In Defense of Humans: The Self Limiting Nature of AI is Becoming Apparent

Can anyone else see the irony of how AI is changing the content marketing landscape in 2024?

This recent BBC article title should say it all: “AI Took Their Jobs, Now They Get Paid To Make It Sound More Human.”

Ponder that for a moment. The LLM AI engines like Chat GPT 4 need to employ humans to make it sound more human. The jobs AI displaced created new jobs, but jobs meant to prop up the inescapable conclusion that AI isn’t truly “intelligent” or able to replace a human.

This is due to one essential reality. AI does not exist in the human world, humans do. And thus AI is limited both in it’s ability to understand as a human would, and also to create new content from the human world. It is self-limited by the content available to it from the digital world.

I’ll get to that point in a minute. But first, let’s look at the irony of humans propping up AI to make it seem more human.

The Humans Propping Up AI

Why do they need to make it sound more human? The answer to that is equally ironic.

Why does a company pay for a copywriter to make a blog post? Because in their content marketing model, they want well written, helpful content that will get ranked by Google and drive traffic to their site.

Why do companies now have to pay humans to make AI content sound more human? Because Google, a primary funder of AI research, is at the same time penalizing and de-ranking low quality content written by AI.

Now why would Google do that? Because their techs recognize what all of us know anyway. Pure AI written content is mostly trash, filled with tremendous inconsistencies, factual errors and weird flowery language no real person would ever say unless English was their fourth language and they learned it by watching old Shakespeare plays.

Google’s bottom line is to deliver content to people in search engines, not AI bots, and that content needs to be something a human WANTS to read and actually finds helpful.

Haven’t you noticed in the last year some of the obvious AI-written garbage – listicles, how-tos and the ilk – filling the search results with difficult to read and sometimes blatantly incorrect content? Who wants to read that?

Companies figured out pretty quick that they still needed humans behind the curtain of the Great Wizard of AI, or it isn’t really going to produce value. If you can’t see through the hype by now, you aren’t watching the same show.

Thus the bitter irony that companies who went all in on AI, fired their content teams, and plugged concept outlines into Chat GPT or other engines fueled by it, now need to hire humans to fix the human-incompatible content that AI is spitting out.

What is the Self-Limiting Nature of AI?

The practical problem above – humans needing to rewrite AI content to make it sound more human – is a symptom of a deeper underlying reality of the self-limiting nature of AI.

What is that reality? AI does not exist in the human world we physically inhabit, is not human, and thus cannot speak in a way that connects with humans. They cannot create new content INSIDE the digital realm from the OUTSIDE human world.

AI engines generate new content for humans in the human world ONLY from their vast pool of existing content already created by humans and put into the digital world.

They cannot CREATE new content from the human world. Why? Because AI doesn’t exist outside its algorithm in our world. All that AI is doing is analyzing existing content and generating content we prompt it to from within that pool of information.

When you ask AI to write a five-paragraph blog post about how to bake the best chocolate chip cookies in the world, it can only create that content from what we, as humans, already gave it for content on thousands of baking blogs. It does not know what a cookie is. It cannot eat a cookie. It cannot go to the kitchen and try some new recipes that nobody every created before. It is LIMITED to the content we gave it on the internet. We humans are the ONLY source that feeds in new creative content to the AI input loop.

How is this a self-limiting reality? As more and more companies boot their human content teams and start generating everything on the web from AI, they are essentially CLOSING THE INPUT LOOP in the digital realm. When all the baking blogs are writing their recipes with Chat GPT, we are only recycling information, not creating it.

Because AI exists within the digital realm, if less and less humans are inputting creative content from the real world to the digital world, AI becomes less and less able to generate anything meaningful for lack of fresh human content.

AI cannot create from nothing or “think” creatively. It is a misnomer in our team’s opinion to refer to it as “Artificial Intelligence” despite current market terminology. A better way to view it is “Algorithmic Intelligence” as the processes going on behind the scenes for AI tools are just highly complex math functions worked out via prompts to their algorithm. There is no real thinking happening. No real intelligence. No genuine novel creation of new ideas.

