I wish the rise of the machines would happen and artificial intelligence (AI) would get to enslaving humanity.
Either that, or we humans need to remember that AI systems are only as smart as the people building them and the models they are learning from.
If we did that, we could avoid needless embarrassment, especially in the journalism community.
On May 18, the Chicago Sun-Times and The Philadelphia Inquirer published a special supplemental section that included books to read during the summer.
Summer reading lists are common forms of content in the media world, and they usually have great suggestions for novels to enjoy by the pool or while traveling on vacation.
The insert, which was titled “Heat Index: Your Guide to the Best of Summer,” was licensed from a company called King Features, which is a subsidiary of the publisher Hearst Newspapers, and it was filled with errors.
In fact, of the 15 books listed, only five were real.
Sure, the authors we real, but AI made up the titles and synopses.
In an interview with 404 Media, Marco Buscaglia, the creator of the special supplemental section, said the following: “I do use AI for background at times but always check out the material first. This time, I did not and I can’t believe I missed it because it’s so obvious. No excuses. On me 100 percent and I’m completely embarrassed.”
Within days, Buscaglia was fired by King Features.
Of course, mistakes happen, and many media organizations have cut book reviewers from their staffs.
But even high school and college students know to double-check their cheating. After all, AI can’t be trusted explicitly. It is known to hallucinate, or “provide users with fabricated data that appears authentic,” as MIT Sloan Teaching & Learning Technologies explained.
Just like author Lena McDonald who left an AI prompt — “I’ve rewritten the passage to align more with J. Bree’s style, which features more tension, gritty undertones, and raw emotional subtext beneath the supernatural elements” — in her novel titled “Darkhollow Academy: Year 2,” Buscaglia was lazy.
Alex Mahadevan, a co-author of an AI ethics handbook produced by the Poynter Institute for Media Studies, agreed.
“I don’t even know where to start with this massive screw-up. Right when newsrooms are trying to build trust with audiences about legitimate ways they’re using AI, this guy comes in and derails it,” he said in a Poynter article by Tom Jones.
Subscribers were rightly outraged, too, as highlighted by Nieman Lab.
Sadly, this isn’t the first time AI has given journalism a black eye.
In August 2023, newspaper chain Gannett saw its experiment of using AI to cover high school sports fail spectatcularly when the Columbus Dispatch published the following with the LedeAI byline: “The Worthington Christian [[WINNING_TEAM_MASCOT]] defeated the Westerville North [[LOSING_TEAM_MASCOT]] 2-1 in an Ohio boys soccer game on Saturday.”
Then, Gannett, which publishes USA Today, had to correct that story and numerous others with AI-induced errors.
Though this should have provided a cautionary tale, Sports Illustrated failed to get the message because in November 2023 the Associated Press reported that the magazine was using AI to write stories under AI-generated bylines.
According to David Folkenflik at NPR, the idea was that “the fake authors were meant to give the AI-created post credibility.”
Such dishonesty has no place in the truth-telling business of journalism. The media already struggles with maintaining audience trust, which makes such actions indefensible and unforgivable.
So next time any journalist considers using AI to cut corners, I hope they channel HAL 9000 and say to themselves, “I’m sorry, Dave. I’m afraid I can’t do that.”
Todd R. Vogts, Ph.D., is a native of Canton, a resident of McPherson County, and an assistant professor of media at Sterling College. He can be contacted with questions or comments via his website at www.toddvogts.com.
