In the 1950s, it was multitrack recording. In the 1960s, the synthesiser. Then came MIDI and hard drive recording. In the late 1990s, auto-tuning came to the rescue of many a struggling singer and created a vocal genre all its own.

The early years of the 2020’s will likely go down as another period in which everything changed: the moment when Generative AI went mainstream.

Even in its infancy, Generative AI is already exerting a profound influence on the way we make, record and experience music.

Limelight readers, it’s time to get up to speed.

Image © Dung Nguyen/Pixabay

What is Generative AI?

Generative AI is a type of Artificial Intelligence that can generate new content (or data) based on huge datasets – text, image and music libraries, for example – it has been ‘trained’ on.

Once it has absorbed and analysed the patterns and structures of those datasets, neural networks (Generative Adversarial Networks – known as GANs) mimic how the human brain works: a ‘generator’ creates new content, and a ‘discriminator’ evaluates it, based on datasets of existing examples. Via this process, the software can create increasingly ‘original’ outputs.

AI is not...