Vokalikes

Vokalikes is a term used in digital music and AI communities to describe vocal-like audio content generated by machines that imitates human singing or speaking. In practice, vokalikes can range from fully synthesized singing voices to processed samples and voice-conversion outputs. The concept emphasizes likeness to natural vocals rather than the identity of a specific performer.

The term vokalikes is informal and circulates mainly in online discussions and experimental music contexts. It is not a formal label defined by a single academic field, and its exact scope can vary depending on who is using it and which technologies are involved.

Production methods commonly associated with vokalikes include neural vocoders, transformer-based text-to-speech systems with singing capabilities, and voice-conversion techniques that modify a target singer’s voice to produce new material. These approaches often rely on large datasets of vocal recordings, raising considerations about licensing and consent from the original performers whose voices or performances are used as training material.

Ethical and legal considerations are central to discussions of vokalikes. Issues include authorship and ownership of generated material, rights to training data, and the potential for misrepresentation or deepfakes. Platforms and communities may implement disclosure requirements or moderation to address these concerns.

Vokalikes appear in experimental music, sound design, and demo production, offering new avenues for vocal textures while prompting ongoing debates about authenticity and copyright. Notable examples include AI singing models and projects such as OpenAI’s Jukebox, which illustrate both the creative potential and the challenges of synthetic vocals. See also AI-generated music, voice cloning, neural vocoders.