Three Updates on Data and Math Careers for College Students
Last week, I spoke at Carleton College’s Math/Stats Colloquium Series.
I wanted to get Carleton students excited by the possibility of a career in Operations Research and AI. So I used plenty of interactive examples ranging from supply chain problems to kidney transplants to automating quality control for potato chips.
This gave me an excuse to add some material to last year’s article: “Two Careers Every Smart College Student Should Know About.” (There are many good links in that article).
Here is some additional material for students who may be interested.
First, many different groups are releasing short videos explaining what this field is all about. I’ve only sampled a few, but I would like to keep adding to the library.
The first one is from the University of Melbourne’s OPTIMA. It does a good job of explaining optimization, explaining that everyone has these problems and that you often need to balance multiple objectives.
The second one is from Northwestern’s IEMS Department1. Industrial Engineering is a name left over from the Industrial Revolution. This video explains that NU’s IEMS department is “data science for decision making.” The video gives good examples, including General Motors designing transportation for the city of the future or determining the proper nutrition requirements for preterm babies.
Second, I think the students benefited from hearing our Ocean’s Eleven2 analogy.
There is still some thought that data scientists should have many different skills, including math, stats, machine learning, optimization, python or R, data engineering, etc. This is the view that data scientists should be like Decathletes— great at many things.
Students need to hear that, in practice, the movie Ocean’s Eleven is a better analogy. Data science teams are made up of many different people with different skills. Students shouldn’t be scared away. There are many ways to be a valuable part of a data science team.
Third, when discussing careers in AI, I think it is important to distinguish between the three types of AI— Artificial General Intelligence (AGI), Generative AI (think ChatGPT), and Practical AI— and what is required.
If students are interested in AGI (and related hard problems like self-driving cars), a PhD-type background is important to break into these fields.
For Generative AI, I told them that no one knows how this technology will shake out. It will be an interesting few years, but you can get into this area with some basic technical skills.
For Practical AI, you can get into this with Machine Learning. However, combining that with some Operations Research (especially Optimization) will make you better off.
Full Disclosure: I teach for this department.
We used this analogy quite a lot at Opex Analytics.