Things often look least likely to change just before they fall apart. So it is with mathematics education. At first glance, the status quo seems indestructible. How much maths you learn is one of the best predictors of your future income. Knowledge of the subject is deemed so important that many countries (among them the US) offer special visa concessions to foreigners with a grounding in the subject.

Beneath the surface, however, there are serious cracks. Most students dislike maths, do not understand its connection to their later lives and do badly at tests even as the questions grow easier. Universities and employers complain that they cannot find people with the right skills. Maths, as it is taught today, will not be a mainstream school subject in 20 years’ time. Either it will succumb to a hostile takeover or it will be run out of business.

A century ago, this might have made sense. Accounting, some physics or some easy-to-model engineering were about the limit of what anyone could hope to accomplish with mathematics. Today, our horizons are wider. Technology means we can develop far more complex models of the real world and gain actionable insights, even where the underlying situations are sophisticated and fuzzy.

Just as in past agricultural and industrial revolutions, technology has changed the art of the possible and the learning needed to master it. Calculation is rarely the obstacle. It is no longer, therefore, the area where most students need to focus their energy.

At its core, maths is a problem-solving process. You specify a real-world problem, develop an abstract representation of it, calculate an answer for the abstraction and then translate back into the real-world language you started with. Before computers, almost all human energy was focused on the third stage: calculating. Now it is usually focused on the other steps instead.

But not in maths education. There, we insist everyone learns to calculate by hand, at least before they use a computer. And we insist they calculate simple, unrealistic problems that are nothing like those they will face outside. We make human imitations of calculating machines. Yet businesses need problem-solvers who are skilled in using modern tools.

Coding should be the foundation of a new computer-based maths — perhaps part of the same school subject

Computers are being used to drive pedagogical change in all subjects. In maths, however, it is the subject matter itself that needs to change. It would be pointless to use computers to help students learn redundant skills of hand calculation. Teaching the wrong subject, however well or with whatever technology, will not make it right. We need to use computers in maths education as they are used in real-life maths — for doing the calculating.

Britain has reintroduced computer coding into its school curriculum. This is not just a workplace skill but a way to represent mathematical ideas for computation. Coding should be the foundation of a new computer-based maths — perhaps part of the same school subject.

Those who are first to embrace this way of teaching maths will reap the biggest rewards, just as early adopters of universal education in the 19th century did.

In some countries, such as Estonia, enlightened governments have been bold enough to look beyond improving their results on tests such as Pisa, which measure international school systems against an old yardstick of success. In others, industry and universities must provide the impetus.

In 1988 Steve Jobs saw the need for a change in maths and urged educators to focus on “the prose of mathematics without getting lost in the grammar”. To make that vision a reality we need to beat the intransigence of our educational systems whatever the technology.

The writer is the chief executive of the Wolfram Group Europe and founder of computerbasedmath.org