This article originally appeared听on Medium and has been reposted with permission听from the author. .听
By Vinjay Vale, 2013 CDB Scholar
Today鈥檚 education system has created a rift between STEM (science, technology, engineering and math) and the humanities. It may seem counterintuitive, but听studying the humanities actually makes you a better scientist. Scholars of literature, philosophy, art, and history develop an understanding and appreciation for what it means to be human. I believe it is vital for scientists to study these fields, especially in a modern world full of rapid technological advances.
Through my own research, I experienced firsthand how the humanities can help scientific research.
My Regeneron STS project was on Artificial Intelligence, specifically teaching computer programs to learn and interpret geometric structures in visual scenes. The current state-of-the-art has poor spatial and compositional understanding; for example, it classifies a leopard-print couch as a leopard despite the furniture鈥檚 lack of a head and tail. Understanding how objects are composed of their parts is critical for complex vision tasks like visual reasoning. My approach to the problem deviated from the modern paradigm of neural networks. Believe it or not, I was inspired by a 1987 psychology paper by American vision scientist Irving Biederman on the Recognition By Components theory for human vision. The theory is based on breaking down complex objects into simple spatial elements called geons.
As I worked on my project, I spent multiple weeks outside the lab where I exclusively read books and papers, on subjects ranging from art to artificial intelligence, some suggested by my mentor Kevin Ellis (who I met through the听). In philosophy class, meanwhile, I learned about Plato鈥檚 Theory of Forms, and realized that I was trying to recreate a similar understanding of the visual world in computers.
Making connections between these diverse fields helped me develop the necessary insights to make headway on my interdisciplinary problem: teaching a computer to learn and interpret visual scenes.
My work falls into an emerging category of AI research called听explainable AI鈥娾�斺�妕hat is, building artificial intelligence systems that can articulate their thought processes to humans. This is an important area of AI safety, which merges ethics and philosophy with the more technical side of computer science.
In general, the gray area where ethics and tech meet is ripe for exploration. The underlying digital buzz permeating all aspects of our lives makes human moral judgement all the more essential. Think about social media, big pharma, self-driving cars鈥娾�斺�妔cientists who understand ethics have the capacity to make a positive impact on the world.听听听听that engaging in arts, history, and literature bolsters morality, compassion, and empathy. As a pianist, composer, and avid reader, I鈥檇 certainly like to think I鈥檝e benefited in this way.
My study of the humanities also has made me a better collaborator, by helping to sharpen my communication skills and compassion. I know that in the future I want to be part of interdisciplinary teams of scientists that tackle significant real-world problems. There鈥檚 no doubt that the interpersonal and collaborative skills that I learn from studying and discussing history, philosophy, and literature will be essential in the future.
Whether we鈥檙e talking about clinical trials (where every day human behavior can make or break a drug) or computer science (as in my experience drawn from art and psychology), understanding human nature is fundamental to doing science. When most people think about the future of humanity, they envision a universe where science has propelled our species towards a better tomorrow. But science alone is not enough to solve the world鈥檚 problems: that science needs to be done by听people who understand what it means to be human.
Like this post?听听to receive more stories, information, and resources about gifted youth straight to your inbox.