AI’s coming inverted pyramid moment for journalism

Technology isn’t just a medium through which communication happens. It has an editorial impact on how journalists collect, filter and ultimately prioritize information as well as how the audience receives and perceives it. This is the great insight from Marshall McLuhan’s phrase, “The medium is the message.”

Because of the pace of change, we already recognize the growth of generative artificial intelligence as a pivotal moment for news. What we haven’t braced for, and perhaps can’t predict, is how the medium of generative AI will impact our message.

One of the most potent examples of technology playing the role of “prime-editor” — the first external influence to have an effect on copy — is the inverted pyramid, which has been a staple of journalism for seemingly forever. It’s so foundational that it feels engrained in the nature of news itself. But the inverted pyramid as we know it today is a direct result of the technological marvel that was the telegraph.

A 2003 Poynter article titled “Birth of the Inverted Pyramid: A Child of Technology, Commerce and History” explains:

But the telegraph had a drawback. It was expensive to use. One of the first charges was a penny a character. Newspapers spent hundreds of thousands of dollars in telegraph costs to report the Civil War. That economic pressure more than anything else influenced a new kind of writing that departed from the flowery language of the 19th century — it was concise, stripped of opinion and detail. Fueling the shift in writing style was a new type of news organization, named the “wire service” after the technology used to transmit the news.

This pattern, where technology acts as a prime-editor, has repeated itself through every technological revolution. Television brought us the anchorman. Search engines wrote countless boring headlines. And the wave of social sharing encouraged “These 10 headlines that you have to click to believe.”

When video came to social media, auto-playing silent videos with text overlays suddenly became ubiquitous. This shift, driven by platform algorithms rather than a deliberate editorial strategy, marked the “pivot to video” era. I’ve written much about this in the past under the name of “platform intelligence.” Study the platform (the medium) and it will inform your editorial thrust.

With all that history to set the scene, the question again becomes: What will be generative AI’s inevitable prime-editorial directive? Will it be as monumental and long-lasting as the inverted pyramid, or will it be a parallax scrolling fad? My hunch, shared by many, is the former, although the jury is technically still out.

The pace of change has been so fast that one can be forgiven for wanting to wait before declaring what the fallout of generative AI could be for news organizations.

In a provocative vision of the future, the influence of generative AI on editorial will flatten the value of producing formats. Today, a lot of time, energy and talent are spent on formatting information in just the right way. We craft stories written with grace, audio pieces meant to entice and tightly edited videos to engage. The cost of production in terms of formatting could drop dramatically.

Before we fear this future, I’m not suggesting the value-add of journalism disappears, but rather shifts from production toward something else. It’s that “something else” where the editorial influence of generative AI will become apparent.

For example, in a world where formatting and production are greatly eased, the emphasis for journalists is to find the “new.” What is the new information that can be formatted? When a local city council passes a bill, that’s new information. The journalist should share that information and other context, such as reactions from supporters or opponents.

It wasn’t too long ago when news organizations had local stringers who would call in information to the rewrite person. Some stringers probably couldn’t write their way out of a paper bag, but they were great journalists who knew how to focus on “the new.”

With emphasis put on gathering “the new,” the role of the journalist is not only to be “in the know” but to also cultivate relationships, personal and digital, to help them stay in the know and, even more importantly, gather unique/proprietary information that will increase the value of future formatting work. If all of this sounds like shoe-leather 101, it is. The journalist-as-writer could shift back to journalist-as-person-who-turns-over-rocks.

Another second-order consequence we can imagine in the aftermath of decreased formatting costs would be a newfound emphasis on content that is provably human-created. If AI-created sales emails reach a fever pitch, we can expect to see a counterresurgence of in-person sales efforts. As cheap and dubious content spreads online, print may find a new value as a provably costly format (print production). Anything that can show the human touch could get new value, and that is the kind of editorial influence from generative AI that many folks who bleed black ink might welcome.

Generative AI will impact journalism on several fronts such as the business models and audience experience and, finally, the editorial focus/mission of news organizations around the world. The medium is the message, and I hope that as an industry we will use the momentum of this shift in our favor.