It wasn't that long ago that AI-generated images were a party trick. You'd type a few words into a box, wait a few seconds, and out popped a picture. Sure, the person in it might have six fingers, the sign in the background might be written in something resembling an alien language, and the dog might have an extra ear, but it was still amazing. We showed those pictures to our friends and said, "Look what AI made!"

Then the voices got better. Then came video, then realistic avatars, then lip syncing good enough to make someone appear to say something they never said. All of it happened faster than I think any of us expected, and somewhere along the way the question people were asking changed. It used to be "Can AI really do that?" Then it became "Is that real?" Now I think we're entering a third stage, where the question is simply "Who cares?"

Here's a detail I find telling. Merriam-Webster's Word of the Year for 2023 was "authentic." Two years later, in 2025, it was "slop," which the dictionary defines as low-quality digital content churned out in bulk by artificial intelligence. That's quite a trip to make in just two years.

You've probably felt it yourself scrolling through Facebook, LinkedIn, Instagram, TikTok or YouTube. Another perfectly polished AI image, another inspirational story that sounds vaguely familiar, another fake movie trailer, another talking avatar, another list of "10 things successful people do before breakfast." And somewhere in the comments, someone inevitably writes, "Enough with the AI slop." That's an interesting turn of events when you think about it, because AI didn't become a problem by producing content that looked bad. It became a problem because the content got too good, too cheap and too plentiful.

We Solved the Wrong Problem

For as long as I've been in business, content has been a headache. You need blog posts, videos, social media posts, graphics, emails, presentations and advertisements, and producing all of it takes time and money that most small businesses don't have. Then generative AI showed up and seemed to solve the problem overnight. One person could suddenly produce what used to take an entire team, and in a lot of ways that really is wonderful.

But whenever something becomes almost unlimited, its value changes. If everyone can create a beautiful image in 30 seconds, a beautiful image isn't special anymore. If everyone can produce a professional-sounding article, sounding professional isn't enough, and if anyone can create a polished video of themselves without ever standing in front of a camera, a polished video doesn't prove much of anything. We created an unlimited supply of content, and now we're discovering that unlimited supply brings its own problem: how do you decide what's actually worth paying attention to?

Seeing Used to Be Believing

For most of human history, seeing something carried a certain amount of proof. A photograph was evidence that something happened, and a recording meant someone actually said those words. You could argue about context or editing, but you generally assumed the person on the screen had really stood in front of a camera.

That assumption is going away. You can now create a convincing photo of a place that doesn't exist, clone someone's voice, generate a realistic person who has never been born, or put words into a real person's mouth, all from a text prompt. And the technology keeps improving. Right now we still look for weird fingers or unnatural blinking, but pretty soon there may be nothing obvious to spot at all.

That creates a strange psychological effect. When anything can be fake, everything becomes questionable, including the real stuff. You see an unbelievable video and instead of thinking "Wow," your first thought is "AI." Sometimes you'll be right and sometimes you won't, but the suspicion is always there now.

The Real Problem Isn't Fake Content

Fake content gets most of the headlines because it's scary. Deepfakes, fake news reports, fake celebrity endorsements, fake political videos and fake customer testimonials are all legitimate concerns. But I think there's another problem that may hurt businesses even more, and it's a lot less dramatic.

People are getting tired.

Not frightened, just tired. Nobody wants to become a forensic investigator every time they open Facebook, studying shadows and reflections to figure out whether a picture is real, or wondering whether the article in front of them was written by a person or a piece of software. So they do something much simpler. They scroll. For marketers, that's the real danger. It isn't that people will hate your AI-generated content. It's that they may stop noticing it at all.

The Authenticity Economy

That opens up an interesting opportunity. If artificial content becomes unlimited, then genuine human experience becomes more valuable. We spent decades trying to make business communication more polished, with better cameras, better microphones, better lighting, professional scripts and corporate photography. Now the thing that grabs your attention might be somebody holding an iPhone and saying, "I just discovered something, and you've got to see this." It works because it feels real. There might be a dog barking in the background, the lighting might be terrible, and the person might stumble over a word, but those imperfections are evidence that an actual human being is involved. We may even reach the point where perfection itself creates suspicion, which after all those years of chasing polish would be quite a reversal.

This Doesn't Mean You Should Stop Using AI

Some people are reacting to the flood of slop by saying businesses should stop using AI altogether. I think that's the wrong conclusion. AI is an incredibly useful tool, and I use it constantly (including, I'll admit, on drafts like this one). The mistake is using it to replace the things that make you uniquely you: your stories, your experiences, your opinions, your customers, your expertise, your mistakes and your personality. Those are exactly the things a business should protect. AI should help you communicate them better, not invent them for you.

There's a big difference between using AI to organize your thoughts and asking it to pretend it had your experience, just as there's a difference between using AI to touch up a photo and using it to generate a customer who doesn't exist. One is amplification and the other is fabrication, and that distinction is going to matter more every year.

Your Imperfections May Become an Advantage

One of the stranger consequences of all this may be that being a little messy becomes valuable. A handwritten note, a live Q&A, a spontaneous video, a photo from an actual event, a customer telling their story in their own words, an unscripted podcast conversation or some behind-the-scenes footage were never considered especially "professional." Now they may become trust signals, precisely because they're hard to fake convincingly at scale. And even when they can be faked, people can tell the difference between a brand that consistently shows real people doing real things and one that publishes an endless stream of synthetic perfection.

This matters most for coaches, consultants and experts. People aren't just buying information from you. They're buying access to your experience, your judgment and your way of looking at a problem. If AI makes generic information almost free, those human qualities are what's left to sell, and they just went up in price.

Knowing What Not to Automate

I don't think we've fully processed what happens when anyone can generate unlimited images, voices and video. We still treat AI-generated content as if it were unusual, but it isn't, and very soon synthetic media will be completely ordinary. When that happens, having access to AI won't be a competitive advantage because everyone will have it. The advantage will belong to the people who know what not to automate.

By all means, use AI to brainstorm, research, edit, summarize, create graphics and save yourself time. Just don't automate the humanity out of your business. A few years ago the goal was to make AI content look human, and the next challenge is going to be proving that human content really is human.

Which brings me back to Merriam-Webster. We went from "authentic" to "slop" in two years. AI was supposed to make content more valuable by making it easier to create, and instead it may have made something else far more valuable. It happens to be the word we picked right before all this started.