An Australian tune written using artificial intelligence has claimed first prize in a new international song competition.
The song, Beautiful the World, was one of 13 entries in the inaugural AI Song Contest, which was hosted by Dutch radio station VPRO.
The competition invited Eurovision member countries to submit a three-minute song in the Eurovision style made partly or wholly by artificial intelligence.
Beautiful the World was composed by a team of computer scientists, producers and musicians lead by Sydney production studio Uncanny Valley in collaboration with RMIT University’s Dr Sandra Uitdenbogerd and the University of NSW.
Uitdenbogerd described the song’s sound as “high entropy Daft Punk” and said its creation took on new relevance as the challenging events of 2020 – including catastrophic bushfires and a global pandemic – unfolded.
“When we started this process we were still reeling from the bushfires,” Uitdenbogerd said.
“So we used our AI processes to turn koala grunts, kookaburra laughs, and Tassie Devil barks into a kind of instrument, reflecting our concern for all the wildlife dying in the bushfires.
“As the year progressed, coronavirus hit and so we all had to stay at home – many of the lyrics speak to this experience, such as the ‘welcome home’ refrain in the pre-chorus, and the line ‘flying from this world that has gone apart’.
“So you can treat it as a bit of fun, but lines such as ‘the music of the earth has arrived’ do also give it that serious message of: ‘look after each other, and look after the planet’.”
All these lyrics – along with the melody – were created using a combination of AI processes and human intervention.
Producers from Uncanny Valley (which boasts production credits on hits from Darren Hayes and Sia) used algorithms to turn samples of 200 Eurovision songs – including perhaps the most famous, ABBA’s 1974 hit Waterloo – into new melodies.
Lyrics were generated by feeding “seed” words into a computer, which then spat out a series of phrases that were selected by the producers.
Uitdenbogerd’s role was to provide an algorithmic “pattern matching” technique, which can match words and melodies together – a process she refined in her PhD.
The winner was decided by a mix of popular vote and a panel of AI experts, including Vincent Koops (from the Netherlands), Anna Huang (US) and Ed Newton-Rex (UK).
OMG we did it! Team Australia won the inaugural Eurovision ! Thank you everyone for your votes and support.
— Sandra Bogerd (@SandraBogerd)
In a written statement, the judges commented on the high quality of the entries.
“Composing a song with AI is hard because you have all the creative challenges that come with song writing, but you also have to juggle with getting the machine learning right.
“Overall, we were delighted by the diversity and collaboration within the teams.
“They not only pushed the boundaries of their personal creativity, but also gave the audience a look into the exciting future of human-AI musical collaboration.”
Contest founder and VPRO editor Karen van Dijk said the teams had made the most out of the creative possibilities of artificial intelligence.
“In my opinion, some songs would not be out of place in the official Eurovision Song Contest,” she said.
While the winners won’t take home any huge cash prize, they are basking in the cheeky glory of claiming the first Australian Eurovision (AI) victory.
Uitdenbogerd, who was up at 4.30am Australian time to watch the livestream announcement of the winner from Holland, said that after a good nap, she was starting to think about what was next.
“It’s been such a great experience collaborating with the Uncanny Valley team, and it’s so satisfying to win an international song contest using my two loves: music, and computers,” she said.
“I’m also really pleased at the awareness this competition has raised about the relationship between AI and creativity, especially when it comes to songwriting.
“There is some talk of composing Australia’s Eurovision song for next year, but we’ll have to wait and see!”