Artificial intelligence is no longer limited to just words and images; it is now learning to compose songs as well. This was reported by Zamin.uz.
OpenAI is working on a new platform for music creation. As part of this project, OpenAI is collaborating with the Juilliard Institute, one of the world's prestigious art institutions.
The goal is to teach artificial intelligence to deeply understand music. Experts are working on annotating scores to bring the system closer to thinking like real composers.
The new system will be capable of composing music from scratch and adding new sounds, instruments, and effects to existing songs. For example, it will be possible to automatically add orchestral elements to a vocal track or transform a simple melody into a full composition.
According to experts, this project will take OpenAI to a new level in the musical artificial intelligence market. Currently, projects like Suno and Udio are leading, but OpenAI's capabilities are expected to significantly change the competition.
Technical details have not yet been disclosed. It is still unknown whether the new service will be integrated into products like ChatGPT or Sora, or offered as a separate platform.
OpenAI representatives have not commented on this. It is worth noting that the company has previously been active in this field: in 2019, it introduced MuseNet, a system for creating instrumental music, and in 2020, the Jukebox system, which creates full songs with words and vocals.
