Spettacolo
SongPad uses GPT-4 to predict the winner of the Eurovision song contest
SongPad is an AI-powered songwriting and reviews platform. The SongPad team used OpenAI's GTP-4 with custom designed prompts and parameters to review each of the songs in the Eurovision song contest, and rank them based on use of rhymes, creative writing techniques, metaphors and emotional impact.
The results suggest that the best song is Germany's entry 'Blood and Glitter' performed by Lord of the Lost.
Based on the AI tests, SongPad predicts the UK entry 'I Wrote A Song' performed by Holly Mae Muller to finish around position 5 in the Eurovision grand final.
Here's what the SongPad review service reported on this song:
The AI suggests that Serbia's Luke Black performing 'Samo Mi Se Spava' ('I just want to sleep') is the song with the lowest quality lyrics, and would therefore be the team most likely to score nil points.
SongPad is primarily a tool for writing and reviewing song lyrics. Powered by GPT-4, it is capable of reviewing emotional impact and use of metaphor, rhyme and creative writing techniques, and structure and repetition.
It is trained on English language content, so the official English translations of all the Eurovision songs were used for review within the SongPad platform.
SongPad reviews are based on lyrics, and the platform doesn't predict what is likely to be popular. It also doesn't take the politics of the Eurovision song contest into account.
SongPad Limited was founded in 2023.
Tim Cole , co founder and CTO, is a web and app developer with 20 years of experience in the tech industry. He has been working with AI for songwriting for over 5 years and is a published songwriter.
You can learn more here: www.songpad.co
View original content:https://www.prnewswire.co.uk/news-releases/songpad-uses-gpt-4-to-predict-the-winner-of-the-eurovision-song-contest-301822459.html