In a recent interview, Meta CEO Mark Zuckerberg shared his views on copyright issues during AI training. He pointed out that many creators and publishers often overestimate the actual value of their works.
Zuckerberg stated that Meta will collaborate with some significant content creators, but if certain creators demand payment, the company would rather choose not to use their content.
Image source: Image generated by AI, authorized service provider Midjourney
Zuckerberg mentioned, "I believe that most creators or publishers tend to overestimate the value of their specific content in the larger scheme of things."
He emphasized that if creators oppose Meta using their works, Meta will also choose not to use them, as this will not have a significant impact on the overall development of AI.
Currently, Meta and many major AI companies are undergoing lawsuits regarding the unauthorized scraping of data for AI training. Last year, the company was sued by a group of authors for training the Llama model, who accused it of using unauthorized works. This case is currently not looking optimistic for the authors, and the judge has also expressed dissatisfaction with their legal team.
Zuckerberg also discussed broader topics related to copyright. He believes that with technological advancements, the legal boundaries of fair use need to be re-examined. He stated, "When you put something out into the world, to what extent can you still control and own it?" This question is particularly important in the age of AI.
Although legal debates continue, many AI companies have already signed paid partnership agreements with major media outlets. Zuckerberg mentioned that Meta's future content strategy will be similar to its approach to news content, emphasizing that the company will only pay for valuable content. He said, "We simply won't pay for content that isn't important to people."
In this process, some artists attempt to protect their works from being used for AI training through unofficial tools, but they often face difficulties due to the terms of use of social media. Zuckerberg concluded that Meta's AI products are still in their early stages, and no one has completely figured out what users need, so most creators probably shouldn't expect to profit from them.
Key Points:
🌟 Zuckerberg believes that creators and publishers overestimate the value of their works in AI training.
🤖 Meta will collaborate with some significant creators but won't pay for content without value.
⚖️ The legal boundaries of copyright and fair use need to be re-discussed and examined in the age of AI.