Why Trust Has Become the Last Thing AI Can Fake

Norbert Elias spent much of his career explaining something that most people experience but never name: that the rules governing how we behave in social space are never really about manners – they are always about power.

His masterwork, The Civilising Process, traces the history of European court society through an unlikely lens: the fork. In medieval Europe, people ate with their hands, shared bowls, and belched freely at the table. Then, gradually, new rules proliferated. Use a fork, use a separate fork for fish, do not reach across the place setting, do not eat too fast, do not eat too slow, and NEVER let your hunger show.

The ruling classes introduced these rules not because they made dinner more enjoyable. They introduced them because each new rule created a new threshold, a new line between those who knew and those who did not. When the lower classes eventually caught up, mastered the fish fork, and learned the code, the code changed. Another rule, another layer, yet another piece of cutlery.

The civilising process, Elias argued, was the mechanism by which distinction reproduces itself. It was never about etiquette – it was always about exclusion.

The Rooms You Had to Be In

For most of the twentieth century, professional power operated by the same logic, just with different cutlery.

To build influence in business, you had to know people. Not know of them, actually know them. You had to be in the right rooms: the conference dinners, the private members’ clubs, the off-programme conversations at Davos or at whatever sector-specific facsimile of Davos your industry ran. You had to be able to afford the flight, the hotel, the conference ticket, and the time. You had to have been vouched for by someone who was already inside.

Access to visibility was, structurally, access to capital (social, financial, and reputational) all at once. The barrier was not talent. It was not even expertise. It was proximity, and proximity was expensive, slow, and deeply unequal.

This was not a flaw in the system; it was the system. The scarcity of access was the gatekeeping mechanism.

The Death of Friction

Then, over roughly a decade, something structurally unprecedented happened. The barrier to visibility collapsed almost entirely.

You no longer need to fly anywhere. You do not need a publisher, a publicist, a PR agency, or a warm introduction. You do not need an institution behind you or a legacy credential in front of you. You need a smartphone, a decent idea, and the willingness to show up repeatedly in a small rectangle of light.

LinkedIn, TikTok, Substack, and YouTube, have not just democratised distribution. They democratised the appearance of authority. Anyone can speak with confidence, anyone can post with conviction, and anyone can build a following.

And now, with generative AI, you do not even need to be a person. AI avatars of real experts can be deployed at scale. Synthetic influencers who do not exist are doing the selling. People are building an audience with content they did not write, in a voice that is not theirs, performing expertise they have not earned. The friction that once made visibility scarce has not just been reduced. In some corners of the digital economy, it has been eliminated entirely.

When Everyone Is Visible, What Does Visibility Prove?

This is where Elias becomes relevant again. In his framework, the civilising threshold is never fixed. When the code becomes learnable, when enough people master the fork, the code changes. The ruling class does not lose its power; it relocates its markers. It finds new ways to signal distinction that the newcomer cannot easily replicate.

What we are living through right now is precisely this moment of relocation. The old markers of authority, conference stage, book deal, institutional affiliation, the right job title, were already being devalued by social media’s democratisation of visibility. And AI has accelerated that devaluation to near-total collapse.

If anyone can be visible, visibility proves nothing. If anyone can generate polished content, polish proves nothing. If a synthetic avatar can deliver a keynote-ready script in thirty seconds, the script proves nothing. The question is: what comes next? What is the new cutlery?

Patina

There is a concept in the antiques trade that I keep returning to when I think about this problem: patina.

Patina is the quality that accumulates on an object over time, through real contact with the world. It is the lustre on old silver, the wear on a well-used tool, the darkening of oak over centuries. It is, crucially, not achievable through any shortcut. You cannot fake it, you cannot manufacture it. You can produce something that resembles it, but then it just becomes a little almost disney like, and ofcourse, every serious collector can tell the difference immediately. It does not hold up to scrutiny.

Trust, in this moment, is patina.

Not the performance of trust, and not the signals of trust – the testimonials, the follower counts, the professional headshots, the carefully engineered content calendars. Those are the equivalent of artificial ageing: chemical treatments applied to make new wood look old. They work, briefly, on the uninitiated. But they do not survive scrutiny.

Real trust is what accumulates when someone has shown up consistently over time, changed their mind publicly when the evidence demanded it, been wrong and said so, been right and let the record speak, and navigated real complexity with real accountability. It cannot be generated by a large language model because it is not made of words. It is made of track record, of relationship, of presence that persisted when it would have been easier to disappear.

The New Gatekeeping

Here is the argument, stated plainly: we are at the moment where the civilising code relocates. As AI makes surface-level visibility infinitely scalable, the thing that becomes genuinely scarce, and therefore genuinely powerful, is the evidence of a mind that has been in the room. Not performed ‘been in the room’ but real-world experience from those who have actually been there. Those who have taken a position that cost them something. They have built relationships over years, not through funnels, and they have produced ideas that predate the content calendar.

I fully realise that this is not a comfortable argument, because it suggests that some of the barriers that collapsed were real gatekeeping mechanisms that excluded deserving people, and some of those mechanisms will be reconstructed. Elias was not naive about this. The civilising process is not progress towards fairness. It is the history of power finding new addresses.

But there is something here worth holding onto. The new address that power is moving to, authentic expertise, real relationship, demonstrated track record, intellectual honesty, is, at least, harder to fake than an invite list or a first-class ticket. It is earned, not purchased. It accumulates through doing the work, not through performing it.

What This Means For You

The practitioners who will survive the AI visibility explosion are not the ones who produce the most content. They are not even necessarily the ones with the largest audiences. They are the ones whose audiences, however small, know them to be real.

Real means: they have said difficult things and stood by them. They have admitted the limits of their knowledge. They have disagreed with people who had more power than them, and explained why. They have published ideas before those ideas were fashionable and been there to say I told you so – not triumphantly, but as a matter of record.

That is not a content strategy; it is a professional life. And it is, I would argue, exactly what AI cannot replicate – not because AI lacks sophistication, but because what is being authenticated is not the quality of the output. It is the quality of the person.

The fork has changed again. And this time, hopefully, it is sorting for is something that actually matters.