Google AI can create music in any genre from text description


Never mind ChatGPT – Music could be the next big frontier for AI content creation. Google recently published research on MusicLM, a system that creates music in any genre with text descriptions. This is not the first AI music generator. As techcrunch Note, projects such as Google’s AudioML and OpenAI’s Jukebox have addressed this topic. However, MusicLM’s model and huge training database (280,000 hours of music) help it create music with surprising variety and depth. You may like the output.

AI can not only combine genres and instruments, but also compose tracks using abstract concepts that are generally difficult for computers to understand. If you want a hybrid of dance music and reggaeton with “spacey, ethereal” melodies that create a “sense of wonder and awe”, MusicLM can make it happen. The technology can also generate melodies based on hums, whistles or details of a painting. A story mode can stitch together multiple details to create a DJ set or soundtrack.

MusicLM has its problems, as with many AI generators. Some compositions sound strange, and the vocals are incomprehensible. And while the performances themselves are better than you’d expect, they’re repetitive in ways that human action can’t. Don’t expect an EDM-style drop or the verse-chorus-verse pattern of a typical song.

Just don’t plan on using the technique any time soon. As with other Google AI generators, the researchers are not releasing MusicLM to the public due to copyright concerns. At the time of publication approximately one percent of the music produced was copied directly from the training songs. While questions regarding the licensing of AI music have not been resolved, a 2021 whitepaper by Eric Sunray (now working for the Music Publishers Association) suggested that there are enough “consistent” traces of the original sounds that AI music reproduction rights may be violated. You may need to get approval to release AI-generated songs, such as musicians who rely on samples.

AI already has a place in music. Artists such as Holly Herndon and Arca have used the algorithm to create album and muse soundtracks. However, they are either allies (as with Herndon) or intentionally unpredictable (as with Arca’s). MusicLM may not be ready for prime time, but it hints at a future where AI could play a bigger role in the studio.


All products recommended by Engadget are selected by our editorial team, independent of their parent company. Some of our stories include affiliate links. If you buy something through one of these links, we may earn an affiliate commission. All prices are correct at the time of publication.

Source


Related News

Boot Camp updates to make Studio Display compatible with Windows but not 100% compatible

Apple has "put the batteries" to be able to make the Studio Display, presented on March 8, compatible with computers with a Windows operating system. This

Improve your Wi-Fi connection with these super offers

To have a good Wi-Fi connection at home, you can count on different devices to improve coverage. We have more and more devices connected to the network and

Cinema offer: 2X1 on Blu-ray and DVD to celebrate the arrival of the 2022 Oscar Awards

Next Sunday, March 27, a new 2022 Oscar Awards gala will be held, where Pen

Is it the end of Windows Defender? Discover this new free antivirus

There are many security companies that propose their projects to us in order to protect our equipment. Some of these antivirus are free, while for others we

Interview with Fabrizio Mazzotta, voice of Krusty the Clown

Return the appointment on the Twitch channel of Pop Culture with Voicesthe show aired on the first and third Fridays of the month it talks about dubbing and

HUAWEI P50 Pro Launched in Turkey

Huawei has launched the HUAWEI P50 Pro, the newest and long-awaited smartphone from the P Series family, which will delight every fan of mobile photography