9

FROM CULTURE TO CONTENT

El Ateneo Grand Splendid is probably one the most beautiful libraries in the world. At No. 1860, Santa Fe Avenue, at the centre of Buenos Aires, the Argentinian culture scrolls through its history. Inaugurated in 1919, Teatro Grand Splendid was one of the hot spots of the tango. People danced here every evening for duple rhythm, while in the Studio Grand Splendid, located in the upper floors of the building, one made records of this standard-bearing music of the country. Soon, Radio Splendid would broadcast from the same building and showcase the tango to a wider audience. In the late 1920s, theatre becomes cinema, when the Talkies arrive. For the first time in Argentina, a film is screened in which the mute hero suddenly says, ‘Wait a minute. You ain’t heard nothin’ yet.’ The reaction is historic as the crowd explodes with joy.

Today, the library preserves, on its 2,000 square metres, the whole mythology of this culture of the yesteryears. Like the Hollywood villas built between the two world wars, Balaban & Katz in Chicago, Ziegfeld Theatre of New York and Fox Theatre in Detroit, there still exists the luxurious hall with its grand staircase and decorated ceiling, giant chandeliers with thousands of bulbs scintillating intermittently, caryatids and stained glass and decorative carpets, and of course, two immense red velvet curtains on the stage, which open and close one on the other. In the stalls of the theatre, the thousand seats were replaced with shelves containing 120,000 books. The balconies accommodate CDs and DVDs; the green rooms are quiet spaces for reading. I see many students, from schools of art and creative writing, reading, sitting on the floor, The Library of Babel by Jorge Luis Borges and Mille Plateaux by Gilles Deleuze and Félix Guattari.

Every year, since its opening in 2000, more than a million people come here for their purchases. In the El Ateneo Grand Splendid bookstore, books and cultural goods are a show. But for how long?

On stage, which itself has been transformed into a café, I meet Hernán Botbol, the co-founder of taringa.net, one of the main sites in Argentina. His office being located upstairs, above the bookstore, he meets me here, at Impresso Café, in the middle of books, music, films and magazines. Looking at the shelves of El Ateneo, furnished with cultural goods, he pauses and tells me, with no empathy, ‘All this is going to disappear! All of it!’ He adds, ‘Except when you go to a museum.’

Taringa is a Latino social networking site, like Facebook, where one can exchange cultural content with an unknown person (the Tweets are called ‘shouts’ and they must contain less than 256 characters, and the same goes for ‘reshoots’). ‘On Facebook, the important thing you share is your offline life that you share online. On Taringa, the important thing is content. People follow you for the content you post or produce.’ Something between a peer-to-peer community site and a blog platform of the Tumblr type, it is an original social network which promotes cultural recommendations and exchanges. The site developed a very useful broadcast feature for streaming music in order to encourage ‘conversations’ between its users. ‘The new economic model of culture wouldn’t come from digital sales which are likely to replace analogical sales. It will come from unlimited subscriptions. Buying music, even per unit, even on iTunes, has no future. Nobody is going to buy music any more. That’s how it is. CDs and DVDs are dead, and also downloading. I believe in subscriptions for unlimited streaming. But that has to go through new forms of copyright,’ predicts Hernán Botbol. Around 130 million people visit his site every month.

For the moment, the business model of Taringa is based on advertising, but the hypothesis of broadcasting cultural content through paid subscription, based on the model of Spotify, Pandora and Deezer, is under study. The team is perfecting the algorithm which would be decisive in this unavoidable evolution. An offer of proper content is also envisaged. ‘The creative industries think of the internet as a distribution tool for their products,’ says Botbol, ‘whereas it is also, and I should probably say firstly, a place of production of new culture.’ The site aims to develop in Argentina and the whole of Latin America, and of course, aims at the Hispanic market in the United States. To prepare itself for this, Taringa opened in the autumn of 2013, an office in Miami.

The experience of Taringa throws light on the many deciding factors of culture in the digital age: subscription, recommendation, algorithm, conversations, new forms of copyright and the quality of ‘content’. Clearly still, this site shows that culture, which was a ‘cultural good’, is in the process of becoming a ‘service’. For Hernán Botbol, who talks to me of ‘art in the age of its digital reproducibility’—paraphrasing the famous title of a premonitory article by Walter Benjamin—the destiny of culture is sealed. El Ateneo Grand Splendid is an amazing example of it: the tango hall has become a cinema hall, records were made popular by the radio, the palace at the city centre went bankrupt thanks to multiplexes, the bookstore took over, and now, it is the turn of the internet. Botbol is persuaded that the digital will wash away everything in its way. And having rented his office on its floors, he has a foot in the place.

Today it is no more about asking oneself if the internet will change culture as we know it. This question no longer appears to be pertinent to most of my speakers. Going over to the digital is considerable, total and irreversible. Rather, the question is about knowing how the internet is going to change the deal, to what extent, and what will remain of art, of cultural hierarchy, critical journalism and business models in terms of this revolution which is just starting. I will simply deal with, in this chapter, the hypotheses of all those I spoke to around the world for this book. They too, like Hernán Botbol in Buenos Aires, spoke of recommendations which will replace cultural journalism, subscription which will become the rule for cultural ‘services’ and algorithms which will become increasingly powerful. They bring up the inevitable mutations in copyright and the new business models for the music industry with its platforms, streaming services, crowd-funding, 360-degree strategies. They also told me of Flipkart, Ozon, and Vkontakte, its Indian and Russian equivalents, which take part in the same disruption. Though they often had intuitions, and sometimes a proper agenda, my interlocutors did not always have an overview. Their analyses depend a lot on their age and the country in which they are living, proving that territory continues to play a decisive role in culture in the digital era. Maybe, this compilation of their random testimonies will make sense of what is in the currently ongoing.

The extension of the non-commercial domain of culture

From the window of his office, in the centre of San Francisco, John Kieser shows me the headquarters of Twitter, a few hundred metres down on Market Street. Kieser is the director general of the San Francisco Symphony (SFS), one of the biggest American orchestras. And he understands that he shouldn’t miss the new cultural revolution, the internet revolution. ‘We understood very early on that an orchestra like ours couldn’t just be playing on stage in Davies Symphony Hall here in San Francisco. It should also play on the digital stage,’ argues Kieser.

‘It is certain,’ he adds, ‘that the Silicon Valley made us aware of the digital before anything else did. As we live here, we get to see the ongoing revolution every day. And that necessarily touches an orchestra like ours.’ On the board of administrators of the SFS, one can find certain founders of Facebook, PayPal, and, until his death, Ray Dolby himself. Besides, the orchestra continues to work with the Dolby company which he founded, specializing in encoding, compression and background noise reduction. Together, they are currently developing the project Dolby TrueHD to improve the sound quality of smartphones and tablets.

The digital has thus become a priority of the SFS. Its site (sfsymphony.org) does not only sell tickets to shows and present the programme of the orchestra. It is a truly global platform for classical music, with videos, television channels, two dedicated radios, podcasts and numerous articles on music. On another dedicated site (keepingscore.org), the orchestra has even increased the number of experiments. ‘We want to create a conversation around a true interactive online experience, because for us, the digital is a part of the education,’ explains Ronald Gallman, the director for ‘education and outreach’ of the SFS. On the site, it is possible to follow the partition, measure by measure, of Symphony n° 4 of Tchaikovsky and stop it as required, to listen to the explanatory comments provided by the conductor of the orchestra, and it is fascinating. Holding the conductor’s baton of course is the forthcoming Michael Tilson Thomas, director of the SFS. And everybody knows him here as MTT.

MTT is omnipresent in the city and online. He has a Twitter account followed by 70,000 people (@mtilsonthomas), an enriched Facebook page and a LinkedIn account. Very well connected with the networks of the Silicon Valley, he is regularly in touch with the heads of Google, Pandora and Twitter, whom he invites to the shows by his orchestra.

‘For MTT, it is a real mission. His grandparents were in the Yiddish theatre, his parents worked for television, so he thinks that his mission, for his generation, is the digital,’ explains John Kieser. Through regular dialogues with YouTube, MTT succeeded in convincing them to launch a dedicated channel (YouTube Symphony Orchestra). The online concert of 2011, performed live under his direction at Sydney Opera, was viewed by thirty-three million people throughout the world. ‘We expect no returns for investing in our digital projects. We consider it simply to be a part of our mission, just like education,’ says Kieser. (He confesses that the site Keeping Score has a budget that surpasses $25 million, running on philanthropic donations and the patronage of the internet giants.)

MTT intends to continue down this line, and other innovations are in the offing. A new YouTube channel will soon be dedicated to the SFS; and video games are being developed in partnership with the gaming department of the University of California in Irvine. MTT would like to go further by creating a web radio and a type of Khan Academy for classical music where short videos will enable learning music.

‘We think that classical music is at a turning point in its history. Its presence on the internet is a full-scale test for its survival. If we cannot create a conversation around it and have great online presence, we won’t be able to convince the audience to physically go to an auditorium,’ concludes John Kieser. While listening to him, I notice, placed on a shelf in his office, the famous collection of all the symphonies of Mahler, led by MTT. In CDs.

In the United States, the SFS is not the only cultural establishment to be a step ahead. The Chicago Symphony Orchestra (CSO) has also developed an original site (cso.org). This site includes a media player and a radio station. It also provides to the public an impressive quantity of resources: Making Music, a guide to learn music; Ingenuity Incorporated, an artistic education platform for schools; and Artlook Map, an interactive map for the artistic places of Chicago.

As for museums, the Metropolitan Museum of Art in New York spearheaded the digital revolution. Its director prognosticates that internet and technology will bring about a transformation in the relationship between museums and its consumers and, at the same time, revolutionise the place of art museums like never before. In 2011, the MET therefore began to change its policy, not only by allowing usage of smartphones in the galleries but also by encouraging them! The digital team of the MET, comprising of 70 employees, created applications, online catalogues and regularly posted photos of their museum artefacts on Instagram. Sometimes, the MET narrates, via photo and video, a true story or an event, such as the acquisition of Portrait of Everhard Jobach and his family, a noteworthy piece by Charles Le Brun, whose arrival in New York in its safe crate, opening, transfer and display, were posted on the Museum’s site. But the MET didn’t limit itself to thinking uniquely about its own visitors, it chose to address itself to the whole world. It took to posting messages on the Chinese social media, on Weibo for example (through sixty posts, more than three million people were targeted). Notably, a meticulous timeline of art history and a detailed artistic chronology dating back to 8000 BC, was put up online, enriched with 7,000 objects and more than 900 texts written by experts. As complete as many works of art history, which serve as reference books, the timeline is today one of the most visited art pages in the world (metmuseum.org/toah). The MET is already thinking about the next step: all the museums would, one day, be interconnected and its timeline would serve as a model for them.

For museums, the Museum of Modern Art (MoMA) of New York (moma.org) offers on its site online visits of the modern art museum, an interactive presentation of paintings and art resources, even if, unfortunately, some tools are only available for members. The Museum of Fine Arts in Boston (mfa.org) offers similar tools and specific content for the social networking sites. Often, these museums make available online excerpts from works when they are digital (The Clock by Christian Marclay, All By Myself by Nan Goldin) or funny supplements (the famous video Good Morning by Kanye West illustrating an exhibition by artist Takashi Murakami, for example). They are also working towards digitizing works of art into 3D. For its part, the MoMA already put forward a proposal for a Kraftwerk retrospective in 2012, promoting the music of the German electronic band, during the exhibition and on its site, especially two of the band’s unbelievably prophetic titles Computer World and Home Computer.

American cultural spaces all have a priority to engage in ‘conversations’ with their audience. Paradoxically, the shift to the digital helps them in doing this. With their sites, applications and their online presence on social networking sites, they grow closer to their members, speak to both teachers and students. They also place themselves better in the ‘community’ by being connected to social educators and the entire sector known as CDC (Community Development Corporation) which irrigate the sensitive areas. Surprising though this may be, the internet is not, in their eyes, a global communication tool, but on the contrary a means to relocalize themselves by participating in the local conversation. For, when the Museum of Fine Arts in Houston or the Philadelphia Symphony Orchestra invest massively in their digital presence, it is firstly to connect with their potential audience rather than with tourists.

These examples illustrate the importance of the not-for-profit cultural sector in the United States and the development of non-commercial activities online. For some of my interlocutors, this expansion of the non-commercial aspect of culture seems to be the future model. If the loss of value of cultural goods takes place due to the internet, entire sections of culture can swing from profit to not-for-profit.

In the United States, symphony orchestras, museums, operas, libraries, ballets, theatre (except Broadway), art house cinema and thousands of film festivals, university presses and most of the universities (not financed by the states) belong to the ‘not-for-profit’ sector. These cultural institutions are, most of the time, neither public, nor even ‘private’, in the mercantile sense of the word: they are independent and not-for-profit.

Their status is determined by an important text, the article 501c3 of the American fiscal code, which gives, by way of compensation for a general interest mission, tax exemption on philanthropic donations. If you finance a not-for-profit museum, you would get a tax reduction. Facing such a major cultural sector in the United States, which is often unknown in the rest of the world, is of course a powerful commercial sector which includes an important part of cinema, publishing (except university presses), recorded music, rock, jazz and rap concerts, contemporary art, Broadway theatre and video games. These creative industries are commercial enterprises and cannot receive philanthropic donations, nor of course fiscal exonerations.

These two cultural differences are fundamental and the American tax authorities keep their eyes open, even though there have been drifts from the norm and sometimes, scandals. The board of administration of a not-for-profit institution must, for example, contribute to the financing of the cultural institution every year, by giving hefty donations—which sometimes run into hundreds of thousands of dollars—whereas the board of a creative industry shares the profit among the stakeholders at the end of every year. They are two worlds unknown to each other, and all American cultural models are built on this fundamental division.

Can one imagine an extension to the not-for-profit domain of culture? This is a hypothesis. Wikipedia is an example, breaking away from the commercial models of classic encyclopedias. In the same way, free software, such as Linux or the search engine Firefox, non-proprietary licence, open content, content-sharing or mutualizing sites offer a multitude of creative tools and content.

