Google Building an AI Bot that can Create "Original" Music

Listen to Podcast:

It shouldn’t come as much of a surprise that artificial intelligence would eventually make its way into the world of music given the extent to which it has already permeated every other aspect of our digital lives, from the creation of original artwork to the writing of essays to conversations with therapists.

It should not come as a surprise though that the tech giant Google would be the first significant player to enter the arena.

Read more: How to Learn Music Production?

Users would be able to type in increasingly specific prompts noting genres and styles or even build songs based on a hummed or whistled melody, according to reports. The company is reportedly building an AI bot that is capable of creating “original” music(Opens in a new window) from both text and sound prompts. Within the company, the upcoming software is referred to as MusicLM.

The details were disclosed in a research article that was published on January 26, and it described MusicLM as a “model generating high-fidelity music from text descriptions.” The publication also stated that MusicLM “generates music at 24 kHz that remains consistent over several minutes.” The article discusses that carefully written captions can be used to make songs and provides several examples:

The primary score of an arcade game’s soundtrack. It has a catchy electric guitar riff, and it moves quickly and enthusiastically. The music has a lot of repetition, making it simple to remember, but it also has some unexpected sounds, such as cymbal smashes and drum rolls.

Along with a library of sounds and other artificial intelligence prompts derived from sources such as art archives, additional sequences of timed text prompts are used to help build the structure of songs (Opens in a new window).

Several examples of the artificial intelligence-created songs have already been published on Google’s Github account (Opens in a new window) as a part of a preliminary release of a dataset containing 5,500 music-text pair examples called MusicCaps.

Read Also: Best Music Streaming Services

It is inevitable that the introduction of such a platform will spark additional conversations regarding the role of artificial intelligence in intellectual property theft and copyright infringement. These conversations will be generated by a multitude of artists and art repositories who have not consented to the public use of their art in the creation of AI bots like these; in the meantime, others are capitalizing on the surge in AI-fronted technology. Additional breakthroughs in AI bring with them a new set of dangers for the people who are responsible for creating the technology, as exploited labor forces bear the burden of data mining and moderation.

The Google AI music maker is not going to be published any time soon, according to the business, which cited persistent concerns about cultural programming biases, errors, and concerns about plagiarism as issues that need to be overcome before the product can be made available to the public.