What is Google's MusicLM?

The Advent of text-to-music

Hello Everyone,

I got to admit I’m pretty excited about Google’s imminent contributions to Generative A.I. and A.I. products we can actually use in our everyday life. Google I/O 2023 conference, either in May or June, 2023 is going to be very big this year as it responds to ChatGPT, GitHub Copilot and many other things.

As such I was very intrigued by the news of Google’s MusicLM. While ChatGPT was released in an unsafe state, Google is more responsible and decided against making the AI trained on a dataset of 280,000 hours of music publicly available due to ethical concerns. That’s something I can respect.

This was announced on January 26th, 2023, and by that I don’t mean a full release. Listen to the examples:

Google announced a new generative AI model called MusicLM that can create 24 KHz musical audio from text descriptions, such as "a calming violin melody backed by a distorted guitar riff." It can also transform a hummed melody into a different musical style and output music for several minutes.

• Google researchers have made an AI that can generate minutes-long musical pieces from text prompts, and can even transform a whistled or hummed melody into other instruments.