Triggered by the popularity of ChatGPT, Google showed up with it’s own AI-based technology that generates music with text descriptions.
Named MusicLM, this AI-based music generator from Google can compose songs in various genres and depths. It’s even reported to make melodies based on humming, whistling, and even text inputs. Though it has it’s own share of problems, it’s definitely a new-gen tool to aid musicians.
Composing Music With Text Inputs
ChatGPT has it’s technology praised by many around the world. After all, it’s able to do anything a man does in minutes, effortlessly. Though there are some catches, such AI-based creative technologies are seen as the next big thing coming up. This led many other companies to come up with their models – the recent one being Google.
As per a research published to GitHub, Google unveiled a new AI-based music generator called MusicLM – that can compose interesting songs with just text input. You can ask for any genre or depth; MusicLM would do it in no time, note researchers.
After all, Google said it’s trained on a vast dataset containing 280,000 hours of music – which would help MusicLM to mix and produce new music from your inputs. Though this isn’t the first-ever technology of such, Google says MusicLM has more scope than it’s previous AudioML and OpenAI’s Jukebox.
You can ask for instrumental music, a combination of genres, and even write tracks based on abstract concepts that can’t be done by regular computers. What’s more interesting is that MusicLM can even craft melodies based on humming, whistling, or the description of a painting and a Story Mode can stitch several of such descriptions to produce a DJ set or soundtrack.
Well, talking about hurdles, Google’s MusicLM said to produce some compositions at times, and vocals tend to be incomprehensible. Also, a few performances are repetitive in ways human works might not. Thus, it still has some room for improvement.
Other Trending News:- News