Nathan Fulton

Nathan Fulton, Ph.D. Philosophy

Nathan Fulton Alumni Interview, August 2017

What does cutting edge research into artificial intelligence have in common with the study of philosophy? For department of Philosophy alum Nathan Fulton (2016) they share a lot. As humanities scholars we are taught that, in theory, the skills we develop in graduate school can be transportable to an ever-widening variety of careers. Nathan exemplifies this ethos of interdisciplinarity in his job at an artificial intelligence firm along with his overall perspective on knowledge, education and society.

As a programmer and researcher, Nathan straddles academic and corporate worlds, but this wasn’t always the plan. In thinking about post-graduate life, “I pretty much expected to be publishing heavily enough in a tight enough area of focus to be marketable to academic jobs…I had a pretty standard view” about teaching and writing philosophy. He discovered that “the academic job market isn’t great, but the philosophy job market is terrific,” if you are willing to seek opportunities outside of the academy, that is. In the interview below, Nathan describes his path, transitioning from philosophy graduate student to artificial intelligence researcher.

What has been your professional trajectory post PhD?

About a year and a half before I finished my dissertation, I applied for and began a job with an artificial intelligence research firm, Cycorp, in Austin Texas. I am an inference programmer for that firm.

How would you describe your present job?

I work on the inference engine for an artificial intelligence called “Cyc”. Cyc performs complex symbolic reasoning (a type of intelligence that interprets symbols within a rigorously specified context) using a knowledge base containing millions of pieces of information expressed in a very precise logical syntax. My job is largely about optimizing and extending this reasoning capability.

What are the skills that you feel are the most important for your job?

Logic is a lot of it: more or less the First-Order Predicate Calculus, along with some higher-order extensions. It’s important to be able to parse ordinary sentences in natural language into a very rigorous form. I do a lot of LISP programming, though that isn’t a requirement for all of our technical staff. A sense of humor is pretty crucial! Though that might be more a matter of temperament than skill.

What do you find most gratifying about your job?

It really is applied philosophy. It’s applied in the sense that we engineer a system that has to perform complicated tasks that will have on impact in the real world, and it’s philosophy in the sense that we seek to represent the world as rigorously and richly as possible, and enable Cyc to reason about it with extraordinary depth and range. It’s a very powerful feeling when you realize that very central, fairly abstract aspects of philosophy can have extensive practical impact across various industries and projects.

Thinking back on graduate school as preparation for this career, what skill or capacity do you think your education best developed that you rely upon at Cycorp?

Philosophy really does improve your ability to think about the structure of the world, to the point where you can describe things in ways uniquely suited to A.I. work. I did various kinds of modeling (for game theory and rational decision theory) that were exceptionally good preparation for this, but even without that sort of stuff, I would have been well prepared for the work I’m doing now by my academic program. Breaking claims down into very elementary pieces, noticing how putting a sentence into the right form changes what you can do with it, separating background assumptions from explicit premises... these skills are all important for developing automated reasoning, and are all taught particularly well in philosophy departments.

How does this relate to what you understand as your disciplinary training?

I don’t think there are a lot of limits on what philosophy can positively impact, so even if I wasn’t working in an environment where (analytic) philosophy was explicitly valued, I would still probably find my training helpful. As it happens, it would be almost literally impossible to do my job well without it.

What skills do you feel were least developed that would have helped you either transition to this field or work more capably in this field?

I think it would have helped me, and would help a lot of people, to understand a little better how different kinds of research tend to get funded. Working at a university, or a private firm that is partially funded by government grants, or an entirely private research concern are all different kinds of experiences that offer different opportunities.

Were there outside (non-graduate school) experiences that you feel prepared you for this job? What were they?

There’s a continuum from “part of my academic program requirements” to “nothing whatsoever to do with grad school”, and valuable things for my current career can be found anywhere along it. Talking about philosophy to people outside my program, or even outside academia, was very important for what I do now, because there’s a kind of translation involved at almost every step: language to logic, logic to code, code to application behavior, behavior to client experience.

Have you maintained a connection to your previous scholarly work some way?

I’m lucky; my company has a semi-formal working group of philosophers who meet to talk about their own ongoing work, separate from company projects. I hope to be work-shopping papers with them later this year, and I hope to publish fairly continuously in the next few years. Admittedly, though, it’s harder to protect your time and energy for that sort of thing when you have a commercial research project you’re heavily committed to.

If you were tracking yourself—and others like you—what is one question that you’d want to ask of your colleagues that would result in knowledge useful to you and to the university (current and future graduate students, faculty, staff, department and university administration, etc.)?

Do you feel that working in your current, non-academic capacity still leaves options open for returning to a university career at some future point, or was it a decisive move away from the academy?

So how would you answer this question for yourself?
[My job] is an artifact of privilege. My options are going to be more expansive than others. I could teach for not just a philosophy department, but also a computer science department. I am very confident about any prospects in a variety of academic departments in the future.

I feel that given how competitive the academic job market is, it is a responsible move to take non-academic jobs and leave the field for others if you can.

So how did you get your first non-academic job post Ph.D.?

A friend of mine (from UCI) started working at Cycorp, and almost immediately sent me a bunch of messages about how I’d fit right in. I wasn’t really sure about the prospect at first---I’d gotten very used to thinking about myself purely as an academic---but he was right, and it didn’t take that long for him to win me over and persuade me to apply. The interview went really well, and I got an offer right there on the phone. I still had teaching obligations for a few more months, but the company was very accommodating regarding my start date, etc.

What advice would you give to other grad job seekers?

First of all, do projects and stay busy. Write, volunteer, organize, code, whatever keep you busy and collaborating actively with others. Second, be open to new descriptions of your own skill set. You may have become proficient with tools for one purpose that have applications to something you haven’t thought about, especially in conjunction with your other skills and talents. Very few people’s careers follow exactly the path they had in mind when they entered grad school, but that doesn’t mean they aren’t using their education, or that they are using it in some less apt way