Question: Is there an AI-powered solution that can identify the most interesting part of a song, like the chorus?

If you want an AI-based way to find the most interesting part of a song, the chorus, for example, SONOTELLER is a good choice. The tool analyzes song attributes like genres and moods and the presence of elements like a chorus, and it can generate detailed descriptions and even autotag music libraries. It's in beta testing, so be prepared for some glitches, but it's a good example of how AI can help you understand music better.

Another interesting tool is Cyanite, which offers auto-tagging and auto-description to extract metadata from audio files. It also offers similarity search to find songs that match a reference track. That could help you find new music and figure out what makes a song tick, including its chorus.

For more general audio processing, AudioShake offers AI-based audio separation that can split audio into its component elements. That can be useful for isolating the chorus or other parts of a song, and it offers mixing and mastering tools as well as audio analysis.

Last, Staccato is an AI-based platform for musicians and songwriters to create and analyze music. It includes an AI Instrument and AI Lyrics Generator that could be useful for generating parts of lyrics in different styles and genres. It integrates with popular digital audio workstations and notation software, so it's a useful tool for songwriting and music production.