Our perspective on Tim Green's article "No Consent, No Credit, No Pay"
What Is Actually Happening
The public debate around generative AI and artists' rights is focused on legal details — datasets, lawsuits, licensing models. But behind all this noise, the main point escapes notice: we are witnessing not a technological improvement of existing tools, but the elimination of the very need for intermediaries between an idea and its realisation.
The chain used to look like this: idea → specialist → tool → result. Now it looks like this: idea → result. AI did not replace the designer with a better Photoshop. It made the designer an unnecessary link in the chain. And this is permanent.
Filters and plugins never cancelled the tools themselves. 3ds Max, After Effects, Photoshop — they remained necessary, which meant the people who knew how to use them remained necessary too. AI became a thin client that replaced both the tools and the specialists in one move: designers, layout artists, retouchers, illustrators, pattern makers. You no longer need them — and you no longer need what they used either.
The Analogy Everyone Misses
The authors of such articles and the participants in court proceedings compare what is happening to piracy or copyright infringement on the internet. That is an imprecise analogy.
The precise one is Gutenberg's printing press. It did not give scribes a better writing tool. It made the scribe an unnecessary link in the distribution of text. The profession did not disappear overnight — new niches emerged, new forms of craft. But the economic foundation was undermined irreversibly.
Even closer is the history of photography. The arrival of smartphones with quality cameras did not destroy photographers in direct competition. It simply turned out that the vast majority of consumers were satisfied with the level of photography available on their phones. Professionals survived in niches: artistic reportage, film photography, studio work. Enthusiasts of film still exist — just as enthusiasts of valve amplifiers do. But the monopoly on quality imagery collapsed forever.
AI is doing exactly the same thing to illustration and design.
What Really Worries Artists — and What They Are Confusing
The article lists grievances: lack of consent, attribution, compensation, style copying. Let us examine each honestly.
Style copying is painful, but it is not new. On DeviantArt, an interesting new style gets copied the very next day — without any AI involved. Style has never been a subject of legal protection in any jurisdiction. AI has merely accelerated and scaled what was already happening.
Legal lawsuits are not built on style but on fact: specific images were physically present in the LAION-5B dataset, downloaded and used for commercial purposes without the authors' permission. This is closer to a real violation — but even here the boundary is blurred. If you train a model on photographs of interiors where paintings hang on walls, purchased by the homeowners — who is the infringer? The law does not yet have an answer.
The demand for attribution from AI seems strange when we never demanded it from human artists inspired by each other's work. This is not an argument — it is a symptom of disorientation.
Compensation is the only genuinely strong argument. Companies earned billions not from specific images but from models for whose creation those images were indispensable. The Spotify analogy works here: the platform pays authors not because it reproduces a specific track, but because it uses the entire catalogue as the foundation of its business. The logic of Getty Images and Sweden's STIM is exactly this — and it is convincing.
But even winning every lawsuit and receiving royalties will not bring back the corporate illustration market. It is gone — just as photographers lost the family portrait market.
Where the Real Protection Lies
Practice shows that those who survive are not those who litigate, but those who keep moving.
Unique technique is real protection. An authorial system built on strict geometric curves, non-trivial mathematical foundations, rare combinations — this is opaque to AI. It averages what appeared frequently in the dataset. What is rare and original it cannot reproduce — it struggles even to describe it.
Fusion — combining two or more styles in non-trivial combinations — produces works that AI cannot decompose into base layers if that particular combination was absent from its training. Papercut art plus liquid art: AI sees the result but cannot see the structure.
However, there is a crucial practical nuance here. "Idea → result" is not yet a straight arrow. AI in its current mass form does not know your visual language. It averages. It drifts. It draws things you did not ask for. Ask it for a shadow from a fountain — and it will try to draw the fountain too. This is not a flaw that will be patched in the next update. It is a fundamental property of a model trained on averaged mass data: it does not understand local logic, a part without the whole.
This means the gap between intention and realisation is still real. And as long as that gap exists, craft does not disappear — it simply changes its form.
Selling process, not just result — the most sustainable model. An artist who sells not only paintings but brush profiles, techniques, and tools sells something AI does not produce. AI uses brushes but does not sell them.
Live contact with the audience creates attachment to the author as a person, not to the genre. This is what economists call switching cost — viewers may go to AI for an image, but they come back to the artist for the human being.
The Deficit That Was Always There
The deepest problem is neither technical nor legal. An idea is never protected by copyright anywhere — only its specific realisation. DeviantArt demonstrates this daily: a new idea survives one day before the first imitator appears. AI merely accelerates an already existing process.
The real deficit — of original ideas — existed long before generative AI. Where did the calligraphers go? They still exist, but the bulk of calligraphy is now produced by electronic and software means. This is not a tragedy — it is a civilisational shift.
What Comes Next — and Why It Cannot Be Stopped Either
There is one more dimension that changes the entire picture.
Right now AI is a mass tool with averaged models. This is like a typewriter: everyone gets the same font. But the moment is approaching when an artist will be able to train a model on materials they have personally selected — including their own paintings, their own visual language, their own logic of form. And this cannot be stopped either.
A personally trained model closes the loop. It knows your visual language. It understands your local logic. It realises your specific intention rather than an averaged one. The gap between intention and realisation — the shadow without the fountain — becomes your problem to solve, not a limitation imposed by someone else's model.
This is no longer "a thin client replaced the specialist." This is the specialist acquiring a personal thin client. A qualitatively different situation. And it destroys the linear picture of "AI displacing the artist." The real trajectory is more complex: mass AI displaces the mass market, but personal AI amplifies the unique author.
The Real Scale of What Is Happening
The authors of articles like this one and the participants in court proceedings are discussing compensation for the past. That is understandable and humanly just. But the future is already structured differently.
The real question is not legal. It is civilisational: what to do with the mass release of creative professions — just as the industrial revolution released manual labour, and digital photography released portrait photographers.
History gives no grounds for panic — every such shift generated new niches, new forms of mastery, new markets. But it gives no grounds for illusion either. What is gone is gone forever.
The artists filing lawsuits are fighting for compensation for the past. Those who will thrive are the ones who understand that the instrument has changed, build a personal relationship with their audience, develop techniques that cannot be averaged, and — when the moment comes — train their own model on their own material.
That is not the end of craft. That is craft in its new form.
And we have to live with that.
Top comments (0)