The case for mathematics in the AI age
(...and the tech bros agree)
By rights, mathematics should have died out with the invention of the calculator.
If it were only about arithmetic.
As teachers and school leaders brace for the AI-inflicted disruption to the school curriculum, I am confident that mathematics will continue to survive, and perhaps even continue to thrive.
One of the major misconceptions around the subject I have taught for the past twenty years, is that mathematical ability is about ‘being good with numbers’. However, ask any professor what makes a good mathematician, and you won’t hear them talk about numerical ability at all: they will use words like creativity and inspiration; they will talk about making connections, resilience, problem-solving and the ability to think laterally and abstractly to find elegant solutions. Above all, mathematicians love the word elegance.
The distinction between effective, efficient and elegant solutions is at the heart of mathematics, and the ability to make this distinction continues to matter.
To use some sporting analogies…
Nadal: effective, Djokovic: efficient … Federer: elegant.
Haaland: effective, Ronaldo: efficient … Messi: elegant.
Effective gets the job done; Efficient gets the job done with an economy of effort; Elegant gets the job done with ease and with a certain style.
This idea of elegance over efficiency is very familiar to mathematicians. There are effective ways to solve problems - often referred to as ‘brute force’; these solutions can be refined into efficient methods - think of the algorithmic techniques you might use for long division / multiplication; and then there are elegant solutions - those solutions that when a student provides them, you almost want to stand back and applaud for their ingenuity, style and ease.
For example, a question I love to ask students is ‘roughly how many basketballs will fit in this classroom?’ Pause for a moment to think how you might solve this.
The vast majority of students work out the volume of the classroom in metres cubed, then divide by the volume of a spherical basketball, often having to correct for units and simplify the formula for the volume of a sphere - they will often arrive at a reasonable, effective solution, based on formulae and arithmetic. But what about the space around the spheres? - ah, perhaps divide the classroom’s volume by the number of cubes each spherical basketball would sit within… getting closer to an efficient method.
On the other hand, there are the small percentage of elegant thinkers, who imagine the dimensions of the room in units of ‘basketballs’ - ‘you just imagine 15 along the width, 15 along the length, and 10 to the ceiling, so around 2000: easy’.
(There is clearly another effective, brute-force approach, which is to fill the room with basketballs - perhaps the first thought of a physicist!)
The same idea maps on to product development in business, and in particular, the AI race.
Listeners of Bill Gates’ podcast, ‘Unconfuse Me’ will be familiar with his interest in mathematics education, and more recently, his pursuit of ‘AI for Maths Education’. Gates’ interview with Sal Khan (of KhanAcademy.org) is just one of several episodes highlighting Gates’ vision for AI disruption in the maths education sector. I have written about the Gates-Khan partnership in this earlier post. Moreover, it is clear that Gates sees mathematics as a particularly fruitful and important area for children to continue to develop in the AI age.
In a separate conversation, with Yejin Choi, ( professor of Computer Science at the University of Washington), we hear Gates explain why a grounding in Maths may well surpass the need for a grounding in programming in the brave new world of AI. They have a fascinating conversation around how the next phase in the AI-race will be likely won by those who focus on more elegant solutions. Large Language Models (LLMs) have a feel of ‘brute force’ to the mathematically-minded: surely there is a better way than churning through all possibilities for the ‘next word’ in the sentence?
Now that the AI can do the programming, it is this more lateral pursuit of elegance, and the ability to ask the right questions that will hold value for employers. As the latest The Future of Jobs Report 2025 | World Economic Forum finds,
Analytical thinking remains the most sought- after core skill among employers, with seven out of 10 companies considering it as essential in 2025. This is followed by resilience, flexibility and agility, along with leadership and social influence.
These are all skills that will be learned in the STEM classroom, as well as the arts and humanities. So, perhaps those of us who are teachers won’t be out of a job just yet.
Let’s finish on a final problem - answers to my direct messages, with copyright disclaimers please ;-)
ChatGPT = effective, DeepSeek = efficient, ??? = elegant
The question marks will be filled by those who can think mathematically.