It should have been obvious when college students started using ChatGPT to cheat on academic papers that the bots were coming for our bylines — and yet, I think, journalists thought they might be spared. After all, any bot could mimic a badly written college essay, but what journalists do? That requires finesse, skill, and critical thought.
At the same time, readers are about to be inundated with bot-generated content. Most of it will be a drag to read — but so is much of modern writing.
Journalists like to paint themselves as brave propagators of truth, without which modern political discourse would crumble into conspiracy theories and vague interpretations rife with half-truths that contribute to widespread insanity. There’s a sort of mythical journalistic figure — perhaps created by movies like All the President’s Men — who operates in the shadows, forces the truth out of sources, and digs through mountains of evidence until corruption is uncovered. It’s a tough thing to aspire to, and unsurprisingly, most of us just scrounge around on social media until we find something to write about.
That, it turns out, is a job ChatGPT (with a little guidance and some fact checking) is just as good at.
ChatGPT Is Invading Opinion Journalism
When the Italian newspaper Il Foglio decided to release a weekly AI insert in its physical magazine and announced that, on occasion, it would be using ChatGPT to write articles in its standard paper (clearly labeled, of course), it caused something of a stir.
News organizations have been experimenting with AI for a while. The Atlantic announced that it would be partnering with OpenAI last year, and the Washington Post’s in-house AI bot, Heliograf, reportedly pumped out some 850 articles in its first year. Some experiments haven’t gone incredibly well; famously, several major newspapers published a syndicated summer reading list with 10 nonexistent titles “hallucinated” by a chatbot just a few weeks ago.
There are, to oversimplify journalism a bit, three general kinds of articles. At the bottom of the pile are the lists, the kind that might be titled “The Top 25 Things Happening This Summer in Your Area,” or “5 Habits to Work on in the New Year” and which tend to be popular in Google search results and take an intern with minimal skill about an hour to write. ChatGPT has no problem chugging these out at a mind-boggling pace.
In the middle you have your everyday news article about the things that happened yesterday: Israel bombed Iran, this girl was forced to watch porn to graduate college, etc. These are also easy fodder for ChatGPT.
On the high end you have investigative reporting and opinion journalism. These are the articles that require either critical analysis or developing sources, and sometimes both, and they tend to be written by journalists who thought their jobs were safe. What Il Foglio’s experiment proves is that those jobs aren’t safe. It’s using ChatGPT to write opinion articles about politics — the kind of articles that journalists (especially those of us in the magazine business) thought needed a critical thinker and a talented writer behind them.
To be sure, it’s not like ChatGPT is going to turn into H.L. Mencken or Hilaire Belloc anytime soon, but neither are most opinion journalists. “If you are not a journalist with enough creativity, enough reporting, enough ideas, maybe you are worse than a machine,” Il Foglio editor Claudio Cerasa told the Atlantic. “But in that case, the problem is not the machine.”
The Age of AI Journalism Isn’t All Dystopia
To be sure, the great thinkers and astute investigators in the journalism world will likely still have a job when all is said and done. After all, LLMs can’t develop sources, engage in thoughtful conversations, come up with new ideas and connections, or develop a unique writing style — real live people can.
This all means that journalists as a whole (not just opinion or investigative journalists) will have to view themselves differently. They’re no longer giving readers just the facts. Objective journalism is so dead that even the lie that it exists no longer sells papers. The only thing that will sell papers is real opinion. Most of it will likely be wild conspiracy; some of it — the best of it — will contain perceptive ideas.
Those of us writers who claimed to be objective because we were neither creative enough to concoct outrageous conspiracies nor genius enough to issue prophecies on the printed page will be going into marketing, editing, or some such related field presently. Those of us who are outgoing enough to develop extensive sources likely don’t have time for writing anyway, and the chatbots will ease our schedules a bit.
At the same time, readers are about to be inundated with bot-generated content. Most of it will be a drag to read — but so is much of modern writing. There will be just a small market for authentic writing available to those who are willing to help pay the salaries of the few remaining human writers.
It’s a reality that most of us haven’t come to terms with just yet, but it’s not all dystopia. AI will force those writers who remain in the field to excel because lazy writing won’t bring in a paycheck. This brutal school of “write better or go hungry” is what the modern age needs to overcome the myth that excessive content is better than quality content.
READ MORE from Aubrey Harris:
A Rare France Win in the War on Porn
Yes, AI Is Taking Jobs From the Class of 2025. No, We Shouldn’t Be Concerned.
