What Criticism Is For Now

The question is asked with some regularity, usually with an implied answer: what is the point of criticism when anyone can post a reaction and every platform recommends automatically? It is a fair question, and it deserves a better response than the defensive one usually offered.

The recommendation function is genuinely gone

For most of the twentieth century criticism served a practical purpose. Seeing a film cost money and an evening, information was scarce, and a trusted critic saved you from wasting both.

That function has been comprehensively automated. Recommendation engines know your history and predict your preferences more accurately than any individual could, and the marginal cost of abandoning something after ten minutes is nearly zero.

Defending criticism as a consumer guide is defending a service that has been superseded. The interesting response is that this was always the least interesting thing criticism did.

What does not automate

Three functions remain, and none is within reach of a recommendation system.

The first is placement. A work exists inside a tradition, and understanding what it borrows, refuses or reverses changes what it means. An algorithm can tell you that people who watched this also watched that. It cannot tell you that a film is arguing with something made forty years ago.

The second is argument. A serious review makes a claim about what a work is doing and defends it with evidence. Its value does not depend on agreement — a well-made case you reject sharpens your own position, which a score cannot do.

The third is description. Articulating why a performance works, or how an edit produces a feeling, gives readers language for their own responses. Most viewers know when something moved them and not why, and criticism supplies the why.

The aggregate score problem

The most damaging development has been the reduction of criticism to a percentage.

Aggregation converts varied arguments into a single number by discarding everything except a binary judgement. A review saying a film is flawed but essential and one saying it is competent and forgettable both register as positive, though they say opposite things.

The number then becomes a marketing input, which creates pressure on the practice itself. Studios cite scores, audiences treat them as verdicts, and the incentive shifts towards being counted rather than being read.

The score replaced the recommendation function efficiently. The problem is that it is also perceived to have replaced the other functions, which it cannot perform at all.

Where criticism is actually healthy

It is worth noting that the practice is not obviously dying — it has relocated. Long-form video essays, newsletters and podcasts sustain serious critical work with real audiences, frequently at greater length than print ever allowed.

What has declined is the staff position at a general-interest publication, which is an employment story rather than an intellectual one. The work continues; it is simply less institutionally supported and more precarious, which affects who can afford to do it.

That last point matters. When criticism becomes something sustainable only by those with other income, the range of people writing it narrows, and a narrower range produces a narrower account.

The honest case

Criticism is not necessary. Nobody requires it to enjoy anything, and a great many people have rich relationships with film and television without reading a word of it.

But the experience of understanding why something affected you is different in kind from having been affected, and criticism at its best is the record of people working that out in public and inviting you to disagree.

That is not a service. It is a conversation, and conversations do not become obsolete because a machine can tell you what you might like next.