Title	MELISSA REED
Brand	THE NORWEGIAN RESEARCH COUNCIL
Product/Service	THE RESEARCH SEMINARS 2021
Category	B10. Use of Technology
Entrant	ANTI Oslo, NORWAY
Idea Creation	ANTI Oslo, NORWAY
Media Placement	ANTI, NORWAY
Production	ANTI Oslo, NORWAY
Production 2	ELLEVILL PRODUKSJON Oslo, NORWAY
Post Production	ELLEVILL PRODUKSJON Oslo, NORWAY
Additional Company	University of Oslo, NORWAY

Name	Company	Position
Gerd Stangeland	ANTI	Creative
Nina Bugge	ANTI	Creative
Torstein D Drogseth	ANTI	Account director
Jennie Andersson	ANTI	Account Manager
Benedicte Bølling	ANTI	Producer
Kristoffer Eidsnes	ANTI	Graphic design/ Motion graphics
Lasse Mejlvang	ANTI	Advisor
Linn Engen	ANTI	Producer
Markus Odland	Ellevill Produkson	Director
Fredrik Hinsh	Ellevill Produkson	Editing/ Film Photographer
Elvina Vladi Kvisle	Ellevill Produkson	Set designer
Sebastian Bergsnov	Ellevill Produkson	Film photography/ Editing
Daniel Vaz Berrum	Chr. Quart Lighting	Lighting design
Alexander Refsum Jensenius	University of Oslo	Development of artificial intelligence
Lars Monstad	University of Oslo	Development of artificial intelligence
Stefano Fasciani	University of Oslo	Development of artificial intelligence
Adriane Loinsworth	ARY	Artist

Why is this work relevant for Media?

Spoiler alert! To get the full experience of this idea, we recommend watching the case film before reading this…. This creative idea is based on a musical experiment, where the technology and the music provided is both the core of the idea and the execution. This is an experimental idea that may change the way we think about music. “Can artificial intelligence replace something as human as creativity?” Was the question we asked ourselves for the annual research festival “Forskningsdagene ''. Artificial intelligence is one of our times biggest breakthroughs, so complex that we don't yet know the

Background

The research festival “Forskningsdagene” is a national, annual festival under the auspices of the research council of Norway. Since the start in 1995 their goal has been to make young adults enthusiastic about research, and hopefully contribute to the recruitment of research-related professions. The topic of this year's festival was «Peace & Conflict», and right now we are on the brink of a conflict that may affect the lives of all young adults - the conflict of man vs. machine. How many jobs will be affected? When will your job be replaced? And can Artificial Intelligence even replace something so human as creativity? Which means young adults' favorite artist ARY may be at risk of disappearing, making this experiment suddenly very relevant to them and their future playlist.

Describe the creative idea / insights (30% of vote)

We made an experiment to see if Artificial Intelligence could replace one of our favorite artists. In this creative campaign AI is the core of the idea and end result. We examined how far the development of AI has come, could it possibly make a better song than ARY? Together with a team of researchers, we trained an AI for a month on ARYs musical catalog. We made it compose melodies and lyrics with the signature/DNA of ARY. When finished, ARY put her voice on and released it for her fans. Waiting to see if they would notice anything wrong. Few did. Then we released the experiment as a whole film on youtube, snapchat and instagram. And people couldn't believe their ears Despite the inherent barriers we managed to break through showing young adults that research actually can be relevant and exciting for them.

Describe the strategy (20% of vote)

How do you make young people not run away when you mention the word “research”? For this year's research festival we wanted to attract young adults, and to do so we had to make them care. That's why we combined research with something that is important to them - music - and turned one of their favorite artists ARY into an experiment.

Describe the execution (20% of vote)

We teamed up with researchers from the University of Oslo and gave them a month to train an AI to make a song. They used “machine learning” to train the AI based on ARY’s music catalogue. When finished we made ARY put her voice, and afterwards invited fans for the experiment in the form of a listening session. We documented the stunt alongside a talk where ARY and the researchers discuss the topic of AI in general and in music. Lastly we released the full length song to youtube and Spotify.

List the results (30% of vote)

The overall goal with the campaign was to reach as many as possible, but mainly targeting young adults. With the modest media-budget of 8000$, we reached nearly 500.000 different people during the campaign periode, with over 2 000 000 impressions, and over 52000 post engagements. Norway is a pretty small country by population, with scarcely 5.5 million people. Consider this results with that in mind. But more importantly we worked with scientists to test the limits of modern music technology. With only a month at hand, we train an AI to write and compose a song that even fooled the truest of fans - until they noticed Melissa Reed.