Local news will come to rely on AI

Because I am prone to occasional bouts of procrastination, I dawdled quite a while on Nieman Lab’s request for a prediction. And then it hit me: Why should I do all the thinking, writing and self-editing, when the new ChatGPT gadget is just sitting there, eager to take a stab?

In that spirit of sloth, I asked my new overlords, “What is the future of journalism?” Within seconds, I felt transported to every lanyard-dangling, crappy-coffee-serving panel I’ve served on or sat through since 2002:

“Journalism is going through a period of transition, and there is still a lot of value in the work that journalists do. But the industry is going to have to change in order to survive…

“One of the biggest challenges facing journalism is the decline in advertising revenue. One potential solution is to charge for online content…Charging for content can be a risky proposition, as it can alienate readers and drive them to other sources of information. It can also be difficult to get people to pay for something that they have become accustomed to getting for free…

“The work that journalists do is important, and it plays a crucial role in our democracy.”

Not bad, right?

A lot of people — especially teachers, or professors like me — are worried about this tool. “The College Essay Is Dead,” warns a recent Atlantic piece. And yes, it does provide a way for procrastinating, slothful students to submit plausible essays. Maybe they won’t get A’s, but most of the pieces I’ve seen would likely be good for at least a B.

But I predict that editors at struggling metro dailies, or thinly staffed nonprofits that are charged with covering government bodies, will someday look at this as a boon. Local newspapers and sites are getting thin these days. It isn’t just that they’re not publishing as many Pulitzer-finalist series as they used to. They also aren’t covering as many school boards, legislative committees, real-estate sales, new-business openings, and the rest of the grist that used to fill the back pages of newspapers. Even obituaries are largely relegated to paid notices from relatives. And as this information dries up, citizens feel more estranged from the agencies that govern their lives and the officials who set their tax rates and hire their superintendents.

There’s good reason for this news deficit. After the budget-trimmers have left you reeling, you’re not going to have one of your remaining reporters mindlessly type in city-commission minutes when they could be out covering news. But if we can automate some of this commodity news, we can provide a lot more information — much of it useful, if not sexy — to people who need it.

There are pitfalls, of course. One is the concern that this will just serve as a convenient way to eliminate more staff. We’ve heard that before. “Automated Game Stories To Make Sports Writers Obsolete,” warned Business Insider about software that generates articles based on baseball box scores. That piece was published in 2010 — or about 12 years before The New York Times bought the Athletic for $550 million in cash.

The bigger pitfall is the garbage-in, garbage-out problem. You can’t simply tell an AI program, “What did the town council do today?” or “Who got arrested last week?” You have to supply it with some raw information. The box-score equivalent would be minutes from a meeting, or incident reports from a police blotter. Would that be perfect? No. Would you want to publish it unedited? No. Could it save your staff a lot of time and generate a lot of goodwill for your readers? Yes.

Editors will do this because they have to. And I think they should — because, and here I quote the experts at ChatGPT, “I believe that journalism is still a valuable and necessary part of our society.” And in conclusion, fellow journalists, I just want to reiterate that “I think the industry needs to adapt to the changes that are taking place.”