It is our pleasure to introduce fellow Evan DiPetrillo, who has joined the office of Senator Todd Young. While there, Evan will focus on matters pertaining to artificial intelligence and national security.
Read on to hear Evan’s fascinating take on the relationship between history and technology and how it inspired him to join the Hill as a technologist:
Like many of those interested in technology, I grew up reading science fiction. Arthur C. Clarke was my personal favorite, and my high school writing projects often centered on the different paths humanity could take once we developed artificial intelligence advanced enough to free us from the need to work. For just as long, I have also been enthralled by history. From classical antiquity and the ancient Near East, to the Napoleonic era and what Churchill dubbed the “terrible Twentieth Century,” history provides a reservoir of lives and lessons to draw upon.
History, however, is only obvious in hindsight. It is easy for us to pass judgment on the wisdom or folly of the Munich Agreement, the Sicilian Expedition, or the Congress of Vienna, but what about the key questions of our own age? Where will they be answered? And most importantly, what can I do to make it so they will be answered well?
After college, I joined the U.S. Navy and served for several years as a submarine officer, during which I completed two undersea deployments in the Pacific. I saw firsthand the tremendous potential of advanced technology (specifically nuclear energy) and the equally tremendous consideration as to how to wield it responsibly. I also saw the meticulous care, day-in and day-out, required by operators to maintain control over such powerful forces.
While in graduate school for economics, I read the book Superintelligence by Oxford academic Nick Bostrom. In it, Bostrom details the inherent difficulty of controlling an entity with greater problem-solving ability than humanity. He makes no claims about when this may be developed, but stresses that the safety challenges we face make it worthwhile to begin thinking about it sooner rather than later. It was clear even then that a similar dynamic was emerging in artificial intelligence, much like what I had witnessed in nuclear energy. There are tremendous possible upsides from AI, but in order to benefit from them, AI’s risks must first be prevented or mitigated to the best of our abilities.
I decided to redirect my career then and there, and enrolled in NYU’s Computer Science M.S. program shortly after. While I had dabbled in programming since at least middle school, where I wrote text-based computer games in BASIC for my friends, this would be my first time studying it formally and learning the underlying theory. Fortunately, my prior studies in economics provided a surprisingly good preparation for machine learning, as both fields are grounded in the principles of mathematical optimization. Since graduation, I’ve been working at the intersection of national security and data science, with my most recent position being with the U.S. European Command in Stuttgart, Germany. That work was deeply fulfilling, but I knew that if I wanted to have the most impact, I would need to work on the most significant problems on a larger scale.
One hundred years from now, one of the most important questions about this era will be if we got artificial intelligence right. This is fundamentally a question of policy, and therefore can only be answered by our elected representatives. As it stands, Congress is perennially understaffed and under-resourced, even in traditional policy areas, and as a result, a novel domain like the governance of artificial intelligence truly stretches the capacity of the institution. Thankfully, TechCongress addresses this by connecting Congressional offices with experienced technologists eager to work on policy. I am excited and grateful for the opportunity to be a part of this initiative and look forward to working on artificial intelligence and national security, and giving back to a country that has given me so much.