The Grok Row Is Real, but It Is Not the Main Event
There is a row in the UK right now about Grok and the creation of sexualised images of juveniles and non-consensual individuals. Let’s deal with that properly, then move on.
There is no scenario in which sexualised images of juveniles should be created by any AI system. None. That should be a hard default, built in at system level, non-negotiable. You could reasonably extend similar restrictions to most AI use involving juveniles, with narrow exceptions for things like family photo restoration or legitimate advertising with consent.
That point is settled. It is not controversial, and it is not where the real argument lives.
The mistake being made is treating this as if it represents the core problem with AI imagery. It does not. It is the most emotionally charged edge case, which makes it politically convenient, but analytically it distracts us from the much larger shift AI is creating.
That shift is not a moral one.
Instead, the shift is the economics around AI images (and all AI) and how this changes society.
Why I Am Not Lecturing Anyone About Morality
I spent 20 years as a police officer. Part of that work involved running informants, many of whom were connected to the sex industry. That gives you exposure to realities most people only encounter through headlines and social media outrage. Exploitation exists. Power imbalances exist. People make decisions under pressure, not ideology.
I am not naïve about this world, and I am not interested in pretending it suddenly appeared because of AI.
Adult content, sex work, and exploitation predate the internet, let alone generative models. Platforms change. Laws change. Human behaviour does not reset every time Parliament passes a bill or a company updates its terms of service.
That is why morality is not the lens I am using here.
Not because it does not matter, it definately does, but because it does not explain what is actually happening.
AI does not introduce a new moral universe. It allows people to supercharge their existing behaviours, world views and desires by removing the barrier of creation.
AI imagery makes anyone the master of illusion and radically alters our world in the process.
Once that is clear, we can stop arguing about what should exist and start talking about the economic change, how that pans out and what you should do.
AI Imagery Is Not the Real Problem
I use AI imagery daily. I create images constantly. I will soon be creating AI video as well. The idea that you can create a short film, even a 20-minute one, using AI is not frightening to me. It is creatively compelling.
The question worth asking is not whether AI images are real or fake, human-made or machine-made. Those distinctions are collapsing anyway.
The question that matters is whether any of this can generate durable value.
To answer that, we need to talk about scarcity.
The Scarcity Inversion AI Creates
For most of modern media history, creation was scarce. Distribution was relatively abundant. If you could write, film, design, or illustrate well enough, you had an advantage. Making the thing was the hard part.
AI flips this completely.
Creation is no longer scarce. It is effectively free. Images, videos, characters, even entire worlds can be generated at negligible cost. Replication is instant. Iteration is endless.
When creation becomes abundant, reach becomes scarce. Attention becomes scarce. Fame becomes scarce.
This is the inversion most people fail to grasp.
It’s not the ability to create that is the problem now, it’s the ability for anything we create to travel online to reach people so you can benefit in some way.
We need to get comfortable with 2 core factors.
That invented things can make money (and already have).
Reach is how they make money.
We Already Know Invented Things Can Make Money
There is nothing inherently strange about making money from things that are not real.
Tom Cruise is not Ethan Hunt. Mission Impossible has generated billions. Tron is not real. Avatar is largely computer-generated. The Marvel Universe is entirely fictional and produces extraordinary revenue.
These are invented worlds populated by invented people, and nobody struggles to understand their value.
So the idea that AI-generated people, images, and worlds could make money should not shock anyone. It should be embraced (as ling as it is legal).
The problem is not that they are artificial.
The problem is that everyone can now make them.
When Creation Is Free, Almost Nobody Gets Seen
You can create an incredible AI-generated film and upload it to YouTube. It can sit there indefinitely with a handful of views. Not because it is bad, but because nobody sees it.
Algorithms do not reward effort. They reward distribution, momentum, and external signals. Quality is table stakes.
This is where most AI optimism collapses. Lowering the barrier to creation does not democratise success. It increases competition and concentrates attention even further.
Anyone who thinks AI imagery is a shortcut to easy money has never spoken to someone trying to earn a living on Instagram or OnlyFans. The hustle is relentless. The competition is brutal. The failure rate is enormous.
AI does not soften this reality.
Instead, the visual world becomes flooded. Any natural moat of beauty, talent, creativity, and skill becomes crossable in an instant.
A 60 year old overweight man could create a social media profile of a stunning 20 year old blonde model.
A person with no photographic talent can create a social media page of stunning wildlife photos, monetise them through printed goods and earn a living without even owning a camera or going outdoors.
But the fact that this is possible actually limits it’s ability to make money in the long run.
Replication Makes Control Impossible
AI adds another brutal constraint. Anything you create can be copied, tweaked, and redeployed almost instantly. Originality becomes fragile when replication is free.
You might spend days crafting an AI character, only for someone else to clone it and flood the same platforms within hours. Control at the point of creation is an illusion.
