Since the topic of Studio Ghibli-style images and AI-generated art is everywhere , I think it’s important to discuss it. I’d love to hear the opinions of the Hivians
In my opinion, we’re at a point in time similar to the late 1990s when the rise of peer-to-peer (P2P) networks made illegal music and movie distribution incredibly easy. This destroyed established business models, causing massive financial losses for distribution companies and artists alike. Similar copyright issues are emerging with AI today, but the similarities end there.
With AI, we’re not talking about a revolution in distribution but in creation, which is not illegal. For example, I can paint a piece in Picasso’s style and sell it as my own copy—art styles are not patented. However, to do that, I would need to invest time in learning how to paint and then create the artwork. The problem with AI is that it bypasses this entire process, creating productive structures in art that surpass human limitations. This allows someone with no technical skills to generate images in a matter of seconds.
This doesn’t mean that an AI user disrespects art or artists—it simply means they’ve found an easy way to do something they couldn’t before, allowing them to express themselves artistically without investing years of effort or possessing natural talent.
On the other hand, artists have spent years training and perfecting their craft, and it’s understandable that they feel this new technology is taking away their livelihood.
Due to the unique nature of this technology, we need to find a real solution. Artists should not be forced out of work (because if they can’t make a living, this will happen within 10–20 years), but at the same time, it’s unrealistic to ban a useful technology (especially since AI allows people without formal training to express themselves artistically).
My view is that we need stricter transparency, explainability, and regulation in AI models. We should know which artworks were used to generate an image or train a model. That way, artists can choose to opt out or be compensated for their work.
We cannot ignore the fact that the foundation of all these models consists of data that took humanity centuries to produce. If artists aren’t paid fairly, they’ll stop creating, and we’ll face a cultural and artistic decline.
Before anyone argues that we can just train AI models on data generated by AI itself (without fresh human data)—
That’s not possible. This phenomenon is called Model Collapse, which refers to the gradual loss of diversity and originality in an AI model’s outputs when it’s trained primarily or exclusively on data generated by other AI models.
Simply put, if AI keeps learning from AI-generated content, its responses become more repetitive, less creative, and poorer compared to the original human-generated data it was first trained on. This creates a vicious cycle where AI produces increasingly uniform content, losing the ability to create anything truly new or innovative.
P.S I asked Chatgpt to generate me an image based on my post.