YouTube is releasing a series of AI music experiments, introducing novel ways for users to compose music using the voices of prominent artists and convert hummed melodies into musical compositions.
YouTube released a blog post detailing a series of new AI-powered visions. This comes shortly after the company released a package of measures to clamp down on forms of AI misuse on the platform, including some forms of deep fake content.
One of these exciting new features, known as “Dream Track,” is already accessible to select creators and designed for integration into YouTube Shorts.
With Dream Track, users can swiftly generate songs fashioned in the distinctive styles of officially licensed artists, such as Alec Benjamin, Charlie Puth, Charli XCX, Demi Lovato, John Legend, Papoose, Sia, T-Pain, and Troye Sivan.
Users have the liberty to specify the song’s mood and themes, with AI crafting music tailored to their preferences and fitting it to their Shorts posts. AI is finding its way into social media content creation workflows, with Meta also announcing new generative AI tools for Instagram.
As YouTube describes, “Dream Track” allows users to “quickly produce songs in people’s style.”
Another compelling tool in the YouTube AI music arsenal is aptly named “Music AI Tools,” designed to support musicians in their creative processes. This reportedly emerged from YouTube’s Music AI Incubator, an assembly of artists, songwriters, and producers helping the company experiment with musical projects.
YouTube notes that participants in this program are “intensely curious about AI tools that could push the limits of what they thought possible” and actively seek ways to enhance their creative processes.
The ultimate goal is to facilitate the transformation of ideas into music with minimal production.
YouTube envisions a world where musicians can effortlessly translate their thoughts and ideas into musical compositions. A utopia for some, dystopia for others.
The tech behind these tools is primarily rooted in Google Deepmind’s Lyria system, tailored explicitly for music creation. Lyria is capable of generating lengthy, coherent audio sequences. Deepmind is developing another melodic AI generator, SynthID, to complement Lyria.
SynthID also introduces an inaudible audio watermark into the music, undetectable to human ears but recognizable by AI tools, thereby confirming that the songs have been auto-generated. As Deepmind asserts, “This novel method is unlike anything that exists today, especially in the context of audio.”
YouTube is spurring on the trend of integrating AI into music. Elsewhere in the industry, AI musicians are even earning multi-million dollar record deals.
As YouTube takes steps to embrace AI innovation while safeguarding the integrity of music and artistic expression, the music industry is bracing itself with mixed feelings.