In recent years, artificial intelligence technology has developed rapidly, which has also brought many challenges, among which copyright issues are particularly prominent. The training of large language models relies on massive amounts of data, which inevitably involves copyright ownership issues. Recently, OpenAI and Microsoft have been sued over copyright issues, which has attracted widespread attention in the industry. At the same time, researchers are also actively exploring solutions, such as developing "forgetting curse" technology to try to delete specific training data from the model to avoid infringement disputes.
AI companies are facing pressure from copyright disputes. OpenAI and Microsoft were sued for infringement by the New York Times. Researchers are trying to develop "forgetting spells" that delete specific training data from models. They used texts including Harry Potter to test the Oblivion Charm. The rich language and plot of "Harry Potter" are also welcomed by AI researchers and appear in some theories of AI concept schools. "Harry Potter" may help the future development of AI models.
Solving the copyright issue of AI models requires the joint efforts of technological innovation and laws and regulations. The research and development of "Forgetting Curse" technology, as well as the standardization of the source and use of AI training data, will be important directions for future AI development. The application of literary works such as "Harry Potter" in AI research may provide new ideas and inspiration for solving these problems.