Journalism, journalism education and ChatGPT
When the world changes, take the opportunity to change with it
Back in the early 1990s, United had this commercial:
For some reason, it’s always stuck with me.
I think it has less to do with the whole “value of doing business face to face” thing and more to do with the idea of taking a negative and trying to turn it into a positive. Something happens in the world, and you either lament the change or you adjust to the way the world is and turn that change into an advantage.
It’s how I’m thinking about journalism, journalism education and ChatGPT.
ChatGPT, the open-AI platform that takes prompts and turns them into writing that sound shockingly human, is turning the writing world and the education world on its ear. Teachers at all levels - especially high school and college - are concerned that this could lead to harder-to-detect plagiarism.
The actualities of ChatGPT seem, to me, to be less of an immediate “threat” (for lack of a better word) to journalism and journalism education. On the page where you type your prompt, one of the limitations explicitly stated is “limited knowledge of world events from before 2021,” and journalism primarily deals with new events.
Because I’m contractually obligated by the Rules of Internet Discourse to do so, I typed a prompt into ChatGPT to show you as an example. My prompt was “Write a news story about the Buffalo Bills game with the Cincinnati Bengals on Sunday.” (What can I say? I’m a masochist.)
The result:
The Buffalo Bills took on the Cincinnati Bengals in a thrilling game on Sunday, with the Bills coming out on top with a final score of 24-21.
The Bengals got off to a strong start, with quarterback Joe Burrow leading the team to two early touchdowns. But the Bills quickly regained their footing, with quarterback Josh Allen throwing for two touchdowns of his own in the second quarter to tie the game at 14-14 going into halftime.
The second half was a defensive battle, with both teams struggling to find the end zone. But with just over four minutes left in the game, Allen threw a crucial touchdown pass to wide receiver Stefon Diggs to give the Bills the lead. The Bengals had one last chance to tie the game, but their drive fell short and the Bills secured the win.
The victory moves the Bills to 3-0 on the season and solidifies their spot at the top of the AFC East. The Bengals, meanwhile, fall to 0-3 and will look to regroup next week.
Josh Allen had a great game, passing for 288 yards and 3 touchdowns, Stefon Diggs had 5 receptions for 103 yards and a touchdown. Both teams will play next week, the Bills will face the Las Vegas Raiders and the Bengals will face the Baltimore Ravens.
If wishing made it so, right?
But let’s move the conversation beyond this specific platform, because this is a bigger issue than the capabilities and limitations of one page and focusing on the trees loses sight of the forest.
The fact is, in sports journalism, we’ve been dealing with this for a good 5-10 years now. The Associated Press, since 2016, has used Automated Insights to create short recaps from minor-league baseball games and assorted college events. These are the classic three-to-five paragraph inverted pyramid recaps written straight off a box score. There’s no human touch, but a human touch isn’t needed for these stories. It’s final score, high scorer, top rebounder, records, that kind of thing.
In theory, this type of software frees up journalists to write deeper game stories, trend features, profiles, and spend their time working on things other than basic roundups. In practice, jobs have just been eliminated.
This is something I’ve thought and written about before. In my sports writing classes, I always start with the classic three-to-five paragraph inverted pyramid recap. But as AI gets better, I struggle with justifying this. Should I spend any of the limited time I have with my students teaching them a skill that a shell script can do just as well?
I always justified it by saying that the basic game story was like the scales for a musician, that you had to learn those before you could move on to bigger stories.
I still think that’s true. But ChatGPT has got me thinking about it with a little more nuance. And that’s the real value of its emergence in writing and education. Like the boss in that United Commercial, it gives us a chance to reevaluate what we ask our students to do and why we have them do it.
That compelling why — something that’s come up in a lot of aspects of my life recently — is where we have a good discussion.
For me, learning the basic game story structure that Automated Insights can create still matters. But why? Let’s move beyond the generalities of “before you run you have to crawl” or the wistful of metaphors about musicians, and get to the nitty gritty about why learning the basic game story matters.
Because it teaches you, as a reporter, to focus. It teaches you to be able to identify the most newsworthy element of a game or an event, and to do so quickly. It teaches you to take a mass of information (a box score and a play-by-play sheet) and pick out the information that is most important for your readers to know. It teaches you economy of language, how to tell a story succinctly and directly.
Now that doesn’t mean we should spend whole semesters, whole units, or even more than a week on the basic game story. I did that in the past, but eventually I accelerated the timeline, so students were moving on faster to other types of stories. The kinds that, as of yet, can’t be written by AI.
But maybe, like the boss in that United commercial saw getting fired as a chance to actually reconnect with customers, the emergence of ChatGPT and other AI programs give us the chance to ask those questions, to justify what I’m teaching and making sure my focus is where it should be.
On what my students need.