Musician Jordi Pons releases track made with AI sound design, not AI songwriting
jordiponsdotme · x · 2026-09-22
Music-tech researcher Jordi Pons released a new track with a clear split: he handles melody, harmony, arrangement, guitar, mixing — everything creative — while AI does only sound design. He uses Stable Audio to generate and transform samples and guitars into uncanny textures, and Moises for percussive layers. He argues AI art's narrative should include underground artists using AI to make sounds that weren't possible before, not just automate existing aesthetics.
More from Multimodal
- John Coogan declares "video is solved", teasing a video generation breakthrough — johncoogan · 2026-09-22
- Creative demo: your daily voice note becomes an AI-painted canvas with a matching melody — iamrobotbear · 2026-09-22
- Runway experiments with responsive generative video interfaces — DanScalco · 2026-09-22
- Redditor shows Qwen's image model can generate famous IP characters — Z3ROCOOL22 · 2026-09-22
- Right-Click Anything: SAM 3.1, Muse Spark and Muse Image combined into one clever demo — nikhilaravi · 2026-09-22
- A Video Prompt That Turns Gravity Upside Down: "The Other Floor" Full Prompt — umesh_ai · 2026-09-22