Stability AI's audio generator can now compose 3 minutes of music

Stability AI has launched Stable Audio 2.0, a higher version of the music-generation platform, which allows musicians to create songs that last up to three minutes. The new version is more extensive than its previous edition which limited the duration to 90 seconds. The company has made it available on its website for free so users can test it without hesitation.

Stability AI’s audio generator levels up and now allows up to 3-minute-long audio files

The new system operates mainly by text prompts though users may also submit audio clips for analysis and imitation. However, all the audio files uploaded in this program must be copyright-free, making it impossible to generate copies of existing pieces of music. Ultimately, this instrument serves as an outlet for creative minds who want to produce original compositions or arrange short clips into long ones.

While the extension of song length is an important improvement, there are inherent restrictions in AI-generated music. For instance, despite an initial fascination with machine-created singing voices often become uncanny verging on unknown language, and barely any melodic coherence at best.

These AI vocals were described by The Verge as “soulless and weird”, similar to whale sounds which contributed towards a rather eerie listening experience. Furthermore, Stability AI’s software sometimes glitches like typical AI-produced content that can cause things to suddenly disappear or duplicate.

Stable Audio 2.0 still needs to iron out some flaws before reaching human-level production

Despite these technical idiosyncrasies, the capability of rapid processing by the platform makes it ideal for quick albeit unconventional music productions. Critics point out that AI-generated music loses the sentimental connection and emotional depths found in human music artworks. Music is a way of expressing thoughts and feelings by its originators which is lacking in AI compositions.

Nevertheless, Stability AI’s Stable Audio 2.0 provides room for creative exploration, particularly in sub-genres such as birthday videos or background music for menial tasks. While it may not replace human musicianship, it offers a glimpse into the potential of AI-driven music generation technology and its role in shaping composition and creativity in the future.