At the risk of sounding self-serving (you’ll see why below), isn’t it time we really looked at how we could be improving the basic unit of news — the story?
The core narrative structure of the news article — lede (whether anecdotal or newsy), nut graf, context, analysis, to-be-sures, all carefully woven into a tight package intended both to pull readers along and discourage editors from touching a single word — has been around for decades, mostly impervious to new forms of reporting, analysis, publishing, and distribution.
Despite huge changes in the technology of news, the structure of a story today doesn’t look hugely different from one in, say, 1932. Sure, there may be slideshows embedded, video added, interactive graphics and so on, but at heart, the core architecture remains the same.
Should it? Maybe it remains the best way to get important information most efficiently into readers’ minds. Or maybe it’s just us journalists doing what we like to do best.
At Semafor, we’ve taken the story form apart and recreated it in sections that delineate news from analysis from counter-argument from different perspectives. It seems to be working well, but is naturally a work in progress. Axios, with its trademarked “smart brevity’ format, has solved for the needs of overloaded and time-strapped readers. Google’s now-dead Living Stories was an exciting experiment in understanding what information readers already know and didn’t need repeated. Homicide Watch DC was likewise an interesting foray into rethinking news judgment (let’s report on all murders in the capital, not just the “newsworthy” ones), reporting structure (gather facts and let technology help assemble them) and building new audiences (family and loved ones of homicides otherwise deemed not worthy of coverage.)
And yet these are all still on the margins of the industry. (Although Semafor and Axios obviously have no desire to stay there…)
What more could we do, especially now that we have far more powerful technology at our disposal?
How about using ChatGPT‘s powerful language parsing and generation capabilities to turn the news experience into the old saw about news being what you tell someone over a drink at a bar? “And then what happened?” “Well, the FBI found all these documents at Mar-a-Lago that weren’t supposed to be there.” “I don’t understand — didn’t he say he declassified them?” “Actually…”
It would let readers explore questions they have and skip over information they might have. In other words, use technology to treat every reader as an individual, with slightly different levels of knowledge and levels of interest.
Or perhaps harness that with engines like Stanford’s Big Local News project or USC’s CrossTown data analytics platform, which pull and automatically analyze reams of local data, find interesting patterns and send them to reporters, either for follow up or as parts of broader stories. Reuters and Bloomberg already have systems like that for financial data; why not build them into more newsrooms?
In the end, we should be taking full advantage of what tools are out there, marrying the best of what machines can do with the best of what humans can do — not to replace each other, but to create a smarter, faster, “cybernetic newsroom” that can serve readers and communities better.
After all, we’ve changed dramatically as a news consuming public over the decades; think about how Twitter threads, TikTok videos, and interactive graphics have all burrowed their way into our news habits. Why shouldn’t our most basic story form change as well?
Gina Chua is executive editor of Semafor.