In context: Artists and content material creators proceed to change into extra entwined with synthetic intelligence. Whereas some view AI as an infringement on the work of others, some creators see it as a brand new medium to current their concepts. YouTube creator and a cappella singer Peter Hollens clearly sees it as a possibility. The social media star not too long ago launched a music video that was 100% created utilizing AI applied sciences.
What occurs once you give inventive license to an AI and ask it to generate a music video? Apparently, nothing fairly as dangerous as we have come to anticipate.
YouTube celeb and a cappella singer Peter Hollens not too long ago proved that it may be one thing extra promising than a pc merely presenting customers with undesirable romantic advances or alerting them to their demise. Hollens not too long ago posted a number of tweets giving followers a glimpse right into a 100% AI created music video to accompany his tune I Can’t Make You Love Me.
What occurs once you make a 100% AI music video utilizing?
ChatGPT for the shotlist
Midjourney for the scenes
RunwayMl for set extension and fixing some property
kaiberAi for the animated vid / utilizing the Midjourney exports
D-iD ai for the singing half
The Future is gonna be superior pic.twitter.com/2TMk6hCwug— Peter Hollens (@PeterHollens) February 22, 2023
In line with Hollens, a number of AI instruments had been used collectively to create the music video. ChatGPT was used to generate a shotlist that described what visuals could be utilized in every scene of the video. Photographs had been generated from textual content utilizing Midjourney, exported, and later animated utilizing kaiberAi.
Hollens’ likeness within the video, seen singing in a number of sequences, was additionally created by AI.
D-ID, an AI device used to create animated avatars, photographs, and movies, permits customers to create something from fantasy-based avatars to speaking portraits generated from a single picture. Extra modifying and revisions had been accomplished utilizing AI-based instruments from RunwayMI.
It is essential to notice that the singing, lyrics, and different audio within the video usually are not AI-generated, a minimum of not this time.
Vocal credit score goes solely to Hollens, a effectively established a cappella singer who has been creating YouTube content material for greater than a decade. With greater than 2.8 million subscribers, Hollens isn’t any stranger to creating extremely produced and effectively acquired audio and video content material.
Replace PT2: What occurs once you make a 100% AI music
‘ ChatGPT for the shotlist,
🎥 Midjourney for the scenes,
🎨 RunwayMl for set extension and asset fixing,
🎞️ kaiberAi for the animated video utilizing Midjourney exports, and
🎤 D-iD ai for the singing half.Need a tutorial? pic.twitter.com/OPpsipUrpx
— Peter Hollens (@PeterHollens) February 23, 2023
That is to not say that AI is not able to creating its personal music. OpenAI’s Jukebox is a neural community able to producing music and singing throughout any variety of genres and kinds. No musical expertise, manufacturing expertise, or vocal coaching is required. Choose a style, choose an artist to make use of as a base sample, and feed in some lyrics. Jukebox will take it from there, and the outcomes aren’t at all times as dangerous as you’d anticipate.
The outcomes will solely get higher because the expertise matures, finally catching as much as actuality TV in proving that typically you do not want expertise to be well-known.