Published by Kiki Choda
Last updated September 2026
Why Your Business Sounds Like Everyone Else’s and What’s Really Behind It
The current conversation around artificial intelligence is largely focused on a technical concern. There is growing evidence that when models are trained on data generated by previous models, the quality of their outputs begins to degrade. Errors compound, variation reduces, and over time the system becomes less reliable. Researchers have termed this “model collapse.”
The evidence is clear. Oxford University’s OATML Group has outlined the mechanism in detail. A landmark paper published in Nature (Shumailov et al., July 2024) demonstrates that indiscriminate use of model generated content in recursive training leads to a measurable loss in output quality.
This is being treated as a pipeline problem. A question of inputs, datasets and training methods. But there is another issue developing alongside it that is far less discussed, and arguably more immediate.
It is not just what is going into AI systems that is changing. It is what humans are no longer contributing.
Over the last 24 months, and with increasing pace as successive AI models have been released, such as GPT, Gemini, Claude, Copilot now embedded as standard into the tools people use every day, I have observed a noticeable change in how people engage with these systems. Initially, AI was used to support thinking. It helped structure ideas, explore options and speed up execution. That was its strength.
Now, in many cases, it is being used in place of thinking.
The distinction matters.
What I am seeing is not simply a decline in original thinking over time. It is the active outsourcing of it. People are no longer working through ideas before they ask the question.
They are asking the system to generate both the question and the answer, and accepting the result with very little interrogation. This shows up in consistent and telling ways.
Business owners describe their company in terms that sound complete, but when asked what makes them different, the answer becomes vague or generic. Marketing teams are producing more content than ever, yet much of it reads as if it could belong to any competitor in the same space. Executives are expecting faster responses and quicker turnaround, which is understandable, but the depth of thinking behind those responses is often missing.
The output looks right. It sounds credible. It is often well formatted, but it is thin. Not because the tools are incapable. Because the thinking that should sit behind the output has not taken place.
This is where the problem begins to compound.
If AI systems are increasingly trained on generated content, and humans are simultaneously reducing the amount of original thought they contribute, then both sides of the equation reinforce the same issue. The system is learning from a shrinking pool of genuine insight, while being fed an expanding volume of recycled material.
Alongside this, access to tools is being mistaken for capability.
There is a growing tendency for individuals to position themselves as strategists because they can produce outputs that resemble strategic thinking. But strategy is not the output. It is the process that leads to it.
A strategist is not defined by their ability to generate answers quickly. They are defined by their ability to sit with ambiguity, challenge assumptions, understand trade offs and arrive at a position that is both clear and defensible. That requires original thought and it requires judgement. It requires context.
What is emerging instead is a reliance on generated responses repeated across formats. The same language appears across websites, social posts, emails and proposals. The phrasing is familiar. The ideas are familiar. Even the so called insights are familiar.
It creates the impression of consistency. What it actually reveals is sameness.
This is not an isolated observation. Content that should differentiate businesses is instead making them indistinguishable. The intention is to move faster. The outcome is often a loss of any sense of what a business actually stands for.
We have said for years that social media changed the way we communicate. But what is happening now feels more fundamental.
It is not just communication that is changing. It is the habit of thinking itself.
In the same way we lost the art of communication, we are now losing the art of original thinking. And I would argue we are losing it faster.
The more people rely on AI to generate responses, the less they practise the process of forming their own. Like any skill, if it is not used, it weakens. I came across a graduate seeking marketing help for a university project 18 months ago. On the side panel of the screen was a question she had typed into an AI tool. Her question was, how do I ask my aunt to arrange a meeting with her?
That moment has stayed with me. And if this is where we are now, what does that mean for the generations coming behind us?
This piece did not begin in research. It began in conversation.
A small number of peers in my network, people I hugely respect and whose judgement I trust had been noticing similar things I had. Something felt different. The outputs were increasing. The thinking behind them was not. I wrote my observations down. Then, in what felt like an appropriate test of the argument, I put them through an AI tool to sense check what I had written.
What came back was well organised and comprehensive. And it kept missing the point.
Every time the system tried to smooth the edges, reframe the argument or improve the tone, I had to bring it back. This is my original thinking, I kept saying. I do not need you to rewrite it. I need you to help me hold it.
That exchange, more than anything else, confirmed what I had been observing in others. AI is a powerful tool when it is working in service of original thought. The moment it becomes the source of that thought, something is lost. And the more it is used that way, the harder it becomes to notice.
If the volume of original thought declines and the volume of generated content increases, the overall quality of what is available to train future systems could shift. The technical ceiling and the human ceiling begin to meet. Or have they already converged? It is in everyone’s pockets now.
For business leaders, this has practical consequences.
Efficiency is right to pursue. Tools that accelerate output are right to use. But speed without depth does not create advantage. When everyone uses the same systems in the same way, without applying their own thinking, the result is a market where everything appears competent and very little stands out.
A simple test. Ask three questions of any business. What do you do, why does it matter, and how are you different. If the answers are polished but interchangeable, the issue is not execution. It is the absence of a clear, original position. These are among the first questions I ask at the start of every marketing strategy session. The answers, increasingly, sound the same and they are unable to differentiate their offerings from their competitors.
The point is not to reject AI. It is to use it with intent.
And I say that having tested the argument myself. I wrote this thinking down. I pushed it through an AI tool. It gave me well formed options, neater sentences, a tidier argument.
But it was not my argument anymore. I kept pulling it back. Because the point of using the tool was not to let it think for me. It was to help me say what I already knew more clearly.
That is the distinction.
These tools are powerful when they extend thinking, not when they replace it. They can help articulate ideas. They cannot decide what your business stands for. That responsibility still sits with the individual.
Because the moment we stop contributing original thought, we are not only weakening our own decision making. We are shaping the future of the systems we rely on.
And that is the part that is not being discussed enough.
I am genuinely concerned, not about the technology. About what people are doing with it. They are taking whatever AI gives them as gospel. They are not questioning it. They are not sitting with it. They are not asking whether it is actually true for them, their business, their situation.
The thinking has to start from the bottom up. You have to form your own argument first. AI might help you find the right word or sense check a fact. But it cannot tell you what you think. It cannot replace the process of working something out for yourself. That still has to happen inside your own head.
And if it does not, you are not using a tool. You are outsourcing your judgement. And that is a very different thing.
The ceiling is not just in the model. It is in us.
