We have written several thousand words on the essay's problems, so it is worth being clear about a distinction that keeps collapsing in this debate.

The claim is that an essay no longer reliably certifies that the person who submitted it did the thinking. That is a claim about the essay as an instrument of measurement. It is not a claim about the essay as an activity, and the two have entirely different implications.

What writing does that nothing else does

Writing forces sequence. You cannot hold a muddled argument on a page the way you can hold it in your head, and the discovery that your third paragraph contradicts your first is a form of thinking that happens nowhere else. Anyone who has written seriously knows the experience of finding out what they believed by trying to write it down.

None of that is affected by whether a model could have produced a similar-looking document. The value was in the doing.

Which means the response is not to stop assigning it

The instinct in some departments has been to retreat from written work toward formats that are easier to secure. That trades away a genuinely irreplaceable form of learning to solve a measurement problem, which is a bad trade.

The better move is to separate the two jobs the essay was doing.

• Keep assigning writing, and keep teaching it, because the activity is the point.
• Stop asking the finished document to also certify authorship and understanding, because it can no longer carry that.
• Put the certification somewhere else, in a short conversation about the work, which takes minutes rather than the hours a careful read requires.

That split has a pleasant side effect. Once the document is not carrying the evidential burden, feedback on it can go back to being about the writing: structure, argument, evidence, the sentence that is doing too much work. Which is what most instructors wanted to be commenting on in the first place, before the last three years turned every marking session into an authorship investigation.