This also creates the problem of the AI engines scraping AI content on other websites, then annexing and analyzing that for new content generation. We are slowly but steadily moving on from AI generating material from human content, to AI now recycling it’s own AI content as well.

This is creating an inevitable downward spiral of increasingly bad content on the internet. By closing the input loop from humans and prompting AI to recycle its own garbage content, we are logarithmically and exponentially increasing content across the web that no human wants to read and no Google search bot wants to index.

AI Didn’t Replace Their Jobs, Delusional Managers Did

The headline in that BBC article is also a bit misleading. AI did not take their jobs. AI is not a person or entity capable of applying for a job. No, they lost their jobs due to the delusions of profit-driven managers who saw what they thought was an opportunity to replace the salaries of 60+ people with a monthly Chat GPT subscription.

I doubt these managers had much technical background or understood the inherent limitations of LLM AI at its current stage. They heard the hype, saw the bandwagon passing by, and on they jumped to temporarily boost their profit margins by cutting staff costs.

Do not think when you read the headlines that AI is taking anyones job. The new wave of junk content littering the internet is not the fault of AI. It is the fault of people seeing what they they think is a golden opportunity and using AI in a way it is not capable of being used.

But what the opportunists will find (and I think are already finding) is that their temporary profit boost due to staff cuts will NOT offset the long term damage to their brand reputation and content marketing efforts due to the prolific stream of AI-generated garbage they’re outputting.

Google certainly isn’t going to reward that with rankings and traffic. And the human customers they’re trying to reach? They don’t want AI junk articles either. Humans want to connect with other humans. The brand that pursues that in this current landscape will win out in the end.

But what a degrading and soulless task for the human to have to rewrite the content from the bots that their managers thought would replace them! Wouldn’t it perhaps have been a good idea to just keep the humans writing content for humans in the first place?

AI is a tremendous tool (just in case you think I am completely anti-AI) when used in conjunction with a competent human writer/marketer. The better managers out there in the digital marketing space figured out early on the limitations of AI. They did not think it was a suitable replacement for human staff, but saw it as a very suitable supplement for their human staff.

Recognize the Limitations and Usefulness of AI for Humans

I’m certainly not advocating we burn AI to the ground and go back a decade as if it never happened. Things don’t work that way. We adopted AI tool sets for enhancing certain content writing workflows years before Chat GPT 4 ever hit the headlines.

But we didn’t fire our human team in the process! We instead used AI to add greater value for our clients, recognizing the limitations but exploiting the usefulness of AI when it is in the hands of a competent human staffer.

At the end of the day, our job is to help our human clients connect with human customers looking for their products to use in the real, human world. AI cannot accomplish that without a human at the helm.

Maybe the day will come when we have a C-3PO walking around in the human world, interacting, understanding and analyzing the human reality in a way that is useful to humans and connects with us. But that day is not yet, and current AI models have real limitations we have to acknowledge.

If you’re reading this and you are a company that fired off your staff to buy in to AI as a total replacement solution, I hope you are having buyer’s remorse. Bring back your humans now before it is too late! You will be surpassed in the digital landscape by companies that understand the necessary synergy between human-led AI usage. You will not connect with humans using bots, only with other humans.

If you have any thoughts or questions about AI usage in digital marketing, feel free to reach out to us with questions or comments. You can contact our team here: Contact Us

Side Note: The banner image used in this post was generated using AI imagery tools from the tool set. Take note of some of the visual inconsistencies even in this AI imagery created from a paid plan. And yet at the same time, the illustration is not half bad!

Sign up to get our latest articles

Don’t worry. We won’t sell your email. We are also really busy managing our clients, so we won’t be filling your inbox with articles every day. We only write them when we have a compelling reason to do so, and some spare time too!