Algorithms with a human touch

Has shaped the concept of “smart curation”: Frédéric Martel. Photograph: ZHdK

What effect is digitalisation having on the future of art criticism and what new income strategies is this involving? In an interview with ISABELLE VLOEMANS, author and researcher at ZHdK Frédéric Martel provides an insight into his extensive studies on the subject.

Isabelle Vloemans: What questions are you trying to answer with your “smart curation” concept and what does it mean?
Frédéric Martel: As a researcher at ZHdK’s Department of Cultural Analysis, I try to understand, to decrypt, and also to imagine the future of the
arts in a digital age. That’s the core of my research here in Zurich. Right now, I’m working in three directions: the future of criticism—how the
digital age will affect recommendations; the changes to the arts caused by social networks; and the new economic models for artists.

The “smart curation” concept, which I coined in a paper for ZHdK, belongs to the first part of the research. It’s an attempt to understand how we will “recommend” music, books, visual arts, or stage plays in the future. This concept can be used for hotel recommendations, tourism, cars, or watches as well. The basic idea is on the one hand the end of reviewing as we used to know it (e.g., film or book critics) given the abundance of content. On the other hand, the very efficient recommendation algorithms (Netflix, Amazon or Spotify) are not accurate enough. I believe there is a third way: combining the machine and human recommendations—the data of the “smart” and personal, opinionated choice (“curation”).

Let’s look in more detail at how “smart curation” works using a few examples. Situation 1: I’m a fan of TV series and am always on the lookout for the next good one to watch. What future scenario will you design for me using “smart curation”?
You can follow the Netflix algorithm and it may recommend a good series to watch (perhaps one that is produced by Netflix, for example “Sense8”). Or you can read the TV-series column in your favourite print magazine and it might recommend another series, for example the new season of “Orange is the New Black” (also Netflix) or maybe “Le Bureau des Légendes” in France. In both cases, you depend on the Netflix algorithm or the traditional critic’s point of view. That might not be accurate enough for you: the first might be too popular, the second too elite. So you can choose another way that will combine data and human recommendations. A tool like “Discover Weekly” on Spotify combines your own playlist on Spotify, the mainstream playlist of the week, and hundreds of influencers (DJs, radio programmers, bloggers, journalists, social media influencers, etc.). That’s the future of recommendation.

Situation 2: I am a singer in a pop band and want my music to be heard — is “smart curation” a threat or an opportunity for me?
I’m convinced it’s an opportunity. If you are an indie band, neither the traditional critic nor the algorithm will have been very kind to you in the past. The mainstream has always been favoured over the indie, exceptions notwithstanding. So, in the end, double or multiple filter tools are better for the indie.

Situation 3: I’m an author and have built up a following on Twitter that’s keen to read and share my posts — who curates what here and for whom, and in what way is it smart?
Social networks are at the core of smart curation. For example, you can follow a book influencer you admire on Facebook or Instagram and like his post. Your friends will see your Like and the Facebook algorithm will increase the diffusion of the content. A very unusual human recommendation on Facebook might go viral—or not—thanks to the mixture of human and algorithmic appreciations. But we are just at the beginning of this reflection and of the development of this kind of tool.

What business models and strategies will grow more important for artists as digitalisation becomes more prevalent?
We need to be very cautious since these evolutions are very complex and predictions are difficult to make. We see and find out every day, even though we might not like this evolution, how artists but also writers or journalists become “brands.” These brands need to be promoted on social networks. New economic models depend more and more on your community and on your social popularity (crowdfunding, the “live” ecosystem, Youtube’s share of revenues, etc.). We are working on these models at ZHdK’s Department of Cultural Analysis. I see a lot of opportunities for start-ups in the sector of smart curation and also for new economic models for artists. At the same time, these opportunities might also become threats without regulations for the Internet giants in Silicon Valley.

What does “smart curation” mean for the future of education at arts universities?
My main concern, as a ZHdK researcher—and our responsibility as well—is to think about the future of the arts and artists. If in 2030 the economic models of artists and how recommendations make them known were to be altered dramatically, we would need to anticipate that shift. I believe we are entering the “digital century.” We are neither in the middle nor at the end of the Internet revolution: I believe we are just at the beginning. We need to understand the transformation, anticipate the evolution, and comprehend the economic consequences of the changes in communication. I’m very optimistic about this new world. But if we are going to help artists succeed, we need to anticipate this future.


Frédéric Martel (, is a PhD researcher at ZHdK’s Department of Cultural Analysis and a French writer. His latest book “Smart” presents multi-year field research conducted in 50 countries on the world’s digitalisation. He published his study on “smart curation” for ZHdK (
Isabelle Vloemans was a project manager at ZHdK University Communications.
Teile diesen Beitrag: