The Era of the Mathematician Has Arrived
I trained as a Pure Mathematician. An algebraist, in fact. By the time I finished my Ph.D., I was so deep into an esoteric specialism that they struggled to even find someone to examine me. I found myself in this rabbit hole because I thought I wanted to be an academic. But then I changed my mind.
Academia was too lonely and too poorly paid, I decided, and so it was time to move on. But where do you go when your only practical experience is spending days writing Greek and Hebrew symbols on black/whiteboards? I was resigned to the fact that it was time to cast my previous expertise aside and start again, that there was no use for deep algebra in my future career.
Fast forward a few years and I find myself proved utterly wrong by that assumption. Today, I would argue that there is no better field to start out in than mathematics if you want to make sense of everything new that is happening around us. And don’t just take this from me — the job market is realizing this too. In a recent report, the US Bureau of Labor Statistics projects that the job market for Mathematicians and Statisticians will grow by a whopping 33% in the next decade, and calculates that Mathematicians already earn almost three times the US average salary.
Why is mathematics suddenly so important?
Even though the market hasn’t always reflected this, I would argue that it has always been important in most fields to have mathematical training. Highly trained mathematicians learn a discipline to their thought processes that inject calmness and assuredness to the people they work with. In today’s fast-paced environments, where quantitative and qualitative facts are coming at us from all angles, there is still a great need for systematic thinkers who have a logical approach and can quickly reduce problems to their crux. Even before the age of big data, I was finding that these skills that I had been so steeped in as an algebraist were really helping myself and my teams in many difficult, tricky, and complex real-life situations.
But despite this, I never imagined even five years ago that the actual content of what I learned would come back to be such a massive part of what I do. How would multidimensional linear algebra, discrete mathematics, matrix theory, combinatorics, logic, and probability be of any use to me in a career in the world of business?
The explosion of big data, data science and AI changed all that. Now, amazingly, I re-engage with all these topics on a day-to-day basis. Here are some examples of what I mean:
My linear algebra background helps me deeply understand how feature spaces in Machine Learning operate, how learning algorithms like Support Vector Machines work, as well as many of the modern embedding techniques in natural language processing. This means I can interpret results more effectively but also explain these methods to others in a relatable way.
My training in discrete mathematics helps me with topics like graphs and networks as well as cryptography, all of which I have to think about or work with regularly.
Underlying classes of matrices and matrix properties like eigenvectors and eigenvalues are critical foundations to many computational techniques today, including algorithms, network analysis, and statistical models.
Combinatorics is an incredible training ground for programming in general, and functional programming in particular, all of which are much-needed skills in the workplace today.
Logic and probability are foundational for statistical hypothesis testing, something I regard as the biggest gap in decision-making skills in today’s data-rich world.
The era of the mathematician is here
The era of the mathematician has arrived. No longer will academic pursuits be the only realistic option for those who want to immerse themselves in the beautiful world that I started out in and have more recently returned to. We are already starting to see and will continue to see, an impressive growth of enterprise roles as mathematics practitioners, many of which are directly impacting outcomes in the world around us.
If you are a young person reading this, and still considering what you want to train in or study, you probably know what I would tell you. If math is in your wheelhouse, stick with it. It’s going to be the skill of the future.
But even if you are a bit further on in your career but have always been mathematically inclined, it’s never too late to become a mathematician. The resources are plenty and the options are varied. I’d highly recommend it if you have the time and interest.
Do you feel you have sufficient training in mathematics to properly understand some of the latest developments in data science, AI and analytics? Feel free to comment.
Data Leader
1mo"there is still a great need for systematic thinkers who have a logical approach and can quickly reduce problems to their crux." I resonate with this phrase. After graduating with a undergratuate degree in Mathematics, I had no idea what I could do with my degree. Being in the data space has showed me the value of the logical thinking that my degree taught me.
AI in the Manufacturing Industry & Automotive | Maintenance | People | Optimizations | Innovations | Education | Virtual reality | Data | Industry 5.0
3mogreat post
Industrial scientist and developer focusing on robust AI systems and evaluation frameworks.
3moAgreed. In 2008 I started a mathematical journey related to attempts to understand how one can build self-evaluation algorithms to grade tests for which we have no answer key - a fundamental problem in AI safety (unsupervised evaluation). The solution required that I learn algebraic geometry, a topic I had never encountered during my doctoral training in Physics. The most mundane task - grading a multiple choice exam without an answer key, just the responses of the test takers to said exam, requires mathematical tools that were not available until the 1960s (Buchberger's algorithm, Groebner bases, etc.). That is new in mathematical history. Who knows what great ideas remain to be mined? As in my case, if you think about a really fundamental problem in your field, you will must surely need mathematics to sort your thinking out on it. Those interested in how algebraic geometry helps build tools for unsupervised evaluation of noisy AI agents, check out the documentation page for my "ntqr" Python package - https://ntqr.readthedocs.io/en/latest
CTO at Geopixels - GeoAI Data Scientist Practitioner
3moGloria Buriticá
MBA, ASQ CSSBB
3moGreat article, thanks for writing and sharing it.