Ultimately, some of my interlocutors think that the line that separates not-for-profit culture from the market culture is intangible. They remind me that the museums and the ballets were sometimes commercial in the United States, in the nineteenth century, before fiscal deductions were offered to philanthropy, bringing in donations and making entire sections of culture shift to not-for-profit. We can guess that some sectors, such as quality publishing, classical music, jazz and art house cinema and auditoriums may be forced from making profits—especially if the internet deprives them of it—and evolve into a system of philanthropic funding by becoming not-for-profit. This would be another way to think of new culture models in the digital age.

Beyond this possible extension of the non-commercial domain of culture, which to some appears to be the future, others would rather have a revision of copyright rules or their relaxation.

Creative Commons

In The West Wing series, he appears in the guise of an actor who plays himself with his real name. A law professor. Mandarin Chinese. And in season six, he spends a part of the night chatting with the president of the United States. As if intellectuals spoke to heads of state! In real life, Lawrence Lessig does not advise princes. He defies them, provokes them and confronts them for their contradiction. ‘I defend free culture in the same way that I defend free speech. “Free” to me means to have liberty, and not free of cost,’ explains this man in his small glasses, professor at Harvard and an activist.

Among the thinkers of culture in the digital age, Lessig is one of the most precocious. Often considered radical, and castigated, he is firstly an innovator. Very soon, he saw that copyright will be profoundly impacted by the digital transition. So, he created the Creative Commons (CC). This licence allows creators to define for themselves the nature of copyright that they want to attribute to their cultural content. There are six of these which allow free sharing of a work on the condition that the name of the original creator will be cited (CC BY). In some cases, it may be possible to remix and freely adapt the content for commercial and non-commercial purposes (CC BY-NC), and often on condition that the final product must be reused in the same conditions and under the same licence as the initial product (CC BY-SA). Lessig explains, ‘Author’s rights licences and the tools of CC complement the copyright system. They go beyond copyright without cancelling it out. At the same time, they give authors and creators, and it is very important to do so, a simple means of choosing the kind of freedom of distribution that they want for their work. So, if you are Britney Spears and if you want to be listened to only by those people who paid for your music, it is your right and I have nothing against it; but you can also be an amateur, a teacher or a scientist and may want to create something for the love of art or science, or to be recognized, without worrying about the money that it may bring you or controlling the way in which your work would be distributed. That is where CC come in.’

Today, hundreds of thousands of sites and millions of contents are under these free licences throughout the world. The most famous are the Wikipedia articles, a part of the Flickr photos, Al Jazeera reports on the Gaza war, some editions of 20 Minutes, the French news site and of course the books by Lawrence Lessig. ‘I am frankly outraged by the functioning of content industries that have an extremist vision of copyright, which has nothing to do with what the historic fight for intellectual property was. This abusive vision jeopardizes the possibilities of the digital age,’ insists Lessig. For him, the reason for the existence of copyright is to encourage innovation and not to protect the model of ‘dominant cultural industries’. Like other intellectuals, he suggests reducing the individual penalties in case of piracy, renew law on private copying and broaden the rules on fair use (exceptions to copyright for general interest, for example, distribution of quotes or short excerpts from films). He also intends to promote mash-ups, musical melanges and hybrid videos, made by mixing different works. And of course, he fights for the reduction of the duration of the copyright which in the United States is seventy years after the death of the author. A pessimist when it comes to copyright, Lawrence Lessig is finally an optimist regarding creative content in the digital age— thanks to the internet, he thinks that culture is becoming more and more enriched and more and more sophisticated.

The CC places itself in the movement of collaborative economy and in the phenomenon called ‘copyleft’ which applies to free software users and those who share and modify them, their reuse also being free. Or how copyright is used to prohibit the setting up of copyright!

‘In matters of copyright, I am not an abolitionist,’ insists Lessig. ‘I think that copyright is an essential element in the modern culture and creative economy. It is necessary to a certain point. It must be an encouragement and an incitement for artists to create great works of art and for writers to write great books. Copyright must be defended as much as possible to defend these objectives. But not beyond this. It is about finding a good balance between necessary protection and excessive regulation. Otherwise, instead of having a free culture, we would have a regulated, and even captured culture, beneficial only to content industries, some artist lobbies or the elite—a culture of sharecropping—and that wouldn’t allow us to celebrate each person’s potential or even inspire most people to create.’ (Lessig deliberately uses the term sharecropping, which refers to a practice in the tenant farming system where a farmer, generally Black, must share profits of his harvest with his lessor, generally White.)

Such words shock certain cultural players. Messiah for some, Lessig remains a Black beast among the author communities. His dissenters accuse him of playing into the hands of piracy and illegal downloading. While copyright is threatened on the internet, they extol, on the contrary, strengthening the protection of artists. Without copyright, art would be in danger and artists condemned to depend on patronage, like during the Renaissance. Some even hope to promote European royalties because it would be the author’s ‘moral right’, more secure than the Americanized copyright. In Brussels and in the European Parliament, there is currently an ongoing debate regarding these issues—between the upholders of royalties and the partisans of copyright and those who are in the line of Lawrence Lessig and want to relax the rules. A harmonization is under way, which, some fear, would take place at the slightest common denominator. The battle within the European Union can therefore contribute to define the future terms of the debate.

In any case, a certain liberal or libertarian spirit of the internet has set a precedent and has many aficionados throughout the world. Everywhere, and especially in emerging countries, I met these cultural players who sincerely believe in ‘free culture’ and sometimes ‘free of cost’ culture. Their champion is a musician who is over seventy years old, Gilberto Gil, a famous bossa nova singer and one of the inventors of Tropicália. When I interviewed him, he was the minister for culture in Brazil, under the Lula government. He confirmed to me that he was a ‘great supporter of CC’, adding that he defended ‘of course’ royalties, but all the while being committed to promoting ‘free distribution of music’. He even said jokingly that he himself was a hacker and believed in non-commercial forms of cultures. He thought of the young Black man that he had been, who grew up in poor Brazil, and now he wanted to help the youth in the favelas to communicate through the digital. Gilberto Gil, on the evening of our interview, sang On the Internet on stage: ‘I want to be online/ Promote a dialogue/ Gather on the internet/ A group of fans from Connecticut/ I want to get in the net to contact/ The homes of Nepal/ the bars of Gabon’.

The end of cultural goods

‘The disc industry has no sense of humour.’ One remembers this famous line by Justin Timberlake, who plays the role of one of the founders of Facebook in the film The Social Network. Unauthenticated or not, the catchphrase captures the complicated relationship between music and the internet. Today, we are not there any more. For, music has found new happiness.

Now, the music industry has got used to the digital. Most songs are available online and the legal offer is satisfying. Its economic model is currently heading towards three main directions—free digital smart radio, linear without download, sharing advertising or remuneration of royalties (like Sirius XM, Samsung’s Milk Music or the iTunes Radio); paid downloading per song or per album (like iTunes); or unlimited access through subscription for streaming, by downloading or via an application (such as Spotify, Pandora or Deezer).

Though the ‘smart radio’ model has its advantages for the consumers, especially as it’s free of cost, it is disadvantageous to the major players of the music industry in not being sufficiently lucrative—royalties that the artists get are smaller in cases of free subscription, even with publicity sharing, when compared to paid subscription. (The payment of royalties is different as the smart radio does not allow the listener to choose the pieces he listens to, even though he can repeat or skip certain songs). Despite these pitfalls, the system is likely to last because it caters to a very strong demand of the consumers, especially in developing countries or in Africa. Web radios are increasing in number everywhere and are becoming professional. (TuneIn Radio Pro, SHOUTcast, iTunes Radio, etc.)

For its part, the model of paid download, popularized by Apple, has reached it’s limit. The invention was decisive, mixing four revolutions in one: an innovative musical file reader (iPod and then iPhone); a new mode of coding offering better sound quality (Advanced Audio Coding [AAC] instead of mp3); a platform which served both as a free personal library and purchasing space for music (iTunes); finally, a major invention, customized sale of music, by title and no longer by the album. This quadruple prowess of the Apple brand however is struggling. ‘People buy music to discover; and now, they discover in order to buy,’ says Patrick Beauduin, the director general of Radio Canada, interviewed in Montreal. A too optimistic opinion? ‘People buy hits. It is the only thing they buy. And moreover if we sell them these songs individually, rather than by the album, it would be difficult to survive. A model based on long tail, an idea that they would buy back catalogues is a myth,’ explained somewhat bitterly in London, Nick Gatfield, one of the managers of EMI (the British EMI has been bought by Universal with a small part [Coldplay, David Guetta, Tina Turner and other artists] going to the Americana Warner). Though iTunes remains the dominant player in the purchase of music online (around two-thirds of the market share of paid downloading), with all the abuse of power that this monopoly entitles it to, it no longer seems to correspond to the practice of consumers. Buying music in this form does not appear, to most of my interlocutors, as a formula for the future.

And there remains subscription. If this model does not replace the losses incurred by the music industries due to CDs and the general shut down of their dealers—from HMV in England to Borders in the United States, and Virgin in France—it appears to be a durable solution for the industry. ‘We are convinced that this third model, subscription, is the most promising for the future and a solution for the consumers and for the music industry,’ says, for example, Pascal Nègre, the CEO of Universal in Paris.

Many platforms already occupy the subscription niche for unlimited streaming, whether it be Spotify, the leader of the market created in Sweden in 2008, or its American counterparts such as Rhapsody, Pandora, Rdio, or the French Deezer or yet, new entrants such as Qobuz or Naxos specialising in classical music and high quality sound (one can also count the new Napster). The internet giants are not far behind: Google with its Google Play and Music Key (YouTube), Microsoft with its Xbox Music, Sony with its PlayStation Music (in partnership with Spotify) and more recently, Apple which, having acquired Beats Music in 2014, launched Apple Music, thus establishing the relevance of the model. (Amazon has a similar unlimited offer, Prime Music, reserved for its Prime subscribers).

It is necessary to mention YouTube again, the indispensable player in the free music ecosystem, which belongs to Google. A hybrid model, it appears to be a sum of services with a million individual visitors every month, who look at six billion hours of video on an average. YouTube is firstly a search engine for videos—the first in the world. It is then a free service of videos and music online, in streaming, with advertisement, but non-linear (one can choose his titles, even if one cannot download them). Now, a majority of its content is legal and YouTube sends around 50 per cent of its advertisement revenue to cultural industries. Mobile became its priority—it already represents 40 per cent of its traffic and one can make the hypothesis that YouTube would place itself essentially on the mobile market for smartphone and tablets. In any case, its original model has had a phenomenal success—YouTube is the leader in free streaming, often with more than 80 per cent of the market share. (It is only challenged in some countries by local players, such as the Chinese Youku Tudou, the Japanese Niconico, the French Dailymotion or the Iranian Mehr.) In parallel, YouTube launched free thematic musical channels which are similar to ‘smart radio’ (or, as it happens here, to ‘smart TV’). In the end, YouTube would have to evolve towards paid streaming by subscription by launching an offer halfway between the Spotify model and Netflix, music and video, with the strike force of the search engine Google. One can even imagine that YouTube, which has probably pacified its relations with creative industries only tactically and temporarily, could, through an expected turn of situation, start making them pay for the referencing of their videos and music! The ceiling of the free offer of music but the start of paid referencing—we see that the models are not stabilized for creative industries.

At the end of the day, Google, Amazon and Apple would evolve from an ‘online music locker’, this personal ‘compartment’ where one stores one’s music, in the ‘online music database’, a collective database that one doesn’t own but which one can access at any time.

This model, sometimes called ‘all you can eat’, echoing restaurant menus with an unlimited buffet, or more seriously ‘open music model’ has the merit of being very adapted to mobility. With respect to ‘smart radio’ which necessitates an internet connection, or paid downloading which is expensive, platforms like Spotify or Deezer have the advantage of satisfying both customers and the majors at the same time. It is an evolution of the spirit! The former, it was said, would never want to pay for music since the Napster phenomenon made them believe, more than ten years ago, that ‘music is free’; they would continue to illegally download their favourite pieces online. The facts prove the contrary. Apparently, they are ready to pay for Spotify, even though the numbers are still precarious (twenty-four million users of the free offer with advertisement and six million for paid subscription without advertisement in 2013). As for the majors of the music industry, their turnaround is still more spectacular. The idea of a free formula for streaming, accompanied by advertisement or a model through subscription, like Spotify, provoked screaming protests less than five years ago. Today, three majors in music—Universal (belonging to the French Vivendi), Sony (Japan) and Warner (United States)—embrace these new models and even want to go faster and further in this direction. ‘The fact that a festival such as ours, entirely centred on music until late 2000s should today be the main digital innovation festival in the world, is a great summary of the evolution of the music sector,’ tells me, in Austin, Roland Swenson, president of South by Southwest or SXSW. (In parallel, the SXSW Interactive Festival has become, every spring, in Texas, the place where new ‘hot’ start-ups are released: Twitter made a buzz here in 2007 and Foursquare in 2009.)

The fact remains that the majors are now beginning to dream—they prognosticate that when streaming by paid subscription would have reached, for example, a hundred million subscribers throughout the world, their financial input to the musical branch would have almost compensated the lack of earnings due to the collapse of CD sales. They repeat that Spotify has already given the industry a billion dollars in royalty since its creation (even if the amount is limited to an average of €0.05 per song heard). Wishful thinking, some would say. But it is better to think of desires as realities than persisting, like the disk industry in the last few years vis-à-vis the internet, in denial of reality.

In its way, this subscription model, which emerges in many countries that I visited, also confirms the major revolution of culture—transformation of cultural goods into services, streaming and subscription. Music is no more an object one can possess but something to which one has mobile access and to which one can listen on all media, where one wants, when one wants, through one and the same general subscription.

This ongoing process is the natural evolution of the digitization of content. However much one sticks to his old disks, personal and unique, and his DVD films painstakingly collected, the pieces of music in mp3 format or ACC and all the movie files downloaded look alike. What is the use of continuing to hold on to them in a ‘library’, a word that, predict some, also belongs to the past. Instead of claiming for itself all the cultural objects in a step focusing on bourgeois accumulation, one will therefore be happy to have access to it—subscription, and not property, will be the future of culture.

This passage from the culture of ‘goods’ to ‘services’ is a fundamental mutation. One is moving from the industry of cultural goods to an industry of cultural services. If the disk industry hadn’t waited ten years to adopt the internet, it might have been able to anticipate this major evolution a few years earlier.

