Silicon Songs: when robots get sing-songy
The fusion of music and artificial intelligence is transforming our most beloved melodies. From synthetic voices filling stadiums to songs created by algorithms, this sonic journey explores the future of musical art.
Technology and musical creativity are intertwining like never before in an era where artificial intelligence is beginning to push the boundaries of human art. In this episode, we explore how algorithms can not only suggest songs, but also compose, perform and bring to life voices that, although virtual, have the power to fill entire auditoriums. Join us on a journey that starts with the earliest experiments with AI in music and moves on to the latest innovations in automated music generation, touching on such fascinating topics as the legality and ethics of these digital creations.
Historical overview of generative AI in music
We begin with a look back. Did you know that the first attempts to bring together music and technology date from the 1950s? Here we take a tour through the pioneers of musical AI, from computer-generated scores by rudimentary computers to the birth of Vocaloid in 2000. We discover how this tool, which began with the simple aim of synthesising voices, has evolved into a benchmark for AI-generated music, with virtual artists such as Hatsune Miku, who has managed to fill stadiums and become a global phenomenon.
How a music-generation AI works and the latest advances
We go to the heart of the episode by unpacking how AI creates music today. We review the most popular tools that allow anyone to generate original songs from the comfort of their browser. We explain in simple terms how portals such as Suno and Udio work, helping listeners understand what the AI is doing behind the scenes. In addition, we explore the parameters that can be customised and the current creative limits of these platforms. Without going into too much technical detail, this section is an invitation to musical experimentation in the digital world.
The field of AI-generated music is advancing rapidly, and this episode highlights the latest innovations. It looks at how the new version of models such as Suno has expanded song-generation capabilities, improving musical structure and allowing longer, more complex compositions.
The episode also looks at how it is now possible to start from a simple hummed melody or musical fragment so that the AI can continue and transform that idea into a complete piece in any musical genre. The magic of AI makes it possible to create everything from techno anthems to powerful metal riffs starting from a well-known Manchegan theme.
Legal aspects and copyright
The rise of AI in music is not free from controversy, and here we enter complex territory: copyright. While some platforms offer users the possibility of owning the songs generated, others face accusations of using copyrighted material to train their models. This section provides an introduction to the main current legal debates, setting out the risks and challenges faced by both AI creators and human artists.