Why We Won’t Use Generative AI at Northern Express

The cost (and failures) of artificial intelligence

By Jillian Manning | June 7, 2025

Some days, I think I’ve woken up at the start of a robot apocalypse movie. Just about every app, every search bar, every website shouts, “Use our new AI function!” It’s on Google. It’s on Instagram. It’s even on the City of Traverse City website.

And I don’t trust it.

Don’t get me wrong—I’m a Millennial who has always been within cherry-pit-spitting distance of the Internet. But this latest step in our technological evolution spells doom in the eyes of that same Millennial who also grew up on The Matrix, The Terminator, and Avengers: Age of Ultron.

How did so many of us forget that in the war of humans vs. robots, humans only win at a terrible cost?

What Is Generative AI?

Artificial Intelligence, or AI, is machine or computer-based intelligence. Versions of AI have existed for decades. It’s the computer playing against the chess master. It’s the recommendation for a new movie you’ll like on Netflix. It’s that gal Alexa who organizes your life.

But generative AI—the type of artificial intelligence that creates something, like ChatGPT (text), Midjourney (images), or Sora (video)—is newer and more disruptive to every field from law to art to journalism. That’s the AI we’ll be talking about here.

AI Is a Thief

Generative AI didn’t just appear out of nowhere. When you ask these programs to create something for you, they can only do so because they have trained off human work, aka a dataset. In many cases, that dataset was stolen from copyrighted content.

For example, earlier this year, The Atlantic found that Meta, which owns Facebook, Instagram, and Threads along with several virtual reality products, “pirated millions of books to train its AI.” You can search the books (which include works from bestselling northern Michigan authors like Brittany Cavallaro and Doug Stanton) here: theatlantic.com/technology/archive/2025/03/search-libgen-data-set/682094.

Generative AI has also trained on your favorite movies and TV shows, which is why it can spit out an email that sounds just like Moira Rose from Schitt’s Creek. It has trained off the artwork of photographers, painters, illustrators, and cartoonists, all so you can share that picture of what you’d look like as an anime character or Barbie doll on social media.

In almost all cases, the copyright holders and creators of the original content were not asked for their permission, nor were they compensated. They were, with no exaggeration, stolen from. Lawyers are now wading into the uncharted territory of intellectual property (IP) vs. AI, trying to find a way to protect creators from IP theft.

And even though we know these companies are stealing, they won’t stop. Because—wait for it—Business Insider reports “firms like Meta and A16z” are fighting copyright rules because they admit it would cost too much to legally acquire the content they need.

Seriously, AI Is Stealing from You

Even if you haven’t written a book or made a movie, AI is probably still training off you. In fact, most social media sites are using your content to train their AI programs.

Meta says this about where their AI training information comes from:

“We use information that is publicly available online and licensed information. We also use information shared on Meta’s Products and services. This information could be things like posts or photos and their captions. We do not use the content of your private messages with friends and family to train our AIs unless you or someone in the chat chooses to share those messages with our AIs.”

But wait—there’s more: “Even if you don’t use our Products and services or have an account, we may still process information about you to develop and improve AI at Meta.”

The kicker: Meta does not allow U.S. customers to opt out of having data used to train AI. All you can do is make sure your accounts are set to private, as Meta has said they won’t use data from private accounts…at least for now. (They are currently embroiled in EU lawsuits over the use of personal data to train their AI models.)

AI Is a Liar

So we know AI is training off words, photos, videos, and more using everything from your Facebook page to Alice in Wonderland. With all that information, surely it will spit out the correct information, right?

Nope.

Most AI engines have some sort of disclaimer attached to them, but it’s getting harder to separate fact from potential AI fiction. For example, when you get an AI Overview at the top of a Google search, you can click a “Learn More” button that explains (when you scroll waaaay down):

“AI Overviews use generative AI, which is a type of artificial intelligence that learns patterns and structures from the data it is trained on and uses that to create something new. While exciting, this technology is rapidly evolving and improving, and may provide inaccurate or offensive information. AI Overviews can and will make mistakes.”

Closer to home is the aforementioned AI used by the City of Traverse City, called “Ask the City,” which is powered by GovGPT. When you open the window to ask the bot a question, a disclaimer says: “GovGPT may produce inaccurate information about people, places or facts.”