Other than these overall evolutions, other leads are explored everywhere in the world: ‘360-degree deal’ management contracts, crowd-funding, sale of telephone ringtones (a sustainable phenomenon in Asia) and, of course, concerts and live presentations. Certain artists, like Jad Shwery, explore all these trails simultaneously.

‘I am starting my second career,’ says Jad Shwery, thirty-three years old, a successful Lebanese singer-turned-producer and director of video clips. When I meet him in Beirut, I understand the importance of his entourage and that, from the look of the passers-by, I am dealing with a star. For ten years, Shwery has been among the most popular Arab pop stars, accumulating hits and first places among the top fifty in Lebanon, Syria in the Gulf, Maghreb and especially in Egypt, where his first album sold 200,000 copies in a country where piracy has for long been massive. One of his hits Warreiny, has been viewed over one million times on YouTube, and Shwery also had huge success with Funky Arabs, Banadilak, Wala Awel and, in November 2013, the very hot We Don’t Care was a huge hit. His videos were such a success on the internet and satellite TV that he got an exclusive contract with a television channel Melody TV which owns its own production branch.

‘This type of deal is called a “360-degree deal”. It covers albums, concerts, by-products and everything that concerns management. Production houses then get a percentage on all these services,’ explains Shwery, during a get-together at his apartment in Achrafieh, the Christian area of east Beirut.

The Egyptian major Melody Music is not the only production house to have tested this model. Its competitors in the Arab world, such as the Saudi group Rotana or the Lebanese Arabica, do the same. ‘We have implemented what we call “a management contract” with our artists. We do, like everybody, 360-degree deals. It means that we don’t sign a singer to make only discs but to manage his entire career: his CDs, his online music sales, and also the concerts and advertisements he acts in.’ Ninety per cent of his artists are now under this kind of global contract, clarifies Tony Semaan, the artists and repertory director at Rotana when I interview him at the headquarters of the company in Beirut.

With the switching over to the digital, this very unique model became gradually generalized. ‘The paradox is that the Arab countries are ahead of the rest of the world,’ jokes Jad Shwery. ‘Piracy rates were so high here, even before the internet, that one couldn’t count on the windfall from the sale of an album or singles. Everything was pirated through millions of duplicate audio cassettes and copied CDs. My first contract in 2005 was already a management contract with a percentage taken by the production house for all my concerts and my publicity. It’s what everybody now does in the United States and in Europe!’

What struck me in Lebanon, like elsewhere in the Middle East, as well as in Asia and Latin America, is the serenity of all my interlocutors in the face of piracy. According to Jad Shwery, it is the industry’s given and one does not take offence. ‘It is in stark contrast,’ he says, ‘to the anti-piracy obsession among the American industries or the European government.’ According to the music professionals in the Middle East, piracy exceeds 80 per cent of the market. And even when CDs and DVDs are sold on the streets, they are almost always pirated. ‘So the peer to peer doesn’t scare us. We are not afraid of internet piracy!’ confirms Shwery. A study of The American Assembly by the Columbia University estimates that piracy is at 68 per cent for software in Russia, 82 per cent for music in Mexico and 90 per cent for cinema in India; (the study also says that 46 per cent of adult Americans have already illegally downloaded music or films and that this number goes up to 70 per cent for the 18–29 age group).

Other specificities particular to Arab countries exist. It is the case of services provided for private parties, birthdays and especially weddings which create regular windfalls for artists and their production houses interested in their percentage. In other countries, such as the United States, debut artists have recourse to ‘house shows’, by playing in private houses to increase their fan base and get small remuneration through tips.

Though he continues to sing, Jad Shwery has taken up a second career today—he is moving towards the profession of a producer and manager. He also directs a number of videos for Arab stars such as Nawal Al Zoghbi, Ouadia Al Safi and Diana Haddad. ‘The production and direction of a clip can bring €15,000 and 15,00,00, so it is a true business model.’ The success of videos on the internet being increasingly crucial to create a link with the audience, online promotion strategies are now proposed by a number of specialized start-ups. ‘The online success and offline marketing complement each other. They both take part in the conversation that the artist must have with his audience,’ Shwery makes me understand. On the other hand, unlike in the Western countries, it is out of the question to hope for windfalls for televised services or passages on the radio. ‘In the Arab world, production houses and artists must on the contrary pay the channels so that their music is broadcast! It is another particularity here,’ says Jad Shwery. He adds, ‘But we are probably at the end of this model because of the internet. Arab artists are starting to make money when their videos get a good number of views on YouTube.’ The site and social networking, like the one that launched artists such as Justin Bieber, now takes part in financing the videos that it broadcasts. One can find similar models under other latitudes. Thus, paying to broadcast one’s music on channels—a generally illegal system called ‘payola’—is frequent in Latin America. ‘Payola is what has for a long time explained the dominance of the majors and the standardization of the market by Anglo-Saxon music. But, today, the internet doesn’t allow such corruption any more, at least not to this degree. There are alternative solutions online. It is good news for us,’ testifies Victor Ponieman of Random Record, whom I interview in Buenos Aires.

As for complementary remunerations such as mobile phone ringtones or by-laws for karaoke, they have become widespread in Asia. ‘Ringtones and ring-back tones are the economy of music now,’ explains, Toto Widjojo, president of Sony in Indonesia. Japanese pop music (J-Pop), Korean pop (K-Pop), Taiwanese and Indonesian have inundated the cellphones, especially thanks to ‘ring-back tones’, call-waiting ringtones (not to be confused with ringtones themselves); instead of listening to the ringing, you listen to an excerpt of a J-Pop or K-Pop song. Not to mention the very popular ‘colour call tone’, the music one can hear in the background during a conversation! As for album covers, they are naturally sold as screen savers for mobile phones. These Japanese inventions that I could see everywhere in Asia, in Korea as in Indonesia, in Taiwan as well as in Vietnam, also bring in royalties.

These new forms of remuneration also extend to syncing (synchronization of music to a visual medium). It holds true for TV adverts and film music, syncing for video games, branding or brand content for big brands, sound designing for websites, etc. Sometimes the copyright revenues that come out of this surpass those generated by album sale. Big music companies have created dedicated departments, and special agencies for music syncing have cropped up. There is no longer a lone economic model for music: one must now speak of it in plural.

Thus, ‘exterior activities’ now constitute a priority for the company Condé Nast (which is the editor of, among others, Vanity Fair, Vogue, GQ or Wired magazines). At the headquarters of the company, a tower of fifty-two floors at the heart of Times Square in New York, they explain to me that colloquiums, conference series, salons, exhibitions are the new economic models for the company, for its writers and its journalists. Henry Finder, editor-in-chief of The New Yorker, with whom I have lunch at the cafeteria of Condé Nast, beautifully designed by architect Frank Gehry and where Meryl Streep was filmed in The Devil Wears Prada, says that the authors of his newspaper, such as Malcolm Gladwell, earn from such kinds of external services as well as from their pen. But of course, he explains, it is easier for a writer who produces best-sellers in a series, like Gladwell, than for lesser known journalists. Success calls success, in music and in literature. ‘It is very unequal and I don’t know to what point it is scalable,’ says Finder.

There remains crowd-funding. It mainly addresses upcoming artists, those that are promising but who have yet to make it. What does one do today, in a world with so much talent and so many projects online, to be recognized and financed? Crowd-funding tries to solve this difficulty. It is about calling out directly to the public who would then bring their support to a project that they like. This collaboration may be based on donation or be an object of real investment.

In recent years, participative finance platforms were created as in the case of Kickstarter, the leader of the sector, and also RocketHub and Indiegogo that many serial and documentary directors, or film directors such as Spike Lee and James Franco, used to get donations and finish their projects. Sometimes, there are more specialized sites, such as Ulele (creativity and solidarity), People for Cinema (cinema), Touscoprod (audio-visual) or MyMajorCompany (music), which are committed to facilitating creation of projects in a precise sector through collection of funds.

One can see that the music sector overflows with inventiveness to find new economic models for the digital age. The industry can benefit from adopting the Web because on it, it can get not only funds but also diversity. Commentators, often hostile towards the internet, have much prognosticated that the digital transition would come through standardization. But today, when one questions the professionals in the field, in many countries, they are not convinced by this analysis. The fear of globalization results in the success of Anglo-Saxon music because the internet is, according to them, baseless.

An analysis of more than a million categorizations of the best-sellers in twenty-two countries, undertaken in 2013 for The Economic Journal, attests, on the contrary, that music remains largely territorialized. Local bands would have increasing influence thanks to the internet. And in reality, while the United States and the United Kingdom, and to a lesser degree, Sweden, Canada and Australia, dominated world sales before the digital revolution, the market has become much diversified today. Of course, the United States continues to dominate the market with twenty-three out of thirty-one artists featuring on a majority of charts. This American mainstream is but one layer, important certainly, but relative, in the quantity of music listened to across the world. The number of national artists has increased since 1990 in the best-seller list in most countries: this is a decisive point as true diversity consists not only about multiplying the supply and the number of titles available in an artificial manner, but they must also be accessible, availed, and enjoy a certain success. It is the case. The MTV group understood this very well and developed its local channels for most of its programmes before the internet duplicated this phenomenon. In particular, by placing live concerts at the heart of the economic model of the music industry, the internet mechanically contributes to increasing the value of local methods of promotion and increasing the sales of local artists. In a connected world, artists from small countries can, from time to time, find international listeners, and, particularly, they can get more recognition at home.

Another factor is the paradoxical effect of algorithms and recommendation. They are global and anonymous tools, and one intuitively thinks that it increases cultural standardization. It is in fact quite the opposite. As they are more and more sophisticated, they take into account sensibilities, nichés and nuances. Their risk, on the contrary, locking the user in his bubble, in his galaxy, and render him communitarian, and only give him what he is already consuming. Therefore, customization of recommendations creates specific conversations which risk strengthening groups and maybe ghettos, even if it can create a bigger territorialization, and let us hope, more cultural diversity.

Finally, it is possible that the global dimension of the internet also paves the way for the distribution of national artists through expatriate communities, which is seen in India, Iran and China. The distribution may be global on a technical plane, but it increases diversity on the musical plane, as it allows members of an immigrant population to take part in conversations of their country of origin. What complex ways does diversity take! At the end of the day, one can hypothesize that music is on the threshold of a new age, and thanks to the digital, its geography will change by becoming more divided and diversified. Much less mainstream.

The other side of the coin is that, paradoxically, it is more difficult to become global. Apart from a few exceptions, such as Psy which is approaching two billion views on YouTube with his Gangnam Style, it is difficult to stand out from the crowd, except for Americans. For many artists, the dream of a global hit is far away, even though the comfort of having carved a niche for themselves becomes real thanks to the internet. This is what José Eboli, the CEO of Universal in Brazil, thinks. ‘Here, in Brazil, the music sold is 70 per cent Brazilian. Americans come only second, with 28 per cent of the market. We are quite proud of ourselves. On the contrary, at the international level, there is hardly any Brazilian music left. Sometime ago, with bossa nova, we were mainstream around the world. But it didn’t last long. Since then, we haven’t been an important player in the music industry internationally. We find ourselves everywhere in the category of “world music”. But we maintain some commercially interesting niches and to cultivate those, the digital is perfect.’

These niches may be particularly rich, as attested by the importance accorded to jazz and classical music online. Classical music featured at 3 per cent in sales of discs in late 1990s as opposed to 33 per cent in the early 1960s. Today, it has made a comeback with true vitality in the digital age. It is estimated that the sales have now surpassed 10 per cent on iTunes. (iTunes does not disclose its sales numbers but this percentage was obtained by gathering rights reserved to disc companies in a study conducted by a special American press.)

The sites which offer classical music online, by streaming or subscriptions, are numerous and special web radios are also increasing in number. ‘The musical radio running on hertz is about to disappear. All that is “disked” as they say in Quebec, has no future. Unless it is live, exclusive or comes with free reign. The future is in special web radios, for example, in classical music,’ says, in Montreal, Patrick Beauduin, the director general of Radio Canada. The site espace.mu, developed by this state-run Canadian group offers a number of web radios in streaming, with a large portion dedicated to classical music. ‘But a classic radio, in hertz or on the internet, cannot speak for more than six minutes an hour, that is the maximum, or it wouldn’t be a musical radio, it will be a “speaking radio”.’ Beauduin thinks that the future of radio, which is a local media, would be essentially mobile on tablets, smartphones and Facebook. ‘Radio is totally compatible with social networking; it is its history. Interacting is key.’ According to him ‘speaking’ radios will also maintain their monopoly in the morning on waking and during the ‘drive time’, the time spent listening to the radio while at the steering. ‘But one would essentially start streaming radio as the bandwidth increases and 4G becomes commonly used. The podcast is a technology of transition, its fault is in being cold and pre-recorded. What condemns it is abundance and the fact that a programme is either swiftly perishable or less impertinent. A sustainable radio does not exist.’ A good example is the smartphone application of the National Public Radio (NPR) in the United States. The efficiency of access to stream programmes on more than a thousand ‘unionized’ stations of NPR, the constitution of playlists, the sharing facility show that the future of radio is in mobile streaming. (NPR has thirty-five million weekly listeners of the average age group of fifty; on a mobile platform, this number falls to thirty-five years.) In Sweden, it is also known that radios have finally chosen to be accessible on Spotify, when the listeners stopped listening. And in England, I was able to see, when I visited the BBC, that forty thematic web radios, of different ‘genres’, were launched, particularly in jazz and classical music.

The classical is also developing on sites for video streaming, including as a part of the Netflix package. An unexpected phenomenon could be seen in the last few years—a persistent interest among the music lovers for a visual experience of concerts and classical operas, which translated yesterday into the rise in sales of DVDs and today into an online video package. ‘It is counter-intuitive, but we ended up seeing that classical is heard through images and videos. The classical web radio will have to offer pictures, it is a fascinating discovery,’ says Beauduin.

Ultimately, classical music has been given, thanks to the digital, a second life and new listeners, keeping alive the utopian dreams of democratization conceived by Leonard Bernstein heading the free concerts by the New York Philharmonic or by Arturo Toscanini and his memorable retransmissions on the 1940s on NBC Radio. Yet, this is only the beginning.

