The AI Mirror

What Machine-Created Content Reveals About Us

I have been spending a goodly amount of time in recently working on these series of articles discussing AI and its relationship to content production and entrepreneurial activity. These two domains of creativity in particular, shape our culture and our societies in many important ways. I have pointed out that AI is a tool–a very powerful one, but it is not a muse gifting us ideas from who knows where. It is a prediction engine, not a soul.

But what if AI isn’t just a machine that serves to mimic us? Not only does it copy us in our thinking and our procedures but it also serves as a mirror. It reflects the messy, beautiful, and awfulness of what we say and do.

AI doesn’t dream up its outputs from the ether, cosmic void, or somewhere mysterious state of being that we can't comprehend. It processes our digital interactions—every tweet, blog, meme, and forgotten forum post we have ever made. It then regurgitates this information as just one big, super-sized remix. The machine isn’t creating—it’s curating our collective mind, and what it reveals isn’t always flattering.

Take the blandness of AI writing and content creation and its tendency to churn out safe, perfectly rendered dross; text that only a committee could love, images full of novelty but strangely unmoving except to the most ardently sentimental.

It is all our fault. If 90% of the internet is rubbish (and I’m being generous), then AI’s sterile output is holding up a mirror to our own mediocrity. Oh no, not that, not me!

Then there is also the issue of bias. As I have written many times elsewhere, I have a deep concern about the relatively tiny cadre of programmers and executives embedding their worldviews into AI’s DNA—Western, tech-bro, ivory-tower biases with values that don't necessarily reflect how the rest of us think and behave. I have been harsh on the gate-keepers in the past. I still think their actions are cowardly and a dreadful reluctance to defend truth and honesty in the face of nonsense because as a class they are more afraid of lawsuits than they care about providing real benefits to the consumer. They may be the most powerful people in the world at the moment but their purported care for others has a sad and pathetic feel to it.

But it is more than just corporate cowardice—it is our obsession with policing narratives. The constant desire we have to dominate others with our ideas and actions. AI doesn’t judge; it just shows us what we are at.

But while it is hard to forgive them, they are only machine-minders in the end–we should take to heart an old maxim from the early days of computing. Garbage in, garbage out. It was true then and still holds true now. Our present day digital factories have the very same problem, but on much more massive scale. Rubbish inputs from the internet and the world wide web produce rubbish outputs. And most of the input is rubbish.

What can we do about it? Is prioritising accuracy and truth something worth attempting considering the enormity of the problem? How can one person make a difference?

For a start don't upload anything that, as far as you can determine, is dishonest, untruthful or misleading. It might not make an immediate dent but you will be in a better place than almost everyone else.

Stay away from garbage stuff. AI is just a huge prediction machine and despite the very sophisticated algorithms in play, it is essentially weighted to assuming that content that gets more attention is more important, and returns the result accordingly. Don’t give the bad stuff any attention. You are just encouraging the growth of more of this distasteful, bland mess. And worse, the obfuscation of the truth.