Today, we conclude this series, exploring the use of AI to imitate the unique styles and techniques of artists.
In general, it is difficult to copyright a specific style of art. It rarely happens, and if it did, we’d miss out on a lot of exciting creations, and it would be difficult to participate in any genre of art without worrying about infringing on someone else’s rights. Imagine if the animation style anime were copyrighted and artists were no longer able to create their own unique versions. We might not have the modern anime we all know and love – Studio Ghibli films, Naruto, or even Pokémon!
Now, AI can remix styles with a simple text prompt. Anyone can use a tool like Midjourney or Stable Diffusion to create a work of art that looks like a Picasso or Van Gogh original. This, unsurprisingly, has many artists on edge, worrying whether their own work will remain relevant if a machine can easily reproduce it.
The line between plagiarism and inspiration is a blurry one. Even when new or unique, style can’t be copyrighted. However, the nature of AI writing prompts means users often specifically name the artists whose style they want to imitate. How much of that work can the user really claim as their own? Read more about our thoughts on this topic here.
Beyond copyright issues, AI will likely impact the sharing of creative content and the open community in general. The legal uncertainty around AI, the lack of transparency of AI algorithms, issues around artist recognition and payment for use of their copyright works as AI training input, the potential flood of AI-generated works, and the questions around authorship of AI outputs, all together constitute yet more challenges for better sharing. Creative Commons wants to focus on working with our community to find a sustainable, coordinated, and inclusive approach that fosters the Future of Open.
This is the last in our series on AI – the possibilities, limitations, and challenges. If you’d like to learn more about this topic, please check out the full blog series – and let us know what you think!
29/05/23, baixacultura, Por Leonardo Foletto*
É quase impossível identificar retroativamente as fontes e autorias retiradas sem autorização de bases de dados pelos bots. Apropriação do saber comum pode estar mais ameaçada – abrindo uma disputa sobre os direitos do autor
O ano de 2023 tem sido permeado pelo tema de Inteligência Artificial (IA) quando se fala sobre tecnologias digitais e internet. Isso se deve em grande parte ao sucesso estrondoso do ChatGPT, uma IA generativa desenvolvida pela OpenAI, uma empresa dos Estados Unidos fundada em 2015 com um investimento inicial de U$ 1 bilhão, cerca de R$ 4,9 bilhões na cotação atual.
Continuar lendo “Como o ChatGPT está influenciando na disputa pelo controle do conhecimento”
comunica 