Algorithm

Less than 500 metres from the Capitol and the Congress of the United States, on C Street, there is a small red brick house, dating back to the late nineteenth century. Discrete and refined, it is like its occupants: the lobbyists of Amazon. While the corporate office of the world’s first commerce group is in Seattle, in the Washington state, in the north-west of the United States, its political office also has an office in Washington, the American federal capital. Emmett O’Keefe, Amazon’s director of public affairs, and Shannon Kellogg, in charge of lobbying for Amazon Web Services, extend a courteous welcome to me. My hosts present very well. With a clergyman’s patience, they respond to my questions with prudence. When, in late 2013, the Snowden Affair was the talk of the town, their arguments were measured to mitigate the concerns of a European. Yes, Amazon is favourable towards Congress intervention to protect privacy; yes, Amazon agrees that each company must pay its taxes locally; yes, Amazon shows each user its Cloud where its data are stocked. The arguments are prepared and my hosts, courteous. Even when I raise my tone, they remain calm. I am sitting face-to-face with seasoned lobbyists.

‘Every time one enters into a new sector, one creates a disruption. And our rule is simple: what is good for the consumer is good for Amazon,’ says Emmett O’Keefe serenely. The man does not wish to be connected with other GAFAs, he throws light on the divergence of view, yet also recognizes the convergence.

The term ‘GAFA’, an acronym for Google, Apple, Facebook and Amazon, denotes companies that have no other link but that of being digital and American. In Washington as in Brussels, they have opened their offices. They finance political associations, think tanks, and hold powerful lobbies. They are in frontal competition with each other—Apple and Google in the mobile; Amazon and Apple in tablet, music and e-books; Amazon and Google in e-commerce and Cloud; and Facebook in social networking, video and photos—but come together to defend their interests. Dematerialized culture is one of them.

To the public, Amazon is firstly a site selling tangible goods from actual warehouses. The company is ‘tangible’, or as they say in the United States, ‘retail’ and ‘brick-and-mortar business’. From books, CDs and DVDs, Jeff Bezos diversified by selling video games, electronic appliances, then clothes, furniture, toys and jewellery, and even original works of arts, fresh produce and wine. Amazon owes its success to the impressive number of products available, its rapid delivery (and the fact that Amazon does not go to countries and regions where geographical communications are difficult), to its redoubtably performing search engine, to its payment platform which is among the most efficient on the internet. One may have reason to fear that in the future, the e-wallet of Amazon would serve, in contest with those of Google, iTunes and Facebook (Facebook Credits), as the obligatory passage for online payment. Facebook login also constitutes a means for online identity verification for thousands of sites, and this all the more so as the company insists that the users provide their original name. For its part PayPal, which belongs to eBay, follows an economic model to facilitate financial transactions or auctions online. Amazon also made a lot of advancements in the financial growth market, where the confidence in a brand and the technical reliability of the back office are an asset. This efficiency has helped the company develop in parallel in the ‘B2B’ (Business to Business), trade between companies, by becoming a platform for selling second-hand goods. There, the Amazon model, founded principally on the new, resembles that of eBay by opening itself to second-hand goods. This double dimension must not be underestimated because it gave Amazon a decisive support base among booksellers, resellers, and other small merchants who depend on it today. According to the famous words by Jack Dorsey, founder of Twitter, who had just launched his own online payment portal (an e-wallet called Square), ‘trade would firstly be a conversation’.

Warehouses and also clicks. Leader in online sales, Amazon is becoming a giant in digital content. With its Kindle, Amazon sells e-books and competes with Apple tablets; with Amazon Prime Instant Video, the Seattle company contends with Netflix in the demand for video streaming; with its Kindle Owners’ Lending Library, it offers lending e-electronic books, like in a real library; with Amazon Cloud Player, it offers an interface similar to iTunes to freely store music and buy it, and might soon switch over to unlimited streaming access based on the Spotify model.

And then there’s the Cloud, where Amazon has become, without putting too fine a point on it, an indispensable leader. Besides, it is this company of Jeff Bezos that, when it was launched in 2006, contributed to the popularity of its Elastic Compute Cloud (EC2). Today, its capacity for data storage, in more than one million servers, connected between them and in real time, is truly stupefying. Private individuals get free access to Amazon Cloud Drive to store documents, images, sounds and videos (the service is paid beyond 5 G). As for professionals, they use unlimited storage capacity, to the point that Netflix and Dropbox, even though competitors, have recourse to the servers of Amazon Web Services. The entirety of the Amazon back office was delocalized on the Cloud. And even the American security agency, the famous National Security Agency (NSA), now uses Amazon to store some of its data, as revealed in a recent book dedicated to Jeff Bezos’s company.

The spectacular infrastructure of Amazon, combined with those of Google and Apple, reveal one of the main characteristics of future culture. We have seen cultural products disappear and be replaced by cultural services: Amazon anticipates this evolution and has developed platforms in which these will be accessible in real time on the Cloud. Decline of CD and DVD sales, stagnation in the market of physical books, these have not escaped the notice of the company’s business model since 1995. If the usages swing to rental, instead of purchases, Amazon would be ready.

Platforms like Spotify, Pandora, Deezer and Google Play Music, All Access, among a number of other services which are on a roll, invest massively in the market of streaming music. Netflix offers the equivalent for cinema; and the sector of books will without any doubt also have its own Netflix where one can subscribe to an unlimited number of books (Scribd, Oyster and Entitle all already trying this, even if Amazon would be the leader of the sector). Over time, the GAFA thus risk being the only ones, given their international hit, their phenomenal storage capacity and their number of users, to be able to offer a global and transversal subscription service for all aspects of culture—music, cinema, video games and books—by unlimited subscription. It is a hypothesis for which we must be prepared.

At the heart of the Amazon model, there is a last recommendation. Thanks to its powerful, constantly refined algorithms, the online commercial site offers to its regular clients purchase suggestions based on their consumer habits and to its occasional client suggestions based on general sales and market trends (with the famous ‘You may also like’). Gradually, it is not just the readers who read the books, but also e-books that read the readers! These recommendation tools, already heavily used by Netflix for cinema, Spotify for music and Khan Academy for distance education, will be crucial for the development of subscription services for culture.

Is it not, however, possible to dismiss all these techno-optimistic hypotheses? Are our literary libraries, collections of discs and DVDs truly condemned? Is the Amazon model going to spread? Will recommendation and algorithms become the rule? To try and answer these questions, field survey and international comparison bring elements of response. In four countries that I visited, Amazon is called Flipkart, Ozon, Alibaba and Rakuten. Where we discover the good news that Amazon is not the only one; it is bad news that a model can become widespread.

Friday in India is ‘casual Friday’. Like in American companies, the software employees in Mumbai, New Delhi and Bengaluru dress in a casual manner. Amod Malviya wears a white kurta.

At thirty-two years, Malviya is the vice-president of Flipkart, one of the biggest Indian e-commerce companies, a kind of local Amazon. Wearing this long shirt, very popular among Hindus and which otherwise reflects belonging to a caste and a certain sense of tradition, he is affected by my comparison with the Seattle company. ‘We are very different from Amazon. We are not at all a clone. Our link with suppliers and the products that we sell, including a wide range of kurtas under “ethnic clothes”, and the manner in which we deliver the orders is very unique.’

Like in China and Russia, Flipkart functions on a ‘cash on delivery’ system. Only a small percentage of Indians have a bank card; the strong distrust vis-à-vis e-commerce and the difficult delivery conditions explain this preference of consumers. Another difference: the problematic state of B2B. Companies don’t know how to cooperate between themselves and it is not possible, for example, to interconnect software systems of suppliers with those of Flipkart, that which Amazon was able to implement in an efficient way.

For all this, Flipkart started like Amazon by selling books, CDs, DVDs and other video games, before extending to software products and clothes. Amod Malviya has his eyes on his American contenders, Amazon of course, but also Facebook and eBay, though he seeks firstly to understand the specific needs of the Indian public. Do rules of e-commerce apply everywhere or are they particular to each country? According to Malviya, there is no doubt that the failure of Amazon in India, which couldn’t adapt itself to the specificities of the local market, confirms territorialization of the online sales.

The site Flipkart also innovates, inspired by social networking sites, to create a kind of ‘social commerce’ (or s-commerce), which depends on India’s favourite social network sites, even if they are essentially American, to sell their products. Between global and local, India thus makes complex round trips. ‘At the moment, we are looking to diversify our products and to better serve our customers, rather than expanding our network of distribution,’ concludes Malviya. With more than 5,000 employees, Flipkart is an Indian giant which is growing strong. But it is also a national particularity, only capable of delivering its products in 130 cities, and almost never, taking into account the disastrous state of the postal network, in the 600,000 villages in the country.

‘The Russian economy is an economy with a lot of cash,’ says, in a factual tone, Maelle Gavet. She is the CEO of Ozon.ru, the principal e-commerce site, a Russian Amazon. ‘For long, e-commerce was not trustworthy in the eyes of Russians. They looked up products online and went and bought them in the shops! We needed to build confidence. We started with books and CDs and then extended to household products, toys, beauty products. Gradually, we became credible.’

Gavet notices, however, three singularities that makes Ozon a site particular to the Cyrillic world. Firstly, payment in cash, almost always on delivery, and not by card during purchase (except for the travel site ozon.travel, which resembles Expedia a little). And then, delivery, which rendered necessary the implementation of a proper model of distribution. ‘We absolutely cannot trust Russian Post. We had to create our own system, Ocourier, a kind of Fedex or DHL.’ And finally, the after-sales services are very important and that need particular attention. ‘Russians need to be reassured, we must clearly respond to all their questions when they are installing a device, when they change it, or when they bring it back if it is not functioning,’ explains Gavet. According to her, Amazon did not succeed in establishing itself in Russia because ‘their site was not able to adapt to local rules regarding cash payment, distribution and SAV.’

Ozon is interesting in its rapport with Amazon because of this subtle dialectic between the original and the copy. Other sites share this schizophrenia: KupiVIP is a copy of the French site Vente Privée and pinme.ru resembles Pinterest to the point of confusing one for the other. Yet, there exist authentic national leaders such as exist.ru which sells separate parts for cars (first Russian site in terms of audience) and holodilnik.ru, specializing in household products (one can translate the name to refrigerator.ru).

In Russia, one can thus see that all sites domesticate themselves. A good example is Vkontakte.ru (or VK.com) which was initially a kind of a Facebook clone and which distinguishes itself completely today. When one looks at this social networking site, whose name can be translated as ‘in contact’, one is struck by the design and colours which resemble Facebook. It however has its differences: starting with the flexibility on the questions of copyright which has for long manifested by the possibility, if not encouragement, of massive illegal downloading of cultural content. This Russian Facebook with 250 million users, evolves today towards a model of legal streaming, by multiplying partnerships with Russian creative industries, and towards online subscription. In its way, Vkontakte encroaches upon Ozon’s territory, which unlike Amazon, did not go in this direction. The Russian Facebook could become Russian Amazon with subscription and streaming.

For its part, Yandex is the main Russian search engine and it reminds one strongly of Google. But the comparison stops there. Firstly because Yandex was developed well before Google, initially for looking up verses in the Bible. The leader site has also liberated from the American model by sticking to Russian expectations. For example, it proposes a very popular cartography service entirely dedicated to Russian road traffic and traffic jams (Yandex Traffic Jam). Finally, its success in terms of search in the order of 60 per cent of the market share greatly surpasses that of Google in Russia (less than 30 per cent). And like Amazon and Vkontakte, it plans to diversify its cultural supply by streaming or subscription.

Localization is therefore indispensable, whether it is for the Russian branches of American giants or for the local clones of Facebook and Amazon. However, sites like Ozon in Russia, Alibaba in China or Rakuten in Japan also confirm that there is an Amazon model which can be found everywhere. Though the local variants are immediately striking to an outsider, they simply attest that localization is always indispensable, whether it is through the choice of the products offered or the methods of delivery proposed. More than the rest of the internet, the online commerce sector remains very territorialized and it will remain so. For clothes or for food, or for getting products of everyday life or products of comfort, it is necessary to be able to get them delivered home. Neither the warehouses nor transport can be global.

This territorialization does not exclude general tendencies. Often, the matrix was cultural, firstly books, CDs and DVDs, before diversification. Everywhere, these platforms of cultural goods are in the process of evolving towards services and subscriptions for streaming. Having precociously invested in the online music sector, Rakuten currently develops by opening an immense ‘online video club’ for cinema (the Japanese site also has a downloading platform for British music, play.com). And like Amazon, Rakuten wants to rely on its search engine and its algorithm to propose more suitable recommendations for its clients. The objectives of the e-commerce sites aim to bring closer prescription of the transaction. Over time, this prescription and critique will be made by the sites themselves. In the digital age, cultural goods are not the only things to disappear: specialized journalists and critics will too.

Recommendation

Cultural Critique is dead! Long Live Smart Curation!

At the entrance of the Headquarters of Gawker in New York: a large flat screen showing the best users of the site in real time. The articles are segregated by the number of views they have garnered and, as the traffic constantly changes, the Big Board, as the screen is called, flickers in all directions. The Gawker sites have, on an average, 100 million individual visitors every month. That makes the number between 3 and 5 million every day!

‘It’s fucking high,’ says James Del, when I meet him on the last floor of an industrial building on Elizabeth Street, in SoHo Square. He isn’t talking of the users of these sites but of the dizzyingly high staircase and its exaggerated steepness even by New York standards. We are only on the third floor but, without a lift, the particularly steep set of stairs renders this visit adventurous. Inside the large, open-plan office, the atmosphere is more horizontal. There is a café to take breaks in, and of course, a Ping-Pong table, the symbol these days of every American start-up. The employees have earphones on, to keep to their bubble, and chat with their colleagues on instant messenger, to appear to be a team. The ‘no hierarchy attitude’ is the norm, even though the jobs are meticulously compartmentalized. One floor is dedicated to journalists, known here as ‘editors’ or sometimes ‘curators’; another is set aside for sales; a third for engineers, accountants, lawyers and customer support. Thus, the employees in charge of content—almost half of the 300 employees—and those in charge of advertisement and data never quite meet, except when they climb up the difficult stairs or clock in with their electronic cards at the entrance of other floors. James Del, who has access to all the floors, constantly goes up and down the stairs and this visibly irritates him. He came to Gawker at 20, upon leaving New York University and year after year, has gone up and down the stairs of this now major start-up. At 28, he is now the vice-president of Gawker, taking care of programming and marketing.