This is why attempts to regulate or restrict AI imagery at the output level will always struggle economically. For every platform that bans something, another will appear. But even on unrestricted platforms, the fantasy that AI imagery represents a reliable revenue stream collapses quickly.
When supply explodes, value collapses.
Which brings us to the real dividing line.
Single Assets Die. Brands Accumulate Meaning.
One AI image does nothing. One AI video does nothing. One AI influencer post does nothing. These are flat assets in an environment drowning in assets.
What compounds is depth.
Value is built through accumulation, consistency, and narrative. Through characters that persist. Worlds that expand. Stories that overlap and evolve.
This is not new.
It is the model that Marvel was based on. It was not one comic. It was thousands. Characters returned. Universes deepened. Audiences invested over time.
AI will follow the same logic.
The winners will not be people who generate clever images. They will be people who build ecosystems. AI influencers with history, voice, and worldview. AI worlds that grow over years, not posts.
This is brand thinking, not creator thinking.
Why Powerful Messages Matter More Than Ever
This is where Rory Sutherland’s thinking becomes critical.
Sutherland argues that powerful messages contain elements of illogicality, extravagance, costliness, inefficiency, disproportion, or risk. Not because rationality is bad, but because rational behaviour conveys no meaning. Meaning is conveyed by the things we do that are not in our short-term self-interest.
In a world where AI makes everything cheap, rational, and easy, power comes from signals that are deliberately none of those things.
Look at MrBeast. His videos are not efficient. They are wildly extravagant. Giving away millions of dollars. Rebuilding entire towns. Running absurd challenges that make no rational sense if your goal is short-term profit.
That is precisely why they work.
The cost is the message. The inefficiency is the signal. The illogicality creates meaning. You cannot replicate that with AI alone, because the power does not come from the content. It comes from the commitment behind it.
The same logic applies everywhere online.
The creators and brands that break through are not the most technically competent. They are the ones willing to do things that look irrational on a spreadsheet. Spend too much. Commit too long. Build worlds when everyone else posts clips.
AI imagery makes this even more important.
When anyone can generate a beautiful image, beauty stops signalling effort. What signals effort is scale, persistence, and depth over time.
Why Platforms Are Strangling Reach
This also explains what is happening across platforms. Reach is being strangled.
Images no longer buy attention. Effort is no longer rewarded proportionally.
The barrier to entry has collapsed, so platforms respond by raising the barrier to reach. This will happen everywhere.
Creation used to be the hard part. Now it is trivial. Distribution is the tax.
As such you will see an economic switch to messages that are powerful
Long form content wins not because it is fashionable, but because it is expensive in effort.
Anyone can post 100 times. Far fewer can post 1,000 times.
Anyone can generate one image. Very few will build a coherent universe around a character for years.
This is how you win online.
Frequency, Effort, and Costly Signals
This is where frequency meets effort. Seth Godin understood this long before AI. He writes every day. He ships relentlessly. Not because each piece is extraordinary, but because accumulation creates gravity.
We are entering an era of fleeting attention. Everything is skippable. Everyone is replaceable. The only durable advantage is sustained depth combined with costly signals.
AI lowers the cost of creation. It does not lower the cost of meaning.
For anyone to truly make money through AI, and in a world of strangled reach, you must build something that is based on powerful messaging.
Illogicality.
Extravagance.
Costliness (money or effort)
Inefficiency.
Disproportion.
Risk.
If you look around and see what’s breaking through the noise online, you’ll see powerful messaging at the heart of that breakthrough.
AI imagery and video, means that to reach anyone, you’ll have to dial up the power supply to the messaging you use.
The Conclusion People Keep Avoiding
AI imagery will not make you rich. Not as a trick. Not as a shortcut. Not as a one-off creation.
What it will do is expose where value has always lived.
In reach. In fame. In accumulated meaning. In brands and worlds that people return to when everything else blurs together.
If you think like a creator, AI will overwhelm you. If you think like a brand builder, AI becomes leverage.
That is the argument. And that is why it is more comfortable to argue about the images instead.
AI imagery is not the problem. It is coming in whether we like it or not. It is going to happen whether politicians like it or not.
AI makes people more of who they are (good and bad) and allows them to do things that they were previously unable to.
And when the cost of doing things reduces to next to zero. Reward comes from powerful messages.
The issue is that most people don’t grasp this.
Every day somebody is creating new software. Every day somebody is trying to create a new AI image or video. This is happening right now in front of us.
I could start and create a software business today. The issue isn’t that I can’t create any longer. The issue is how to reach and how you acquire customers.
Like it or not, everything comes back to brand, to fame, to reach.
You must build for the long term.
Because that’s where the real value will happen in the future.
AI isn’t the bad guy in our new world story.
But it changes our economics. Once you understand this, you will start playing the game of powerful messages.
Thanks for reading.
Andrew Holland
Originally published on LinkedIn on 15 January 2026.