The Machine Can Write the Essay. Now What?

There is a question that most schools have not yet answered honestly, and most parents have not yet thought to ask. If a machine can now produce a competent essay on any topic in under thirty seconds, what exactly is the point of spending thirteen years teaching a child to write competent essays?

This is not a rhetorical provocation. It is a genuine design question, and the answer to it should be reshaping how we think about education right now. It largely isn't. Most schools are still assessing children on the same outputs that artificial intelligence can now generate on demand, at negligible cost, with no particular effort. The credential that results from doing this well is becoming a signal of something considerably less valuable than it was even five years ago.

To be precise about what has changed, because precision matters here and a great deal of the conversation around AI is either catastrophising or dismissing, neither of which is useful.

Artificial intelligence systems can now perform, with increasing competence, the cognitive tasks that formal education spends most of its time developing. Retrieving information. Producing structured written analysis. Generating fluent output on demand. Executing well-defined problem-solving tasks within known parameters. Summarising complex material. These are not peripheral capabilities. They are the capabilities that school measures, rewards, and uses to sort children into their futures. A machine can now replicate most of them at a cost that approaches zero.

What AI cannot do, at least not yet and not with any reliability, is think from first principles about a problem it has never encountered before. It cannot bring genuine lived experience and real judgement to a situation that is uncertain, ethically complex, and consequential. It cannot build a relationship of real trust with another human being. It cannot exercise the kind of creative synthesis that comes from a person who has built deep knowledge across multiple domains and can see connections that nobody has previously mapped.

These are not consolation prizes. They are the capabilities that become more valuable every year as machines take over more of the cognitive middle ground. The difficulty is that they are also precisely the capabilities that standard education is least designed to build.

A parent paying attention to this shift is not asking whether their child will be replaced by a machine. That is the wrong question, and it tends to produce either paralysis or false reassurance. The right question is considerably more specific. Is my child developing the capabilities that become more valuable as AI improves, or the capabilities that become less valuable? Because those are two very different development paths, and most educational systems are currently pointing firmly in the wrong direction.

The essay is not the point. It never really was. The question is what kind of thinking produced it, and whether that thinking is being built deliberately or left to chance.