Traverse City Commissioner Heather Shaw has repeatedly admitted to using ChatGPT to inform her comments or votes. Per our sister publication, The Ticker, Shaw said in November 2024 that ChatGPT is “an amazing resource for legal opinions.”

And yet generative AI has often been in the news for its foibles, especially in law. Bloomberg Law reports that “In June 2023, two New York attorneys filed a brief written by ChatGPT, which included citations to six nonexistent cases and erroneous quotes. This was one among several high-profile incidents that highlight some of the risks of using artificial intelligence for legal professionals without strong oversight or scrutiny.”

These kinds of made-up answers and false facts are called “AI hallucinations.” AI is getting more accurate as it gets fed more data, but in the meantime, it can and will spit out incorrect or invented information.

How can you know if an AI generated response is fact? You’ll have to look it up and see if reputable sources corroborate the same information. (Like you would have if AI weren’t around to give you the answer.)

AI Is a Cheat

All right, well if we can’t trust AI to give us the truth, the whole truth, and nothing but the truth, at least it can do our work for us!

Many adults use AI to write emails, reports, resumes, and more. TheVerge.com found that as of December 2024, ChatGPT had over 300 million people using the AI chatbot each week.

(And students are using it to write their papers, do their homework, and skip the pesky chore of, you know, actually learning. If you haven’t yet read New York Magazine’s report “Everyone Is Cheating Their Way Through College,” do so now…but prepare to be horrified.)

But all that AI-outsourced work comes at a price. Microsoft—which ironically has its own generative AI programs—and Carnegie Mellon University published a study called “The Impact of Generative AI on Critical Thinking” in early 2025. The key finding was that “…the more humans lean on AI tools to complete their tasks, the less critical thinking they do, making it more difficult to call upon the skills when they are needed” (Gizmodo).

AI Is a Killer

The final AI nail in the coffin is its effect on our planet.

The United Nations declared this past fall that “AI has an environmental problem,” noting that “proliferating data centres that house AI servers produce electronic waste. They are large consumers of water, which is becoming scarce in many places. They rely on critical minerals and rare elements, which are often mined unsustainably. And they use massive amounts of electricity, spurring the emission of planet-warming greenhouse gases.”

Teen Vogue reports that “one estimate found that a single ChatGPT search uses 10 times the energy of a normal Google search.” And in September 2024, The Washington Post found that a “100-word email generated by an AI chatbot using GPT-4 once requires 0.14 kilowatt-hours (kWh) of electricity, equal to powering 14 LED light bulbs for 1 hour” or “519 milliliters of water, a little more than 1 bottle.”

That’s just emails, and short ones at that. For those keeping score, this story is about 1,700 words. If ChatGPT had written this for me, I would have 17 bottles of water lined up on my desk.

AI Is Here

It’s too late to turn back the clock—Pandora’s evils are out of the box. But, if you know your Greek myths, you know that when all the worst things have come out of her can of worms—sickness, death, etc.—the box still contains hope. (Some read that version of hope as a gift, others as a curse. We’ll be optimistic here and go with the former.)

Here’s what I hope: I hope that when people turn toward generative AI, we are informed about its pros and cons. That we demand better protections for intellectual property and privacy. That we do not allow it to replace human art. That we use it sparingly and intelligently. That we put our planet first.

I can tell you how we will use AI on the editorial team at Northern Express. We’ll use spellcheck. We’ll use services that transcribe our interview recordings to text. We’re a small, lean team, and we’re not against new technology. If you are ever curious or concerned about our use of AI, I encourage you to reach out.

But we won’t use generative AI to write or produce our content. You will never see a byline like one I recently spotted in The Battle Creek Enquirer, which said, “Janis Reeser, reporter assisted by AI.” Not ever, not on my watch as editor.

We will continue to get our quotes directly from individuals in our community and our research from trusted organizations, publications, and resources. We will have real journalists writing real stories about real people, places, and events in northern Michigan.

We are creative, curious humans. We love our craft. And we take our jobs far too seriously to let a robot do them for us.

Jillian Manning has served as editor of Northern Express since 2022 and was a contributor for several years before that. She is passionate about protecting the integrity of journalism, publishing, and media.