The invitation was simple.

Send us your questions.

I had been doing something else.

I was sending observations, connections, corrections, sometimes an argument that had developed far enough to have edges. Things I had noticed and thought might be useful.

Answers, more or less.

That left a question behind.

How could a news organization handle more than questions?

For most of television history, the answer may have been practical rather than editorial.

It couldn’t.

An audience can be enormous. Even a tiny fraction responding to a broadcast can produce more material than a newsroom has people available to read. Asking for questions narrows the aperture. A question is recognizable. It can be selected, routed to someone who knows the subject, perhaps answered on air.

“Send us what you’re noticing” is a very different invitation.

What comes back might be anything.

A correction.

A memory.

An observation from someone who works in the industry being discussed.

A connection between two stories.

A disagreement with the premise of the segment.

A possible story.

A sentence.

Most of it may lead nowhere.

Some of it may be very good.

The difficulty is finding out which is which.

That difficulty is beginning to look different.

Researchers at MIT built a system called AudienceView for journalists dealing with more audience comments than they could reasonably read themselves. It uses a large language model to identify themes and topics across the submissions, show how those themes are distributed, and preserve links back to the individual comments from which they came.

The machine makes the volume navigable.

The journalist still decides what deserves attention.

A journalist can see that hundreds of people are talking about one aspect of a story, then go back and read what particular people actually said.

That is a small change in machinery with a potentially larger consequence.

The old invitation was written for the receiving capacity behind it.

Perhaps the next invitation can be wider.

Send us your questions.

Send us what you noticed.

There is an odd inversion here.

Much of the discussion about AI and journalism has concerned production: machines producing words, images, summaries, perhaps entire articles.

The receiving side may turn out to be just as interesting.

A large audience already contains experience, expertise, memory, local knowledge, disagreement and curiosity. Until now, much of that has been difficult to receive except through narrow doors: a question, a letter to the editor, a call to a producer, a comment beneath a story.

The material was there.

The listening capacity was small.

I still like the invitation I saw on television that morning.

Send us your questions.

I just wonder whether the technology has reached the point where the next line can finally be added.

What else have you got?

WE&P by: EZorrillaMc&Co.