Math learning for adults

I know plenty of people who are scared of math. And their story is always a variant of how they struggled with some topic in school and then assumed that they’re not cut out for math. And they could never catch up.

The school system is partly to blame - the one-size-fits-all approach. The underpaid teachers - I don’t really want to blame them, but I’ve seen a few bad teachers who rubbed students the wrong way. The parents - who want the best for their child but can only outsource education to a school or tutors. The kids themselves - who can be disobedient brats who make life hell for teachers.

I did some surface-level research over the past few days and found plenty of EdTech initiatives aimed at fixing education for the next generation. I’ve been reading The Math Academy Way which is a treasure trove of knowledge, compiling learnings and research from the last ~50 years, showcasing what pedagogies have been shown to work best and how modern technology can be used to make learning effective for every student.

One of the most widely quoted findings in the field of education that I often come across is “Bloom’s 2 sigma”:

In 1984, Benjamin Bloom described experiments in which students receiving mastery-based one-on-one tutoring performed around two standard deviations above conventional classroom students. The result became famous as the “2 Sigma Problem”: how do you reproduce the benefits of individualised tutoring at scale?

In simpler terms, imagine two batches of 100 students. Batch A is taught using a traditional classroom setup. Batch B gets one-on-one tutoring, with teaching adapted to each student.

At the end of the term, all 200 students take the same exam. The average student from Batch B would score about as well as the 2nd or 3rd best student from Batch A.

That’s the scale of the difference implied by two sigma.

This bodes well for those who can afford personal tutoring. But not everyone can. Not until now. I’m writing this in 2026, when AI is showing promise of being a capable one-on-one tutor.

The hard part about making learning effective seems to be finding a way for mechanics and motivation to compound with each other. AI is getting very good at mechanics - creating course content, adapting it to the student, applying different learning strategies, and measuring mastery. Motivation is about making learning fun, getting students excited, showing them their growth, getting them in the zone and maybe even competing with time-sinks like TikTok brainrot for their attention. I have a hunch that the world of gaming might hold the playbook for motivation.

Coming back to what I started with - math for adults. I’m considering whether it’d be worthwhile to build a product for adults to learn math. Something that is fun and educational at the same time.

I think it solves a real problem. Math appears in so many aspects of our life that we don’t realise. We’d all make better decisions if we were better at math. Financial decisions, big and small, require a good understanding of mathematical concepts.

Has my investment beaten inflation?
How much mortgage can I afford?
Is £15/month or £150/year cheaper?
Should I take a £70k job with more equity or an £80k job with less?
£4.50 for 750g vs £6 for 1.2kg

But as I was writing this, and thinking about examples for the previous list, I realised that a lot of these are now just questions I’d outsource to ChatGPT. And I’d trust what it’d say.

Maybe it’s not a bad thing - More people might take better decisions. Maybe it is - and we’ll lose our mental intuition.

I just went ahead and asked ChatGPT to give me some more examples. Turns out I was not thinking broadly enough. There are a lot of scenarios, mostly money related, right in front of our eyes, where I probably won’t pause to ask an AI, and would benefit from being better at math. Even if I did ask AI, I don’t think I’d be able to internalise the answer without working the math out myself.

But I don’t think people want to be ‘better at math’. That’s not a burning problem. Stop making bad financial decisions, yes. Save more money, yes! Make more money, absolutely. Grow in your career, for sure. All these seem more like burning problems.

I’m going to talk to more people and seek their views on this. In particular, I want to collect moments where people felt that being better at math would have helped them make a better decision. I’ll compile and publish what I find.

If a situation comes to mind where you thought, “I wish I was better at math,” I’d love to hear about it. Send me an email at akshay@akshaylal.com.