‘Our model is based on “curation”. We read everything on the internet and when a story seems interesting, we take an idea, an information or a paragraph, and we write an article and create a buzz,’ Del describes ingenuously. Sitting in one of the ‘meeting-rooms’ of Gawker, he has his coffee mug placed on his table in front of him. I am in the office one of the main media people in the United States, a fun and cruel site, regularly lambasted by its counterparts for being ‘creepy’, ‘trashy’ and ‘sketchy’! Del, however, speaks professionally, efficiently, seriously and almost too wisely. He has this disarming way of accepting the criticisms against Gawker, its excesses, its futility, its voyeurism and its borderline attack on privacy and its cynicism. He has heard it all; and he is no longer touched by it. Of overarching vision, editorial moral, respectability, he seems to say, ‘That’s not what the web is about!’ On the contrary, sharing, participation, curating, algorithm and recommendation form the heart of his job. Del points out that the readers of the site can themselves participate in its composition, through a dedicated platform, a meta-blog called Kinja: ‘Everyone can publish on it. This helps to bring out new content and new talents, by shattering the traditional line between articles and readers.’

Partly curating content produced by others, partly creating original articles with lots of algorithm, Gawker is specialized in all that is buzz, cool and hip, with a lot of media gossip. Its baseline, proudly displayed on the site, reads: ‘Today’s Gossip is Tomorrow’s News.’ When they are not robots, the writers are cool hunters, trend-setters and taste-makers. ‘Our main aim is to be of service to our users,’ says James Del.

It is in 2002 that Nick Denton, an English journalist living in New York, launched Gawker from his apartment in SoHo. Like in the Bright Lights, Big City, the cult novel by Jay McInerney, Denton is fascinated by the private lives of the New Yorker journalists and the other magazines of Condé NastVanity Fair and Vogue. Perhaps, as he was unable to make it to the editorial board of these ‘seniors’, he creates Gawker to talk about what happens backstage in all these places. His first ‘confidential’ pieces on his favourite punching bags—Tina Brown, head of Vanity Fair who later became head of New Yorker, or Anna Wintour, the star-editor of Vogue—hit the bull’s eye. The tone is cynical. Infiltrating, along with his interns, the Condé Nast Tower which was then situated at 4, Times Square, he can talk of the food habits of the journalists, their dress code and says that they wear sunglasses in the elevators. Importance is given to stories taking place in the cafeteria designed by Frank Gehry, where Meryl Streep has lunch in the film The Devil Wears Prada.

From the beginning, Gawker is a mean site but now it’s becoming meaner by posting compromising photos taken at private parties and revealing the exorbitant contract rates of the entertainment industry. ‘We have a very good legal team,’ says James Del, innocently.

The establishment is repulsed by these unethical practices but it rushes to read the site, nevertheless. ‘I am not proud of it, but I look at Gawker every morning,’ admitted Chris Anderson, the founder of Wired (which also belongs to Condé Nast), in an interview. And one day when I interviewed Tina Brown at her home in the Upper East Side, she pretended not to be interested in such Gawker gossip, calling it ‘New York fakery’.

On its official introductory page, Gawker defines itself as the ‘one-stop guide to media and pop culture’. The site claims to reveal to the readers information that is more or less private, the off-scoops and secrets of stars, media and the entertainment industry. However, Gawker has gone beyond this initial aim and become a media group, a chain of sites and a true franchise.  The economic model here is both horizontal and vertical. Each nano-site is given its autonomy, identity and its verticality; but all its users are taken in horizontally when it comes to selling an advert. From the main site (gawker.com), users can reach a dozen sub-sites and sub-blogs, niches specializing in Hollywood Celebrity news (Defamer), secrets relating to Washington politicians (Wonkette), the backstage of Silicon Valley (Valley Wag), the sports industry (Dead Spin) or porn (Fleshbot), television (Morning After), video games (Kotaku), the army (Fortress America), science-fiction, geek-info and other electronic gadgets (io9, Gizmodo and Lifehacker). It even projects introducing a mini-site dedicated to opposing the Tea-Party Movement. A sub-blog is specialized in fact-checking the rumours circulated online. And another, Gawker Review of Books, is dedicated to the publishing world and literary criticism—as though it is a literary supplement of the New Yorker.

The ambition of Nick Denton, the founder of Gawker, is apparently to have twelve sites, according to a funny rumour. That is as many sites as Condé Nast has magazines.

Abundance. ‘Every 48 hours, we create as much content as was created between the advent of humanity and 2003.’ This statement, often repeated by Eric Schmidt, executive chairman of Google, shows that abundance is one of the principal characteristics of the internet. This profusion translates into streams and currents, at the risk of inducing logorrhoea. When culture, which until yesterday was made entirely of ‘cultural products’, goes digital and becomes a sum of ‘cultural services’, some guidance becomes indispensable. How else can one distinguish oneself from Spotify which has more than 25 million titles? How else can one locate oneself amid the 300 hours of videos uploaded to YouTube every minute? The supply is unlimited to the point where it is dizzying. The expression ‘all you can eat’ is many a time symptomatic. Like in Chinese restaurants where such menus are offered, one can eat to the point of satiation – and sometimes to the point of indigestion.

In their way, the curators at Gawker are supposedly helping users to navigate this abyss of information. These new-media critics sort, select and orient the information. On the official blog of the site, one word appears many times and reiterates the mission of Gawker: recommendation. Of course, their journalism takes a particular route: it follows media errors, secrets, the bustle around town, but it announces, according to James Del, the future of internet and digital prescription.

Gawker comes in the new generation of pure-players revolutionizing journalism and who are largely dedicated to viral information: BuzzFeed, Vox, Vice, The Daily Beast, Upworthy, Mediaite, Twitchy, Dose, GivesMeHope, OMG Facts, Reddit, Mashable, or to a different degree, Huffington Post or Slate. As rightly or wrongly predicted by Eric Schmidt, the media are becoming aggregators, filters, content providers, counting increasingly both on curation and validation to ensure credibility of the information, and on first reporting.

Often, the journalists here are replaced by curators, data analysts, data scientists, or even ‘Chief Aggregator of Viral Content’ and ‘Chief Trend Hunters’. Sometimes, it is simply algorithms that do the work of humans. In some cases, such as Dose, GivesMeHope or OMG Facts, three sites belonging to serial entrepreneur Emerson Spartz, their principal media innovations are not so much concerned with producing content, as they are with the manner of selecting from among the pre-existing articles and curating them by repackaging them, before they are promoted on social media. This cannibalisation is also key to Gawker’s success.

Bill Keller, former head of New York Times, had this killer-statement to make about new media in general and about Huffington Post, in particular, that they ‘found out that, by putting together gossip on stars, videos of cute cats, blogs by unpaid journalists and information borrowed from other publications, and adding to all this a left-wing tone, millions of people would read these sites.’ The person that this notion targeted, Ariana Huffington, nicknamed ‘the queen of aggregation’ indirectly replied to him in her own newspaper, the New York Times: ‘I didn’t kill the newspapers, darling. New technology did.’

Curation and aggregation technologies vary, but they generally aim at creating viral posts which influence the audience, thus bringing in advertisement revenue. The New York Times journalists, despite the remarks of Bill Keller, have already addressed these constraints of the internet: they change the titles of articles so that they are more easily read by the Google robots, contain links and keywords in their texts and have recourse to professional services to ensure that they go viral, increasing the traffic of their messages on social media.

In certain places, web-journalists are taught to write in short sentences using more periods and fewer commas: commas and colon marks are to be proscribed, says Emeron Spartz. Writers everywhere adopt Search Engine Optimization techniques (SEO) to help search engines read their content. A title comprising of keywords (chosen with a keyword optimization tool) for example, is better indexed by robots, just as an article with outgoing links. A title in the form of hyperboles, with past participles and ending in prepositions, is susceptible to quickly go viral. A title written to tell a story is even better. At Doze, catchy lists and titles are multiplied; at Upworthy, teasing titles with hashtags are created to make the reader’s mouth water; at Gawker, users are induced to share content on social media; at Reddit, provocative photos are given priority. All these techniques are conceived to please Twitter threads and Facebook newsfeeds. (The traffic on sites like Gawker, MuggleNet, OMG Facts, GivesMeHope largely depend on the traffic generated by Facebook).

These secrets of titling and indexation still depend on human labour. But it is possible to go further by entrusting this work to robots. On Mediaite, an algorithm aggregates content from 1500 media outlets every morning and once curated, they are proposed to the users. On Reddit, an algorithm automatically compiles the likes, votes and preferences of the users, and determines its welcome page accordingly.

At Gawker, explains James Del, user comments are selected by a new algorithm called Powwow, which organizes the conversation without human intervention. Sometimes, like at OMG Facts and Dose, a ‘Headline teasing’ algorithm is used: when an article is created, it is randomly proposed on twenty different posts and platforms with different titles. The Headline-teasing algorithm then compares the buzz created by the articles, the number of clicks and their speed of diffusion. A few hours later, all the titles are automatically changed to the one that has gone most viral.

These major evolutions of web-journalism necessarily affect the culture of critique, which in turn evolves towards ‘likes’, ‘blurbs’ and ‘Tweets’. But can one really give one’s opinion on a book or a film in 140 characters? When recommendation on Twitter becomes a kind of blurb, those small, positive quotes from journalists used to promote a bestseller or a blockbuster, it becomes the degree zero of critique, according to the many journalists who were spoken to. ‘Twitter is indeed a concise media,’ notes Antonio Velazquez, blogger, hacker and co-founder of the site Horizontal in Mexico, before adding: ‘We can tweet web links which redirects to the long articles.’

Nevertheless, on social media, critiques are increasingly becoming publicity slogans: ‘The Best Family Film This Year’, ‘Holiday Classic’, ‘Wow!’, ‘Absolutely Brilliant!’, ‘Hilarious!’, or the more frequent ‘****’. On the internet, it is the victory of ‘Two Thumbs Up!’, a system to judge films, conceived by critics Robert Ebert and Gene Siskel on their channel At the Movies on ABC. They evaluate a film with their thumb, and only three judgements are possible: 2 thumbs pointing upwards if both critics like the film; two thumbs facing downwards if they don’t; one upwards if they are divided. Thus, the user knows whether the film is a ‘must-see’ or a ‘turkey’. After the demise of Siskel and the retirement of Ebert, the show was taken up by two journalists, one of whom is none other than the head of the cinema service of New York Times, A.O. Scott. He now holds his thumb up or down!

WHEN I CLOCK IN AT THE ENTRANCE OF THE World Trade Centre One (WTC 1), on the Northern side on Vesey Street, the lift identifies and takes me directly, with no other stop, to the floor of my meeting: Floor 34. This is where The New Yorker and the entire Condé Nast group (including Vogue, Vanity Fair, GQ, Wired and the site Reddit) moved to, in the autumn of 2015.

To get access to the most famous New York magazine, no need to climb stairs, but one does need to prove his credentials. Built on the ruins of the Twin Towers destroyed during the 9/11 attack, the World Trade Centre One is not accessible to public. The works are yet to be finished. It is necessary to furnish one’s identity proof beforehand, pass through many levels of security and be meticulously photographed.

‘The problem with media is that of discoverability,’ begins Henri Finder, sitting in Room no. 1, on the 34th floor, with a spectacular view of the monument dedicated to the victims of the World Trade Centre, the south of Manhattan, and further away, the Statue of Liberty. Editor in Chief of the New Yorker, Finder personifies this elitist magazine which defends a critique of culture ‘with standards’. Its journalists see their mission as that of a gate-keeper’s. Finder himself is a bit of a prohibited intellectual, both sophisticated and low key, his discretion erected in the art of living, but who can, suddenly, demonstrate an exhilarating, off-the-wall humour. In the American slang, one might say he is ‘book smart’—intelligent and very, very, very well educated. He doesn’t have a Twitter account.

‘How can one make a good article known today? This is the serious challenge of “discoverability”,’ says Finder, employing the fashionable word. Faced with the abundance of content online, the question of access and selection becomes crucial. We can, of course, resort to counting on chance: it’s what we call serendipity or finding content without having searched for it (a word borrowed from a Persian tale, used by Voltaire, meaning ‘happy chance’). Conversely, if we believe that chances are rarely good, one can then trust algorithms and automatic selections. We can also go the more traditional route, if we are afraid of algorithms, and trust the traditional critiques—like those of the New Yorker.

The New Yorker of William Shawn, in the 1950s, was on a mission to defend ‘Art’ and protect ‘Culture’—in capital letters—that is, stand up to barbarians who want to abolish the cultural hierarchies between ‘High’ and ‘Low’ culture and demolish, by the same stroke, the line separating taste from mediocrity, the elite from the masses, culture from entertainment. The New Yorker of Pauline Kael and her renowned critique of cinema in the ’60s and the ’70s, then preferred—following a counter-intuitive method that has since ensured the success of the newspaper—taking popular culture seriously and write on ‘high culture’ in popular language. Tina Brown’s New Yorker, in the 1990s, finally became convinced of the suspicious nature of the European cultural hierarchies and decided to permanently blur the lines. Mixing genres has become the rule: it critiqued Metropolitan Museum and Star Wars, Shakes and the Monty Python, the novels of John Updike and, on twenty pages, the AOL-Time Warner fusion. ‘I started a chronicle called The Annals of Communication to follow the major evolution of studios, television and, particularly, the entertainment industry,’ said Tina Brown, during an interview with me for my book Mainstream. ‘The New Yorker must speak of that which the people are speaking of,’ she added, in an obvious tone.

For Henri Finder, who today embodies the New Yorker in digital disruption, the barbarians are now called click, blurb, algorithm, recommendations of Amazon, Gawker and maybe even more so, the Gawker Review of Books!

