A new AI music synthesis service called Udio, created by ex-DeepMind employees, has entered the scene, allowing users to generate high-fidelity music from written prompts, including lyrics. Similar to Suno, Udio offers a range of genres like country, classical, and hip hop. Despite its impressive capabilities, some musicians fear these AI tools could threaten their livelihoods, echoing concerns raised in a recent protest letter signed by over 200 artists.
While Udio’s AI-generated songs demonstrate technical prowess, there are limitations in the quality compared to human-created music. The service relies on a five-step workflow involving human input to enhance the AI-generated compositions. Users can input prompts to create custom tracks, and Udio provides options for extending or remixing songs. The platform generates two distinct song snippets for users to choose from, allowing for customization and sharing on various platforms.
Udio’s model utilizes a large language model like ChatGPT to generate lyrics and a music synthesis method undisclosed by the company, likely similar to diffusion models seen in other AI music tools. The platform emphasizes user rights over the generated music, permitting commercial use. To prevent copyright infringement, Udio employs measures to block tracks resembling specific artists’ work, ensuring originality in the AI-generated music.
Despite advancements in AI music generation, some individuals express concerns about the automation of art, questioning the need to replicate human creativity. AI’s ability to mimic art forms, although not always perfect, showcases the potential for generative AI to create notable and intriguing compositions. As AI technologies expand into music creation following advancements in image and text generation, the emergence of AI-generated music signals a new frontier in creative AI applications, sparking debates about the impact on human artistry and expression.