OpenAI is reportedly making a significant return to the generative music space. The artificial intelligence company is developing a new tool capable of creating music from user-provided text and audio prompts. This initiative marks a deeper move into creative media for the firm behind ChatGPT and Sora. The new model aims to compete in the growing AI music market.
This new AI tool is designed to generate a variety of audio content. Expected use cases include adding original soundtracks to existing videos. Furthermore, the tool could generate instrumental accompaniment, such as guitar tracks, to complement pre-recorded vocals. This functionality is crucial for content creators and video editors.
It remains unclear when OpenAI plans to launch the tool. The company has not yet confirmed if it will be a standalone product. Alternatively, it could be integrated into existing platforms like ChatGPT or the video generation application, Sora.
In a strategic move to ensure musical quality and authenticity, OpenAI is collaborating with the Juilliard School. Students from the renowned music conservatory are reportedly assisting the project. Their role involves annotating musical scores to provide high-quality training data for the new AI model.
This approach suggests OpenAI is prioritizing structural musical knowledge. By incorporating annotated data, the company aims to improve the coherence and musicality of the generated output. This strategy also helps mitigate legal risks related to copyrighted material. However, the legal and ethical debates surrounding generative music remain a prominent industry challenge. The AI music space already includes major competitors. Companies like Google with MusicLM and startups such as Suno and Udio are also actively operating in this domain. OpenAI’s entry will significantly intensify the competition.
 ORIGIN: Techcrunch