‘Sites with large traffic have less and less critiques. Those with fewer audience stick to niche critiques. That’s why the New Yorker is still relevant,’ says Finder. He believes in sustainable journalism with a ‘long critical trail’, beyond acceleration and buzz, with recommendation for a long length of time. The New Yorker model is virtuous on the editorial and journalistic levels, but online, it is confronted with two problems like all other paper media: firstly, the economy of the ‘long tail’—fewer sales in short term but sustainable and durable sales in the long run—in the end, bring very little money; secondly, the cost of production of a New Yorker article is astronomically high compared to a Gawker post, while they both have the same readers online and generate same advertising revenues. (Some whisper that Gawker’s ‘publishers’ are frequently freelancers who are paid a dozen dollars per post, plus a bonus depending on the audience of their article).

‘There are fewer and fewer critiques and fewer and fewer literary supplements,’ says Finder desolately. ‘We must therefore count more on critique.’ He adds: ‘We must remain who we are: that is what our reader expect and value. Our most-read articles are often the longest ones. But we must also be part of the conversation, which is about being present on all the digital platforms. We should be everywhere, but we must remain the New Yorker.’

Traditional editors shared this state of mind. ‘I have no fear of algorithms! I am ready to duel any algorithm whenever you want!’ says Jonathan Karp, the CEO of Simon & Schumer, one of the biggest American publishing houses. In his vast office at Number 1230 on the Avenue of the Americas, New York, he brags about his desire for battle and flexes his muscles. But he admits, a little later, to having directed his editors and authors to increase their presence on social media. ‘We saw that Facebook is a great tool to bring attention to a book,’ says Karp, fascinated, as though he had just found out that string cuts butter! And he cites novels from his publishing house launched almost exclusively on social media (such as We Are Not Ourselves by Matthew Thomas). On the Simon & Schuster site, buttons to share, like and retweet information on books are as visible as on the Gawker site. As for YouTube, Jonathan Karp swears to me that a dedicated team from his publishing house is on it. Simon & Schuster is all set for the duel with the internet.

The crisis of the critical function. Although the New Yorker, with its current 1.1 million subscribers, does not seem threatened in the short term neither by blogs and social media, nor of course by sites like Gawker, the crisis of the critical function is all too real.

Traditional prescription has not disappeared but, everywhere, journalists with whom I spoke admit that ‘something is happening’. Internet by nature brings about the end of hierarchies, disintermediation, decentralization, and disappearance of elitist legitimacies—evolutions which unavoidably affect critique. We are entering a culture characterized by ‘conversations’ and not by arguments of authority, a culture where recommendation is becoming central, but where the prescribers are also multiplying in number and infinitely. Legitimacy on the internet no longer depends on social status, degrees or knowledge acquired, like in the world of print, but has new criteria such as e-reputation, popularity, community to which one belongs or that one is surrounded by. The ‘top-down’ hierarchy of traditional critique is suffering everywhere. This is the great disruption of hierarchies.

An evolution that blogger Antonio Martinez, spoken to in Mexico, welcomes: ‘The traditional media have moved away from their mission and their social role. They stopped talking to the citizens. Criticism has become cynical, remote and opaque, almost deadly. But information wants to be free! And the explosion of social media, the growing influence of the hacker culture and whistle-blowers are changing the game. A new critique is emerging and will challenge the media system as a whole.’ Patrick Beauduin, the Director General of Radio Canada, interviewed in Montreal, is less radical but he also thinks that the future of critique lies in recommendation: ‘An audio-visual group such as ours will no longer be a broadcaster but a prescriber, who guides and gives an opinion. We will be curators.’ Antonio Martinez Velazquez even believes that the new prescriber will the hacker. The hacker plays, according to him, the same role in the web culture as the black kid or gay cool did in pop culture. He invents a new cyber culture of rupture which he himself fabricates with code—and then changing the codes. The hacker becomes the influencer and prescriber. In one word: a hipster.

Already, film critics seem to no longer have much influence on the box office which they did before and literary supplements no longer sell books

(Different qualitative studies on the reading of the press show that between 85 per cent and 90 per cent of readers do not even open the literary supplement of a daily newspaper). In the United States, critiques often tend to disappear due to the lack of support from editors, who do little advertising around it and would rather prioritize cooperative agreements with Amazon. Though the literary supplements of the New York Times and the Wall Street Journal remain influential, it is not the case with other publications. ‘Book World’, the Washington Post supplement, created in 1967, ceased to be an autonomous weekend supplement in 2009 (it has joined the ‘Style & Arts’ section of the print version); the ‘Book Review Section’ of the Los Angeles Times was also fused together with the running edition of the paper in 2007; as for the San Francisco Chronicle, its supplement was reduced from six to four pages in 2006.

On the internet and social media, ‘populism is the new model of the cool; the elitists are the new nerds,’ predicts journalist Alexandra Molotkow, in a New York Times article. The problem, however, is deeper, and, in some ways, more serious. A series of fundamental mutations that are underway may permanently transform critical function: the recognition of a visual culture that softens the power of writing; the development of accurate audience measurements that reveals the invisibility of critics; the abundance of internet that requires a ‘filter’; the ‘long click’ which upsets the order; finally, the unlimited cultural subscriptions that are made possible by the rise of the cloud.

The first mutation is the blurring of the distinction between media, the blurring of borders. Digital acceleration translates into a change in shape, formats, and temporality. Alistair Fairweather, who runs the site of the South African magazine, Mail & Guardian, explained this upheaval to me when I spoke to him in Johannesburg: ‘What is funny is that we are a weekly, but on the internet we become a daily newspaper.’ One can also see that this mutation and this acceleration of the media have multiple and contradictory effects: books are transformed into short essays; essays, into open columns; columns into blog posts; and blogs, into Tweets. In image and sound, the radio becomes the podcast and the podcast becomes streaming; television is changing to connected screen, SVOD or Netflix; as for MTV, today it has become YouTube!

Alistair Fairweather also notes that the websites of television channels, radios and newspapers are similar on the web: they have text, images and videos. ‘It is as if all media have merged in their form on the internet,’ says Fairweather. On smartphones, the rapprochement is even more noticeable, the sign of a culture ‘becoming more and more visual’. According to Fairweather, who has long been a film critic, these developments have major consequences for cultural criticism. First, the ‘finished’ article is disappearing: it is constantly updated and corrected. Second, although titles and reference brands remain, readers are interested in articles more than in an entire newspaper. ‘The key is that they are choosing. They still trust brands like the Mail and Guardian, but they do not try to read a single newspaper anymore. The power of writing is eroding on the internet where it competes with other media, other modes of expression, and other kinds of legitimacy.

The second major change is the possibility of knowing the precise readers of articles. ‘The biggest change the internet is making in the media is not immediacy, or lower costs, it is measurability. And it is scary if you are a traditional journalist,’ said Nick Denton, the founder of Gawker (in a long profile dedicated to him by … the New Yorker). However, these precise audience measurements tend to confirm what we already felt, without being able to prove it, namely, the few readers interested in the criticism. From now on, each article having its statistics of reading and its number of page views, one discovers dumbfounded, in the newsrooms, the infinitesimal audience of the critics of dance or opera and, more generally, cultural critics which are rarely read. A whole new hierarchy of information is being put in place which excludes home page critiques because of their unpopularity, as they are suddenly becoming more invisible. (On Gawker and many other sites, readers can see the number of views of an article).

The third mutation underway, linked to the unlimited abundance of internet, the saturation and, sometimes, to the frustration it generates, is the need for ‘filters’ In the old critical world, cultural journalists were dealing with relatively rare cultural products. There were fewer artists, fewer novelties and they had to be ‘found’. Film and documentary releases were limited to a few dozen a month. But today, their number has increased tenfold on YouTube, not counting television series or video games, which are at the heart of creative innovation. Novelties in music are literally endless on Spotify, Deezer, iTunes, Apple Music, Music Key (YouTube), Prime Music (Amazon) and Soundcloud, as well as on Taringa (Argentina), Xiami (China), MelOn (Korea), Saavn (India), Anghami (the Arab countries), and even on what remains of MySpace (which is in decline since its purchase by Rupert Murdoch and then by Justin Timberlake). Television programs have multiplied exponentially while everyone can become their own live video broadcaster and be streamed on Meerkat or Periscope (which belongs to Twitter). And we can think that new authors will appear, self-published, on platforms like Scribd or the Amazon Kindle—perhaps these are the future YouTube of books. Even before exercising his function of recommendation, the critic must therefore sort through this unlimited offer.

He can choose, of course, to stay out of the digital culture, but will miss most of the current and future works. By limiting himself to the Western ‘canon’, elitist or academic, he deprives himself of commenting on essential parts of contemporary culture. And, if he decides to switch to the digital, he will therefore resort to functions of ‘filters’ before being able to exercise his mission of ‘recommendation’. Without technology, there is no possibility to find—and no choice. But to resort to machines already means switching to the new critical function.

‘The Long Click’ or the end of critique as we knew it

MOST NET GIANTS now use web-based recommendation algorithms: Amazon (with Amazon Prime), Facebook (with its EdgeRank-driven newsfeed), Apple (with iTunes Radio and Apple Music), or Google (with Google Play and Music Key). Prescription algorithms are also at the core of cultural service models in streaming, such as Netflix for cinema; Spotify, Deezer and Pandora for music; Steam and Twitch for video games; and Scribd and Oyster for books. (It should be noted that these unlimited ‘independent’ streaming services are based on the cloud most often and on Amazon Web Services, the main cloud that popularized the expression in 2006 and democratized its use- the solution chosen, for example, by Netflix ... nevertheless, a competitor of Amazon).

Thanks to powerful algorithms that are constantly refined by user behaviours—the so-called collaborative filtering—sites and applications offer their subscribers suggestions based on general sales and market trends and on their consumption habits (this is the famous: ‘You may also like’ from Amazon).

With the mutation of the traditional critic, some think that algorithms, combined with big data, will become the cultural prescribers of tomorrow. According to some of the most positive prognostics, they can even provide internet users, recommendations that human beings were not able to do without software or computing power. The movements of opinion can be perceived, artists who are having a breakthrough are being spotted, and special niches are being discovered.

This revolution was made possible by the invention of the ‘cookie’, a mini-program that helps recognize the user: sites use this ‘cookie’ to keep in memory the history of his activities. This major innovation, created in 1994 by a Netscape engineer, Lou Montulli, helps feed most customization algorithms. We are talking about ‘Long Click’ to summarize this change: when we surf the internet, by simply clicking on a content, ‘liking’ an article, or ‘retweeting’ a message, we contribute, without knowing, to feeding different algorithms. They then retain the user data and can deduce the user’s centres of interests or his preferences of online purchases. In return, the algorithms helps customize their responses and bring in target advertising, which has paved the way for all the recent spectacular innovations in online advertising, social marketing, and real-time bidding.

Advertising is currently undergoing several major revolutions that intersect with the ‘curation’ debate: with Google, the advertiser pays the search engine only when the user clicks on a banner, such as AdSense, AdWords or DoubleClick. On YouTube, the model oscillates between spots called ‘pre-rolls’, broadcast before the free video, or ‘stream ads’, when ads are mixed with content. RTB or real-time bidding models are automated real-time advertising auctions, for example, along with Metamarkets or MoPub, owned by Twitter. At the same time, advertising increases its relevance on social media by being simultaneous, personalized, contextualised and ‘customized’. An advertiser can ‘buy’ advertising space, but he can also ‘broadcast’ the ad itself or ‘create’ it, which is new. There are three forms of advertising: ‘Ads you buy; Ads you spread; Ads you create’. Geo-localization accentuates these mutations at the risk of making the cookie a more formidable spy, whose performance is increased tenfold by performances of mobility and applications, each smartphone having a perennial identifier based on its telephone number. Finally, a whole stream of advertising, and ‘Native Ads’, is trying to enter into ‘conversation’ with the user. The new mantra of Twitter becomes, for example: ‘Think content, not advertising’. This is the typical template for ‘Sponsored Tweets’ and other ‘Promoted Tweets’. In advertising in the digital age, one should not only think in terms of advertising, as in the days of Mad Men, but also in terms of content and conversation. We are moving towards true ‘conversational’ advertising.

ALGORITHMS OF RECOMMENDATION are not, however, infallible. They are not necessarily ‘accurate’ either, as there exist distortions of recommendation.

There is, first of all, the problem of their manipulation for commercial purposes. Little is known about it, but as revealed in a survey by George Packer in the New Yorker, the American giant Amazon uses on its site a highlighting system called—a fine euphemism—‘Co-Op’. These famous ‘cooperative advertising agreements’ become ‘recommendations’, presented as the result of the objective criteria of an algorithm, whereas they are publisher-funded advertising. This manipulation, unknown to customers, may explain in part the relative inefficiency of Amazon’s suggestions that lead to fewer purchases than is normally said. (Amazon does not offer this ‘pay-for-display’ system in the form of advertisements but it negotiates fiercely with the publishers, outside the anti–competitive laws, an additional percentage on the sales gained this way—from 3 to 5 per cent—which can in total come up to, for example with Random House, 53 per cent of the price of a book).

A second type of recommendation distortion comes from the very limits of the algorithm. ‘Machines’ excel in all forms of measurement, aggregation, usage statistics, collating consumption histories, satisfaction evaluation, mass data processing, etc., but find it difficult to predict behaviours and tastes, to choose, and what is more, to judge emotions and sensibilities. Unbeatable when it comes to finding the best price of a plane ticket or a hotel room, they anticipate badly, contrary to what is sometimes believed, when it comes to the cultural prescription.

One cannot but notice that on Pandora, Spotify or Deezer, that after listening to a certain type of music for a long time—soul for example, as was my experience—the algorithm tended to indefinitely recommend the same musical genre (for me, it was uninterrupted titles of Motown, Atlantic or Stax, without the possibility of switching to pop or classical music). The algorithm does not innovate, venture out, imagine, or associate ideas, leading inexorably to increasingly narrow recommendations. (This phenomenon, known to econometricians as ‘attrition’, tends to reduce choices, especially if the user does not make his opinion known. To ensure diversity, engineers then, from time to time, artificially enter random musical proposals to elicit user reaction and allow the algorithm to be renewed).

In general, algorithms fail to perceive the expectations of users who have eclectic tastes, the so-called multi-genres, or tastes that evolve constantly. Thus, one might want to listen to such music, hip hop for example, in the morning when waking up, but prefer pop at work, salsa while driving a car and finally ‘smooth jazz’ in his room while going to bed—so many situations, sensibilities, moods that the algorithm cannot evaluate.

