You will see it splashed all over this website as well as anywhere else on the internet that talks about careers in artificial intelligence: it’s a field that absolutely demands a strong command of advanced topics in mathematics and statistics.
If you are among the nearly 25 percent of American college students who suffer from moderate or severe levels of math anxiety, that’s bad news. And if you are interested in careers in artificial intelligence or machine learning, it probably leaves you with a big question: can you get a job in AI if you are bad at math?
The answer may surprise you.
An Epidemic of Innumeracy Threatens American AI Progress
If you believe you have poor math skills, you are not alone. According to Associate Press reporting in 2023, American students score lower than students from 36 other countries worldwide. Only one out of every five high school students are ready for college math courses when they graduate.
This is a big challenge for the economy in general, since STEM positions in general are expected to grow by some 30,000 openings per year for the next decade. AI jobs will be a big chunk of that number.
But it’s a challenge with solutions, including some that will work for you.
It’s Not That You’re Bad at Math, It’s the Way it was Taught
First, let’s get rid of a problematic misconception: people aren’t just bad at math. Math skills are a function of perseverance and the right education. No one can give you motivation, but you can find places that will teach you math skills to the standards required in the artificial intelligence industry.
In fact, considering the scope of the problem, many American colleges are taking steps to help students get up to speed. Northeastern University, for example, offers a Bridge to Calculus program for high school students that serves as a sort of boot camp that pushes kids from underserved communities in Boston to develop skills sufficient to take Advanced Placement Calculus and earn college math credits before they are even out of high school.
Systemic failings in American math education have robbed many students of the proper instruction they need to tap their innate inner mathematician.
It’s important for many people to understand that it’s not a personal problem. The education system is falling down when it comes to preparing American students for STEM jobs. But with recognition of the problem growing more widespread, so are available solutions.
Why You Haven’t Been Taught What You Need to Succeed in AI Engineering
Most American schools teach math on a conveyor belt. Pressure to teach to standardized test combined with the inability of those tests to adequately measure genuine understanding means most kids don’t get the solid foundations they need. Pure rote memorization gets you through a test, but doesn’t convey real understanding.
When you end up confronting genuine problems, anything you haven’t been coached to work through, then, you know you are unprepared. Psychologically, this builds up. Most American students simply come to believe they don’t have the skill to perform advanced mathematics. This belief keeps them from ever seriously trying.
This leaves a lot of high school graduates way behind when it comes to the math foundations needed to launch on a Bachelor of Science in Artificial Intelligence, let alone moving up to a master’s or doctoral degree. But the reality is you can handle any equation that comes your way, with the right training.
Solutions to Get Your Math Skills up to Speed for AI Engineering
Your first step is to get the idea that you just can’t do math right out of your head. After that, you will find that there are plenty of programs that will help build up your math skills enough to get the foundations needed for careers in AI.
Free online courses like those offered through Kahn Academy will help you work back to your point of pain in the fundamentals and drill them in solidly so you will absorb them this time around. YouTube videos offer up lectures to help you come at the same concepts from different angles, until you find what works.
Many colleges, and particularly community colleges, have also taken note of the problem and set up special classes to help math-challenged students get up to speed. University tutoring centers can pair you up one-on-one with someone to help walk you through the hard parts.
Math Gets Easier When You Are Using It to Achieve Something Real
Just as important is the shift from thinking of math as a purely theoretical exercise to something that is of ultimate practical importance in your field.
Students in middle and high school can be forgiven for finding math dry and unenjoyable. They are typically introduced to concepts in a vacuum detached from their daily desires and expertise.
By the time you get to a point in your education where you are seriously considering a career in AI, you can come back to those concepts with more specific goals in mind. Random studies in linear algebra may be hard to find the motivation for. But if you can take those tools and suddenly apply them in machine learning algorithms for invariant subspace reductions to tear into image data, you might be a little more motivated to absorb them.
Math Is Important in AI Engineering, but It Doesn’t Have to Be Your Biggest Strength
Although you must learn high-level maths to succeed in AI, it’s not the whole story. You may absorb the bare minimum and never come to love the cold equations. That’s okay, too.
While the heart of machine learning is and will always be complex math equations, AI as a whole is a strongly interdisciplinary field. And just like everyone on a baseball team doesn’t need to be a strong hitter, there are roles in AI development where you can lean on strengths other than math.
Coding, for instance, is about logic and creativity. It’s possible to become a programming all-star based on your ability to break down problems and state things simplistically and elegantly in code. Even if you’re implementing algorithms rooted in math, you may find that your coding skills set you above even strong mathematicians in AI.
Something similar can be true about data science. Understanding relationships, visualization, and the various tools for managing and manipulating large data sets is key in modern generative AI training. Someone who has mastered these mechanics is going to be a valuable part of an AI dev team even if their math isn’t first-rate.
Finally, communication skills are a long-time must-have in any kind of collaborative development process. You may not be the math whiz in your AI engineering team, but if you have a knack for helping the math geniuses talk to the coding gurus and helping both of them make sense to the business staff, you will find there is a spot for you in AI development no matter what.
While AI can perform miracles, it hasn’t yet been able to negotiate, convince, or collaborate with humans on a large scale. If you can do that, you’ll never be replaced in AI development… no matter what your math skills are like.
Still, don’t count yourself out of the AI wave on account of math deficiencies just yet. After all, large language models are famously terrible at math, and no one is throwing them in the garbage. And like machines, humans have a capacity to learn and improve.
Do that, and you’re sure to succeed no matter how hard the equations get.