Raw LLM Responses

Your viewpoint on using images for research and then transforming them into commercial use provides food for thought. However, it's not uncommon for research to eventually find its way into commercial use. And on the issue of copyright, I think there are instances where it goes a step too far (Disney springs to mind). Data scraping, especially without an opt-out or opt-in, doesn't always sit right with me. However, from the early days of the internet, we've had the saying, "If it's on the internet, it's there forever." This implies that if you're not comfortable with your work being seen or used, it might be best not to publish it online. One strategy artists could employ is always keeping a high-resolution copy of their work secured offline as a safety measure. This can help during any disputes and can make it harder for an AI to accurately imitate your style. Copyright does accommodate derivative works, but they must be distinctly different to be deemed independent from the first. If an AI recreates your work after being trained on your material, it could potentially fall under the original copyright, although this remains a legal gray area. Looking towards the future, I suggest a few strategies: - Firstly, keep the highest resolution copy of your digital art secured offline as a safety measure. This can help during any disputes and limit the imitation potential for an AI. - Secondly, if your artwork appears somewhere without your permission, consider submitting DMCA takedowns and copyright notices. For artwork to bypass your copyright, it needs to be distinctly different. - Thirdly, be aware that if you place your work on the internet, it could potentially be misused. The permanence of the internet means your work could be utilized in ways you're not comfortable with, and your ability to rectify this may be limited. - Fourthly, secure your work by placing it behind a login, ideally with an entity that is dedicated to protecting your interests. This can make public scraping and other unauthorized use less likely. - Lastly, advocate for developers or others to create detection systems. Such systems should be able to link works or styles with the original artist, although this may be a monumental undertaking. This would enable those who want to use AI for generative works to reach out to the original artist for permission, and potentially provide a portion of any commercial gains to the artist who (willingly or not) provided their training data. - Remember, if you don't protect your intellectual property now, it may be much harder to defend it in court later. AI scraping can seem unfair, and there might be instances where it is. However, some of this issue lies in educating artists on how to protect their works. AI is here to stay and will certainly continue to improve. But as of now, it can't create wholly original works, where original refers to things involving human creativity. I'd go so far as to encourage artists to use AI to enhance their current works. Those who are opposed to how some AI tools have been trained will need to seek out more ethically grounded alternatives. As the saying goes, "If you fight the flood, you'll be swept away by it." While I haven't personally created or published commercial art online, I understand the fear that these tools might replace you. But I encourage you to try them out. You might find that it takes as much skill to get what you want from these tools as it does to create an original piece, although your results may vary. Learning to use them now could give you a huge leg up in the future. Written by a human, proofed by AI, adjusted by human, adjusted by AI, final copy edited and proofed by a human! Also Definitely not Legal advice, do your own research!