The case of classical music is even more illuminating. Spotify and Deezer are firstly criticized for their chaotic presentation of titles: it is indeed very difficult to listen to the songs of an opera in the right order, or Beethoven’s symphonies one after the other. As for choosing one’s performance, one’s soloist or one’s conductor—Martha Argerich or Sviatoslav Richter on piano, Furtwängler or Karajan to lead Wagner—it’s a challenge. It is the fault of metadata, often presented in a very insufficient way, but also of the algorithms which do not know how to classify classical music well or to evaluate it. ‘These algorithms do not work. They do not process enough data, they work with sample or with forms of serendipity. Those who program them do not understand the singularity of the classical music. Above all, they do not curate enough,’ explains Klaus Heymann, the international founder of the portal of classical music and jazz, Naxos (interviewed in Hong Kong). For particular niches, Heymann is convinced, curation will be decisive. And we will need, for classical music, as for the other genres, according to the New Yorker music critic, Alex Ross, true ‘aesthetics and ethics of streaming’.

We find these distortions of recommendations with Netflix and Amazon. Even when the criteria are refined, and the technique improves and the algorithms begin to ‘learn’, the results remain quantitative, not qualitative, and ultimately impersonal. They produce what is called in the jargon ‘noise’, that is, irrelevant data and content. For one thing, it comes from the fact that ‘data mining’ works based on samples, because of the lack of computing power needed to process all the data collected. The results are permanently corrected. This makes the recommendations, already complex to propose, even more precarious. ‘Filters’ are getting better, keywords and metadata are being refined, automatic aggregation is making progress, but the results are still disappointing and ultimately not very personalized—they are without flesh or aura.

There is another problem here—the limitation of the ‘filters’. Even when deliberately selected by users, filters produce ‘noise’. This partly explains the failure of RSS feeds like Google Reader, and ultimately, the likely exhaustion of podcasts. In all three cases, the recommendations come up against profusion. Due to lack of time, it becomes humanly impossible to consult all the content listed by the RSS feeds or proposed by Google Reader. The same fate may befall podcasts: you can subscribe freely and for free, but you do not have time to listen to them as they accumulate in the iPod or smartphone. In the ‘filter’, it would be necessary to add a new layer of ‘recommendation’.

‘The problem with Google Reader and RSS feeds is the amount of information,’ says Fairweather. ‘At first, we found it useful, but very quickly we were drowned by the quantity. Relevance diminished and the abundance became unbearable. The recommendation largely helps sort all this avalanche of content. And paradoxically, it gives the users an active function: make choices, no longer be passive. Recommendation can feed conversations and help make choices.’

There is an immense field of analysis of the distortions of recommendations on social media. The case of Facebook is revealing here. Unlike Google, Apple or Amazon, Facebook is the only net giant that does not want to sell cultural content directly. Instead: Facebook subscribers can make ‘recommendations’ to their ‘friends’ and Facebook business partners can make ‘suggestions’ based on their marketing agreements (Facebook’s preferred customers include Netflix for video, Spotify for music, Zynga for video games, Warner Studios, Miramax and Lions Gate for film, the Washington Post and Yahoo for information, Ticketmaster for concerts, and several other brands for travel and catering). If they have not verified their privacy settings on Facebook—complex, numerous, variable, elastic and, alas, often incomprehensible—, Spotify or Netflix users may be surprised that their music playlists or the movie choices find themselves posted automatically on their Facebook account, almost without their knowledge. In the end, social media can help recommend content related to commercial agreements at the risk of causing a confusion between prescription and marketing.

Other recent developments of Facebook cause distortions of recommendation. Like Google and its PageRank algorithm, Facebook has since 2006 implemented a ‘newsfeed’ managed by the algorithm EdgeRank. This initially included three criteria: the user’s affinity, the importance of the content and the time factor. Since then, EdgeRank has become more refined, now combining multiple criteria, up to 100,000 parameters constantly renewed and updated. Among these, mobility and geolocation are increasingly taken into account, as half of the one billion Facebook users access it today on mobile (and half its revenue also comes from mobile advertising).

More recently, Facebook has chosen to ‘editorialize’ the content of its users, especially the publication of ‘Pages’ (those managed by media, brands or personalities, unlimited in number of fans), but also to a certain extent, personal accounts (limited to 5,000 friends). From now on, when a media, a company or an artist posts a message on his Facebook ‘Page’, the algorithm allows him to reach only a tiny percentage of his own fans (around 5 to 7 per cent). Facebook deliberately limits the scope of posts and, via its algorithm, determines the dissemination of content either by its initial buzz (if the message is liked, shared or commented on, the algorithm extends the publication), or by the purchase of ‘sponsored’ spaces. In other words: to reach the friends subscribed to his ‘Page’, a Facebook user must now buy advertising. (On the other hand, when Facebook users ‘like’ a ‘Page’, they innocently think they are subscribing to it, while in fact they are essentially feeding the social media algorithm and its advertising model, but they hardly receive on their ‘Timeline’ the messages of the ‘Page’ they have liked). A similar algorithm exists for YouTube, Google+ and LinkedIn. On the other hand, Twitter has kept the publishing system that led to its success: Tweets are published without algorithm, linearly, in reverse chronological order and in real time, but for how long?

This commercial technique of Facebook highlights the insecurity that characterizes social media. For example, what is the point of an author, a publisher or a film producer building his community on Facebook if, without warning, the Menlo Park firm can nullify this investment and claim that they must buy advertising to be able to reach the fans of his ‘Page’? The same risk arises in the medium term for publishing content on Instagram, Pinterest, Path, Tumblr, Snapchat, Meerkat, Periscope, Vine, etc. In the end, the media, the cultural industries, companies and anyone who wants to increase the visibility of their content or their criticism on Facebook, are caught in their own trap of the ‘law of click’: staking everything on their visibility on social media, they must now pay to reach their own community. As a result, information or recommendations that go viral on social media are not necessarily those that are carried by a buzz, by the best ideas or by greater creativity, not even those that are best identified by the algorithms: sometimes, these are just messages sponsored by purchase of advertising space.

There is more. Little is known but social media, Facebook and Twitter, in particular, systematically use the content and recommendations of their subscribers, for their research, which is then marketed. For example, one can take ‘F-commerce’ for this marketing evolution of Facebook. It works in two stages: on the one hand, social media algorithms carefully analyze subscribers’ conversations, posted photos and videos, and incorporate personalized, ‘contextual’ and ‘native’ advertising. Mixed in with the threads of conversations, devious and misleading in their camouflage 2.0, bordering on mixing genres, these ‘native ads’, ‘sponsored Tweets’ and other ‘driven trends’ seem less intrusive while bringing in a lot of money (several hundred thousand dollars for twenty-four hours of exposure on all Twitter accounts in the United States, for example). The social media then sell the recommendation data of the millions of messages exchanged each day by their users. Advertisers, such as those responsible for the creative industries, media or consulting and audience measurement agencies, are looking for accurate data in real time, based on the interests of consumers, their uses, movements of opinion, well beyond only the ‘trending topics’ made public. In the end, cultural industries, businesses and media publish content on their Facebook ‘Page’ now by buying advertising to reach their own community, before being asked to pay again, and a high price at that, to get the results of the studies that they themselves helped generate.

It may be thought, as I believe, that social media will play a major role in the future in the prescribing and dissemination of information relating to culture, but one must keep in mind all these distortions of recommendation

SMART CURATION. If traditional cultural critics are an endangered species, if the algorithms can be biased and if they struggle, in any case, to propose really relevant recommendations, it becomes necessary to invent a new form of prescription. This is what I have chosen to call in this book: the ‘smart curation’.

Faced with the Achilles’ heel of the internet—the abundance—the return to the traditional model of criticism is no longer relevant. Made of flesh and bone, intrinsically linked to traditional analogue or print culture, this model becomes obsolete because of its elitism and its inability to ‘filter’ efficiently and quickly the mass of accessible content. Above all, it offers a unique vision of ‘good taste’, takes into account a small number of criteria and is unable to provide varied recommendations in accordance with the processes, situations, niches and—let’s dare use the word—cultural ‘communities’. But today, in the age of the internet and cultural fragmentation, there can no longer be a single universal criticism valid for all. There are spheres of taste; therefore, a plurality of recommendations is needed.

The internets are decentralized, decentred, and plural: disintermediation characterizes them. And it is unlikely that we can go back to an elitist model where the judgment is left in the hands of a small number of critics, which already enraged Balzac’s Lucien de Rubempre.

However, the second solution, involving ‘machines’, strictly mathematical, delegating this prescription to automatic algorithms, does not seem more effective. It is too imperfect to be efficient.

‘Smart curation’ offers an alternative solution: it is a combination of the two models, the algorithm on the one hand, and curation on the other. It is a ‘double filter’ that brings together the power of ‘big data’ and human intervention, the association of machines and men, engineers and acrobats. This algorithmic curation will be done by both those who use words and those who use numbers.

‘The algorithm can help identify what’s popular, but it’s unable to say why it’s popular,’ says Fairweather. It relies on the mass, the average or the opposing, ‘I like/do not like’. At best, the technique tries to predict that ‘people who have loved such content may, in the absolute, love another’—a valuable but insufficient proposal. For Fairweather, it is therefore essential ‘to have both the “big picture”, that of the statistics, and the “small picture”, that of an informed person, who has a certain expertise, who makes choices, filters the information and gives his opinion. We cannot be satisfied with the five stars of a good Amazon recommendation.’

Smart curation ‘is a form of intelligent editorialization, an automated and then humanized selection that allows you to sort, select and recommend content to readers.’ It can take various forms but I will define it essentially by three elements. It is, first of all, a recommendation that benefits from the power of the internet and algorithms, but also from a human treatment and a personalized prescription by ‘curators’. This function of ‘double filter’ is indispensable.

Hence, and this is the second point, the need to have recourse to an intermediary or an escort or a third person, for this second filter of curation. This filter cannot be the producer of the content itself, nor the consumer. An author, for example, cannot do his own ‘curation’ himself (while an author can promote himself or a reader may have his own judgment). The second ‘curation’ filter is therefore a mediation that needs an intermediary.

Here, the very term ‘curation’ is interesting. Ambiguous certainly, it still has in Europe and in the United States, an elitist connotation that links it to museology and art museums. At MET, MoMA or the National Gallery, a curator is one who presents and organizes an exhibition in a museum of fine arts. The word has spread in recent years in cinematheques, museums of contemporary art or libraries, before being taken up by the digital culture, becoming one of its ‘buzz-words’.

Finally, the third element of ‘smart curation’ is part of a ‘conversation’. It is not a question of recreating a hierarchical judgement ‘top-down’, supposedly universal, but most often arbitrary or peremptory. It should be a dialogue that paves way for exchanges, round trips, pluralities of taste and is elaborated in different ‘spheres of judgment’.

Ultimately, simple ‘curation’ is the opposite of the aggregation proposed by the algorithms. Curation or aggregation, what must we choose? What if we combine the two? We must defend cultural exception but with mathematical power.

While the name of ‘smart curation’ is new and an expression imagined for this book, there exists many illustrations of the concept. For example, Facebook likes, Twitter retweets, Pinterest pins, little hearts on Tumblr dashboards, and Google+ ‘+1s’: all these network-specific tools participate in this ‘smart curation’ approach.

When a person posts a cultural recommendation on Facebook and that this one is ‘liked’ by his friends, the algorithm takes into account these indications and multiplies the visibility of the initial post (the algorithm used on the ‘Pages’, even more than on personal accounts, is indexed to the number of ‘likes’ and comments). The mathematical power intervenes but, at the beginning, it is a human person who recommended to his ‘friends’ a cultural content that he liked. So we have a mixture of ‘smart’ (the algorithm) and ‘curation’ (the singular appreciation of a person through his ‘like’ or his comment). The peer recommendation is increased tenfold by the mathematical power.

Social listening or curated playlists are another example of smart curation. The technique is found on Spotify, Deezer and Pandora. The algorithms of these unlimited subscription music platforms are, as we have seen, not very relevant and pertain more to ‘smart’ without ‘curation’. On the other hand, when these sites show the playlists of other subscribers, they are doing smart curation with this social listening. (Spotify went further when it introduced, at the end of 2014, the ‘Top Tracks in Your Network’ function, displaying constantly updated playlists of the ‘friends’ followed).

James Iovine, who cofounded along with rapper Dr Dre the Beats headphones, has chosen to redesign iTunes and Beats Music, and launch Apple Music with the model of ‘smart curation’ (Apple bought Beats Electronics in 2014 and has consulted Iovine for its new streaming music service). He was convinced that the recommendation by algorithms, as used on Pandora or Spotify, remains insufficient and that it is necessary to combine the data collected by the algorithms with the judgements produced by human beings to obtain more relevant results. This led to the creation of a musical social media in parallel with Apple Music.

In the radio sector, National Public Radio (NPR) has also innovated in this direction. This American radio network, independent and not-for-profit (but not public contrary to what its name might suggest), made a bet early on in the digital to rejuvenate its audience. The NPR now has 20 million individual visitors to its site and 27 million podcasts downloaded each month. As a result, the average age of listeners listening to the NPR over the air is 52; those who connect to it via internet or podcast are on average 36 years old—almost twenty years apart! A dedicated application on smartphone offers an original experience of ‘smart curation’: the radio is fully personalized, whether through its programs, local stations or playlists. Powerful algorithms, ‘conversation’ tools and ‘geo-targeting’ system for local information ensure a great fluidity of programs, their ‘customization’ and their sharing on social networks. Recommendation is central, key to the success of syndicated programming (and even to TED conferences available on the app). In doing so, the NPR also demonstrates that the podcast is a transitional technology and that mobile streaming, combined with powerful application, recommendation techniques, and relevant algorithms, could be the future of radio.

Harry Potter and les Booktubers

LAUREN BIRD LOVES HARRY POTTER AND WAFFLES. With her teenage physique and funny glasses, she reminds me of these female characters from the novels of J.K. Rowling and Suzanne Collins. And she seems like she is just out of childhood.

