The Short-Sighted Nature of the AI Slop Argument
Writing off AI as slop denies clear evidence of its impacts.
Unhappy with current stock images, I used Adobe Stock Firefly to generate a new image to represent the final piece of an AI puzzle.
Within a 24-hour period, I read several critics who wrote off AI announcements and content as slop. It’s easy to understand how people would view over-the-top hype and poorly or uncurated AI-created content as proof points. But to write off all AI uses as slop – regardless of the field of execution – is short-sighted at best and, in the most dramatic cases, career-ending.
Certainly, every profession and field is now experiencing poorly executed AI. This is the by-product of a technology boom filled with hype and unasked-for use cases, from bad products to crap content. However, for every poor execution, there are most certainly several good ones. While these use cases may not achieve expectations of 100% miracle-producing solid gold or pixie dust, they are good enough to create efficiencies with time spent, improved accuracy, and quality, all while generating cost savings and ROI.
The results are so impactful for some companies and even in some sectors that companies are now reducing their workforce, particularly those that perform rote and repeatable tasks. For example, Bloomberg just reported that Wall Street banks are expected to cut at least 200,000 back office and middle management jobs over the next five years. Between productivity increases and cuts, AI is expected to add $180 billion to banks’ combined bottom line.
The same article noted that AI was texted to impact more than those who lost positions but also those who remain. They will become managers of processes whose work is augmented and supported by AI. While JP Morgan noted that jobs will not be displaced, rather they will be augmented, CEO Jamie Dimon paints a utopian image of better lives with 2 ½ day work weeks.
The Marketing Sector
Let’s consider one of the professions that dominate AI criticism, marketers who declare all AI copy is slop. Marketing is already impacted by AI with a 30.37% decrease in writing jobs posted and a 20.62% decrease in web development jobs, according to Harvard Business Review. In the same time period, searches for writing jobs dramatically increased as well as new positions that included ChatGPT in the job description.
Yet, marketers are rightly slamming the poor (often unrefined) writing and image-generated content produced by AIs. They complain about AI and how terrible the content is. And it is. It’s robotic, windy, inaccurate, and visual depictions
This bad content is usually produced by an unchecked AI or, worse, a human who uses the AI and thinks that raw, unrefined, untrained output is acceptable for public content. How can people lose their jobs or get forced to incorporate these weak tools into their workflows?
Yet social media streams are littered with posts of people filled with workers who have lost their jobs, or who are struggling to find new gigs. The Klarnas of the world have used AI to automate a large majority of their content. Marketing departments are not turning back and filling new positions with workers who refuse to use AI. Worse, competing companies are forced to adopt the same or similar types of automation to simply stay in business.
The “AI is slop” argument is the cry of an endangered worker and business owner. It ignores the painful realities of the rapidly evolving job market. In reality, the dismissal of all AI as slop is a form of denial. AI adoption is becoming existential.
Adopt or Retire?
Midjourney helped me restore and reimagine a fuzzy Radiohead concert stage photo I captured at Lollapalooza 2008.
So what to do about this situation? Adopt or retire? Most people are not in a position to simply retire, and they don’t want to find new positions either. Instead, they need to embrace how AI can help them improve their workflows rather than throwing out the baby with the bath water.
First, we must begin by understanding that there are both very good and bad AI use cases. While the overhyped fails and glaring errors are the most discussed examples, less sexy examples are shaving double-digit percentages off error rates, time spent, and other cost-saving efficiencies. Here are just a few of them:
Personalized outreach campaigns that better understands context
Automated customer service powered by records from hundreds of thousands of past cases
Meeting notes and summarization
Better editing with generative-AI-enabled tools like Grammarly Pro
AI-generated stock creative developed by the buyer
Literally, there are millions of successful AI use cases now. The "AI is slop" argument ultimately reflects a failure to engage with the technology's nuances.
More Discerning Criticism Is Needed
Second, instead of dismissing AI entirely or accepting poor implementations, we need to develop a more sophisticated approach to AI criticism. One that acknowledges both AI's limitations and its practical benefits when properly applied. The future belongs not to those who blindly embrace or reject AI but to those who learn to wield it with skill and judgment.
Given this reality, direct, smart critiques of AI use cases will address the execution, not the entire movement of AI technologies across every sector. For example, criticism should be directed at the humans who publish AI slop without cleaning it up. Online targeting poorly executed advertising is as old as social media itself. Blaming generative AI is akin to blaming the car for an illegal left turn.
Instead, critiques should also look for ways to improve AI use cases and provide best practice examples of how they are using AI. If they haven’t found a good use for AI yet, why is that? It's probably because they dismissed it out of corn, fear, or some other cause.
For example, my biggest criticism is not with AI technology itself but with companies like OpenAI that present their algorithms as cure-all solutions while failing to grasp their practical applications. These companies position themselves as visionaries while essentially being toolmakers—creating powerful but unrefined instruments that require expert hands to shape them into truly useful solutions. I believe the hype cycle would be less toxic if OpenAI engaged in better product marketing.
This distinction in criticism matters. While I remain skeptical of OpenAI's approach, I regularly use AI tools like Anthropic's Claude, Adobe Firefly, and Midjourney in my daily work. The result? I've gained back roughly two hours each day – time now spent on strategic thinking rather than routine tasks.
In conclusion, we should not debate the wholesale adoption or rejection of AI. Our focus should be on understanding AI as a toolset that demands a critical eye, discerning use, engaged human management, and professional refinement.