What else can we believe?

A competition, a dictionary, a YouTube clip — and the same mechanism everywhere. Anyone who hears something often will accept it as true. That applies to people. And, as it turned out this week, also for the AI ​​chatbot I asked about it. A thought exercise about language, facts, and why “everyone says it” is never a source.

The reason: Formula 1 and words that didn’t belong there

I was a Formula 1 fan from a young age. I remember seeing the first race with my boyfriend Erik in Harderwijk in 1977. I’m dying of Ayrton Senna saw it live in 1994 — one of those moments that stays with you, even if you were still young. I’ve been watching ever since. But lately I’ve been noticing things in the comments that, to my ears, shouldn’t be there.

Take the driver who leads the most in a season. In the past you simply said: “he is most in charge”. Nowadays you hear “he is in the lead most often”. It sounds almost the same, and new viewers — who have been joining since Max Verstappen — won’t notice. But for those who have followed the sport for some time, it sounds like a word that was only coined for something that already had a name.

Or the start of a race. In the past you started first. Now you hear commentator say that someone has the race “has begun”. What I once learned as a Germanism — a corruption from German (anfangen) that you don’t use in Dutch — has suddenly become a normal commentator word choice. Not once. Soon more often.

So I did what you do these days: I asked a random AI chatbot. Whether that was “normal” language. The answer, from different models, was remarkably unambiguous: yes, this is just normal language development, language changes.

And that got me thinking — not about the word itself, but about the answer. About how something becomes “true”. And about the consequences if you no longer recognize that mechanism.

Aside about language: the dictionary that follows the street

(By the way: an aside, but one that matters.) There is a year that can be mentioned – I leave the exact date to your own research – when “people” started to listen to what the street says, and then include it in the dictionary of the Dutch language. The idea sounds modern and popular: language belongs to the people, so the dictionary follows the people.

But pay attention to the mechanism. It’s the same as the commentator. You do not adhere to the agreement that was made – the meaning that the word had, the rule on which the language was built – but you normalize what the masses now use. Even if that was once a fabrication, a sloppiness, a Germanism that has stuck. Not because it’s right, but because it has been said often enough.

This means that the change is no longer a fact, but a consequence of repetition. And the circle closes itself: once it is in the dictionary, it is “official”, and no one asks whether it was ever correct. The AI ​​chatbot that answers “yes, normal language” does nothing but repeat the new reality — because that reality has now become the average.

Who has known this for a century: Edward Bernays

When you first recognize this mechanism, you wonder how long it has been working. The answer is: more than a century. The man who first systematized it is called Edward Bernays — cousin of Freud, and the inventor of what we know now public relations call it, but what he himself is honest propaganda until the word became loaded. His insight in one sentence: you don’t have to force people — you just have to present the same thing to them repeatedly until it becomes reality.

Bernays literally wrote that the task of the propagandist is to “order public opinion” so that people make the desired choice on their own. No coercion, just guidance. And the most powerful guidance is not the argument, but the repetition.

Those who want to delve deeper into this themselves: the investigative journalist James Corbett of The Corbett Report has covered Bernays and the history of PR/propaganda in depth and with context. Worth your own investigation:

• 🌐 corbettreport.com — look for Edward Bernays / propaganda for its episodes and source documents.

An excerpt showing the principle: The Trueman Show #287

How misconceptions arise — through misinformation and propaganda, by misleading people — That we can have differing opinions is aptly illustrated in an excerpt from The Trueman Show, with Kees van der Pijl as guest. The fragment is not about language, but about how a small group with a common agenda can turn a hobby into a global influence.

It concerns delivery #287 — “The 1984 9/11 Plan & Mossad in The Hague”. I point it out as an example of the mechanism, not as a statement about the content. Judge for yourself — that’s the whole point of this blog.

The excerpt below starts at 53:03, the question “what is the influence of Freemasonry, and who and why are involved?”. Rob Elens also comes by and explains how he helped the corona committee exposed; until approx 55:45 it explains how “the far too rich of this earth” (Bill Gates is mentioned by name) have turned a personal hobby into a global influence. If you are interested in this topic, listen until at least 1:05:35

What I recognize in it

Whether you share the explanation or not, the principle is recognizable: a small group with a common goal, repeating that goal over a long period of time, and adjusting the definition of what is “normal” along the way — until what was once a special statement has become the starting point. That is precisely the mechanism of the commentator who introduces “most often”, and of the dictionary that then approves it. Difference in scale, not in kind.

---

The digression: how an AI chatbot actually works just like a human

And now the step that connects everything. Why did those chatbots answer “yes, normal language”? Not because they know language rules — but because they do exactly what we do.

If we hear something often, we start to see it as true. Edward Bernays built up an entire field there. The NLP techniques that came on top of that say the same thing: repetition anchors something in your brain. The textbook example from pop culture: Barbara Papa in the Barbapapa stories, which repeats everything three times — so that it sticks. Innocent, and therefore clear: repetition is how something is repeated into reality.

A chatbot essentially works the same way. A language model is a huge combination of “what is most common”. Ask it if “most often in the lead” is normal, and it searches its body of text for the frequency — and whatever occurs more often as an answer becomes the answer. The chatbot takes something for granted if it is more common than the alternative. Exactly the mechanism that Bernays described in humans, only now without anyone consciously controlling it.

This means the chatbot is not an oracle, but a mirror. It reflects the average of everything we’ve written and said together — including the repetitions, including the fabrications that have become “normal.” Anyone who wants a fact will get one from a model frequency back. Anyone who wants the truth must continue to check for themselves.

“Everyone says it” has never been a source.
Not in humans. Not with a dictionary. And certainly not with an AI chatbot.

What we can still believe — and how

The answer is not “nothing”. It is: nothing on the authority of repetition alone. A few habits that help:

1. Ask where it comes from. A word, a statement, a “fact” — who said it, since when, and what was its original meaning?
2. Note the frequency drop. If the only argument is “that’s just how it’s said,” then there is no argument — there is repetition.
3. Use a chatbot as a mirror, not as a source. The answer reflects the average. That’s helpful — it’s not the truth.
4. Read the original thinkers yourself. Bernays, Corbett, Kees van der Pijl — listen, read, and judge for yourself. Copying repetition is exactly what they warn about.

The commenter says “most often”. The dictionary follows. The chatbot nods. And somewhere there is someone who remembers that things were once different, and that it was no small matter. That is the place where “what we can still believe” begins: not with what is often said, but with who continues to see the difference.

This blog belongs to a larger theme. We build and manage our own infrastructure — without dependence on big tech, and without blindly trusting what has become “the standard”. Mail, hosting, payments, password vaults: in your own hands.
Please contact us.