Lauren Bird arranged to meet me next to Google’s new headquarters in New York on 9th Avenue and 16th Street in Chelsea. In her spare time, she crashed the offices of YouTube (which belongs to Google), not far from there, where she can use editing studios for her videos. To access it, you must spot the discreet elevator, hidden by a shoeshine, inside Chelsea Market between the 9th and 10th Avenue, and go up to the fifth floor. There, when their YouTube channels achieves some user success, many bloggers, like Lauren Bird, can benefit, free of charge, from a kindly ecosystem. I see geeks sitting in leather sofas with a small exhibition of contemporary art and a large poster promoting a book published by a ‘Booktuber’. Bird is a ‘booktuber’: she talks about Harry Potter in short clips posted on YouTube. Elsewhere, on another channel of hers, she specializes in ‘Waffle Irons’. As others test iPhones or the strength of skateboards, Lauren Bird tests these Waffle Irons to see if they are as resistant as their promoters claim. On her channel, she tries to cook eggs, sushi, gherkins, snickers and even a pumpkin in this waffle maker! The result is sometimes surprising and always hilarious.

More recently, Bird, who graduated from New York University, joined the Harry Potter Alliance, an association of Harry Potter fans who are committed to improving the working conditions of American employees, such as McDonald’s or Walmart. The videos she posts on this topic have already collected several million views on YouTube.

Emerson Spartz, like Lauren Bird, started his career as a simple Harry Potter fan. At the age of 12, he launched MuggleNet, his first site—now one of the main platforms for J.K. Rowling’s novels. The famous novelist invited Spartz to her private estate in Scotland, and she also supported the initiatives of Lauren Bird and her Harry Potter Alliance.

The phenomenon of ‘booktubers’, which initially appeared in Argentina, Spain and the United Kingdom, is clearly part of ‘smart curation’: young readers or students share their enthusiasm after reading a book, facing the camera. Using a smartphone, a GoPro camera or more recently Periscope, they share their passions with staging, simplicity—and talent.

There are as many formats and genres as booktubers: their videos can be serious, crazy, arty or more mainstream, as illustrated, for example, by the YouTube channels of Christine Riccio, Jesse, Raeleen, Ariel Bissett, Priscilla, Kat O’Keeffe or Regan.

Sometimes these ‘booktubers’ start writing, dreaming of becoming, in turn, writers. The editors, as confirmed by Jonathan Karp, CEO of Simon & Schuster, scrutinize their channels in the hope of discovering new talent. The algorithm of YouTube also identifies these videos and may, on occasion, make them go viral on the internet. Without being deemed a ‘great writer’, the ‘booktuber’ can become a ‘brand content’.

Other models of ‘smart curation’ also exist. It can be social media dedicated to reading and ‘fan-fiction’: Wattpad, for example, is Facebook for literature. Based in Canada, this group already has more than 45 million members with nearly 100 million stories uploaded to it (the young novelist Anne Todd has published After, a fan-fiction written on smartphone and read by twelve million people). More than a publication interface for all, Wattpad, a real online reading club, is also a space for comments, sharing, and therefore ‘curation’. The algorithm ensures media coverage of the most popular stories.

In another style, a site like The Conversation, developed in Australia and the United Kingdom, aims, as the name suggests, to recreate a ‘conversation’ with a broad network of academics and intellectuals. With the baseline ‘Academic rigour, journalistic flair’, The Conversation aims to shed light on ideas based on the analyses and points of view of the researchers. In doing so, it allows them to gain new visibility and, aided by social networks, become popular.

Finally, GoodReads, acquired by Amazon in 2013, combines algorithmic search, automated reading lists, and custom book reviews. Internet users can build their own library, rate their books and those of their friends, even as GoodReads offers them recommendations. It is a kind of social media dedicated to books, with 20 million members who can link it to Facebook or Twitter to increase their ‘sociality’. (It is interesting to note here this renewed interest of Amazon for the curation, especially as it is observed in parallel with the acquisition of IMDb in the cinema and Twitch in the video game, two sites based on evaluations and recommendations. Initially, in 1995–2000, Amazon had developed its own tools in-house, and a team of about 20 journalists, writers and salaried editors from Village Voice or the New York Review of Books wrote leaflets, did interviews or broadcast their recommendations, but in 2002, Amazon’s ‘editorial’ department was closed, with ‘human’ publishers being permanently replaced by algorithms).

AT GAWKERS HEADQUARTERS IN NEW YORK, James Del, vice-president of Programming and Marketing, tells me about the ‘Gawker Review of Books’ (review.gawker.com). This site has joined the family of ‘sub-blogs’ launched from the main portal, with the aim of publishing excerpts of new books, to do interviews of authors and for literary recommendations. Although being digital pure-player, sometimes up to the point of arrogance, Gawker also knows to place its bet on an ‘old media’, like the edition of books. The turnover of the sector—around thirty billion dollars a year—has not escaped the founders of the site, which fascinates the bestseller industry. But Gawker Review of Books treats ‘books’ in its own way! To create a buzz, it reveals some secret deals in publishing, talks about defamation lawsuits, and talks about the death of a famous writer. The posts seek to gain audience: ‘The 50 Best First Sentences of Novels’; ‘How Harlequin Became the Most Famous Publisher for Romance Novels’; or, recently, an attack on the New York Times and its list of bestsellers deemed too ‘Caucasian’ (that is, in American speak, too ‘white’, without black or Latino elements). Other times, articles from the Gawker Review of Books, especially long interviews with promotional writers, can be serious, relevant, and well made. This is one of Gawker’s secrets: to recruit the best students from creative writing programs or the most prestigious journalism schools. And every week, the site publishes its ‘Gawker Review Weekend Reading List’—as if it were a respectable mainstream daily.

Gawker hopes to create a new ‘conversation’ around books, like Emerson Spartz wanted to create a conversation about Harry Potter when he launched his MuggleNet site. They are not the only ones.

Other generalist pure–players are interested in cultural criticism in general, and books in particular—to not leave this niche market to new entrants. This is the case of Slate which inaugurated in 2014 a digital supplement for book reviews. For their part, sites like Vice and Politico have recently launched paper versions, in which the books are discussed. With regard to Buzzfeed, known for its short formats and viral posts, the site has chosen to promote long formats and book reviews in a dedicated section. ‘The Buzzfeed website also publishes very long articles,’ insists Henri Finder, editor of the New Yorker, as if to reassure himself.

For, competition is fierce. Everywhere, experiments are numerous and new sites are imagined. In the United States, the interesting Literary Hub selects literary novelties every day and publishes excerpts of books. In Germany, Perlentaucher for its part bets on recommendation, and Enobii in the United Kingdom. In France, sites like EntréeBook, Booknode, Sens Critique and NonFiction have specialized in online criticism, while BdGest/Bedetheque focus on comics. As for the German site dedicated to the theatre, NachtKritik, it is updated collectively at night, to give an account in the morning, of the pieces seen the day before.

The new ‘paper’ literary journals are not falling behind and try to invent bi-media models. In the United States, interesting formats have been devised by n + 1, McSweeney’s, The Believer and Tin House. Appearing three or four times a year, these hybrid journals have developed an economic model based on philanthropy, book publishing and sale of the derivatives. To an extent, these journals believe in quality journalism and do not rely on ‘clicks’ but on ‘Pulitzers’ (to quote researcher Angèle Christin). This does not prevent them from having an innovative and original presence on the web.

Other examples exist, beyond the cultural sphere. A site like Techmeme (techmeme.com), dealing with technological information, manages to mix these quantitative and qualitative approaches: it automatically identifies contents and ‘hot stories’ and then, with the help of ‘human’ editors, validates, hierarchizes and reformats them. Pearltrees, an online tool for ‘collaborative curation’ aims for a similar goal. Other services like Storyful, Vocativ, Dataminr and ReCode offer mixed models that bring together ‘data’ journalism and recommendation, with the intervention of data analysts or Chief Content Officer, alongside traditional editors.

As for the specialist start-up Outbrain, it offers its B2B (business to business) media clients turnkey solutions to the sites to identify the best articles likely to provide them with the most audience, but by adding a linguistic and geographical dimension.

Finally, although more anecdotal, we can mention the initiative of Mark Zuckerberg, the head of Facebook, who launched in 2015 a ‘Page’ dedicated to his readings. He took a resolution to read a book every two weeks and post his opinion on this Facebook account. The Gawker Review of Books, which found that the page in question was not updated, criticized his dilettantism in a post insidiously titled ‘Mark Zuckerberg Is Not Oprah’. We are entering an ‘era of mini-Oprah’, commented the New Yorker, referring to these small prescribers, compared to the great prescriber and literary priestess that Oprah Winfrey was for long in the United States. A beautiful turn of phrase that may sound like the spirit of the time.

Era of mini-Oprahs

THE EVENING IS ON ‘INVITATION ONLY’. Henri Finder opens his beautiful private apartment of Tribeca for a ‘Book Party’. Some of the greatest New York journalists, intellectuals and other publishers are here. Hipsters, start-up heads and other ‘barbarians’ are poorly represented. It looks like the appetisers were made by a star chef; like the ushers and waitresses were preparing to compete for Miss America; like the book which is launched on this occasion, written by a professor at Princeton University, is the most beautiful of the year. Raising his Mojito glass, spiced with fresh red pepper, as a tribute to the petulant, Finder pronounces, subtle and gripping without laughs, a little speech full of references to Henry James! A prestigious ‘Book Party’ but a moment, it seems to me, from another era. All smiles, Finder reassures me: that evening, the book in question—a Note Book—is, he tells me, ‘a curious collection of poems and mini-essays initially posted on Facebook!’

Publishers, journalists, authors: many in the United States are not afraid of the internet. Not impressed by algorithms or social media. They have already integrated them into their lives, even when they pretend the opposite and do not talk about it. ‘We will, of course, have many new voices on the internet that will talk about books, make criticism, and give their opinion. It’s just great,’ says Ken Auletta, the famous New Yorker media critic. He adds: ‘Before it was through word of mouth and now we have “likes” and “links”. That’s great!’

Stepping back is impossible: Ken Auletta and Henri Finder know it well. Quite rightly, Auletta adds a reservation, when I ask him about curation and algorithms: ‘There is no magic formula to be successful on the internet.’ With his flowing articles consisting of dozens of pages, and twenty interviews on the media, entertainment and internet industry, Ken Auletta is one of the veterans of American journalism. He knows he has nothing to prove. Therefore, algorithms! He too is ready to take the plunge and duel modernity. When I leave his big apartment on the Upper East Side, I ask him provocatively if he thinks that an ‘Auletta algorithm’ can emerge. He says: ‘No!’ Before conceding that in the long run, he does not know … and that he will be dead anyway.

New Yorker vs Gawker: this could be the summary of the titanic fight that is unfolding before our eyes. Cultural critique, indexed to the future of the press and the paper book, is preparing for new battles. ‘The digital media revolution is a hundred-year-old battle. We are just at the beginning,’ predicts Gawker’s James Del. He believes that the intersections between media and their audiences, between the authors and their readers, will change fundamentally in the years to come. Conversations, ‘engagement’, and curators will be essential. And algorithm will become the cornerstone of this future. (To avoid insulting the future and gaining a kind of respectability compatible with advertisers’ expectations, Gawker’s management has put some of its sites on standby: the Wonkette, the political media has been resold, also Fleshbot, porn site; Oddjack, the sub-site for online gambling and betting has been disconnected).

I share the point of view of Gawker manager’s on the algorithms. Contrary to the vision of techno-sceptics, I do not believe that machines standardize or impoverish the web. A frequent mistake in the superficial analysis of internet mutations is to see in the ‘big data’ and in the algorithms, the phenomena of standardization, homogenization and ‘mainstreaming’. The machines would always lead internet users to blockbusters, bestsellers and hits, to mass culture and entertainment. This is referred to as ‘locking algorithms’ or ‘addiction algorithms’.

This may be true but it is not systematic. Algorithms are tools that depend on their programming. They can lead to the mainstream; they may be linked to advertising contracts; On the contrary, they can lead to niches, world music, world cinema, avant-garde literature or the most contemporary art. Everything depends on the criteria and the parameters that are fixed to them.

Above all, as the algorithm progresses, machines will be able to improve their performance, specialize and adapt to the most complex interests of consumers. They will consider the sensibilities and the nuances. The real risk is therefore not so much the mainstream as the partitioning content into hermetic niches without intersections or interactions. Instead of imposing the taste of the masses, machines tend to lock the user into his ‘bubble’, and provide only that which he already consumes, and makes it exclusive. It is not so much a process of standardization; rather it is the process of distinction and differentiation. This can lead, at worst, to the atomization and multiplication of hermetic niches; at best to happy fragmentation or diversity. The algorithm is not the enemy of the cultural exception—it can even be a tool.

With recommendation, subscription, algorithm, crowdfunding and the new forms of copyright, the need for ‘smart curation’ appears as one of the fundamental evolutions of culture in the digital age. Culture, which was a ‘cultural product’, is becoming a ‘service’ where ‘content’ is available on all media and platforms. All that remains to be done is to create the conversation to talk about it.

This is what the Gawker’s kids, the Booktubers and countless new ‘mini-Oprahs’ do. And, like them, I do not believe that we can escape the algorithms, take a step back and fall back on the world of old-fashioned critique.

However, like Henri Finder, Ken Auletta or Jonathan Karp, I do not believe that we can be satisfied with a world where all cultural content depends on the algorithms. Robots will not take over from journalists. The machine will not be the future of critique.

‘Smart curation’ can reconcile these two worlds. It can even become one of the new internet battles and a way to ‘disrupt disrupters’. Many new and traditional media are already interested in it, experimenting with amazing or implausible algorithmic tools that combine mathematical power with human judgement. On the academic side, comprehensive research programs on smart curation will bring together social scientists, algorithm engineers and academic cultural critics. Finally, countless start-ups are also working on these ‘double filters’, bringing together investments and recruiting at every turn.

One of them will soon be on Fifth Avenue, one of New York’s most prestigious addresses. Its name: Gawker. ‘We’re going to move this summer: The Fifth Avenue! Yeah,’ exclaims, James Del, pleased with the power of the symbol. It’s David, who’s defeating Goliath. It’s the underdog who is jubilant to join the establishment. And as proof of the ambition and the success story of his start-up grown up, he adds: ‘And this time, we will have an elevator.’