Why are tech companies obsessed with the “solutions” and not the “problems”? How many of your peers are involved with Code for Good? Do you, the reader, actually know what MIT GOV/LAB even is, or even of its existence? (I certainly didn’t until I wrote this piece.)

I'm not really disgusted so much as I am disappointed that MIT, whose mission is to do scientific research that benefits all of humanity, is following the data science and AI trend that the rest of the corporate world is obsessed with. At the end of the day, society wants data scientists and analytics, but we need climate scientists, biologists, engineers, urban planners, public policy wonks, and so many other types of people so much more. We should spend our money to produce more of these types.

— Faraz Masroor ’21