As innovations like driverless cars and trains, smart vacuum cleaners and smart fridges, and exercise and dream tracking programs become part of our everyday lives… Could we really expect books to remain unchanged? The artificial intelligence and big data boom is likely to radically change the way we access and consume all kinds of cultural content. In the midst of the current uncertainty and change, we try our hand at forecasting and venture to predict the future of the book.
Artificial intelligence is by no means new, it is something that humans have been working on and studying for at least five decades. But over the last few years, the potentially unlimited access to large amounts of real-time data on human behaviour – also known as “big data” – and a rapid increase in computational power have led to an explosion in the potential uses of artificial intelligence. As a result, it is no longer a mere scientific concept but a social reality that is already transforming everyday life, including the way we access and use all kinds of cultural content.
Just like electricity in the late nineteenth century, artificial intelligence is set to become a structural commodity that affects most economic activities as well as the business models used by all kinds of companies. In other words, all business sectors, including the cultural industries, will base their operations on AI. Almost everything you do while you are using screens, as well as most apps that you interact with on your mobile phone, collect all kinds of data, which are then used to deliver better and more personalised services: the cookies that monitor your web browsing to determine your level of interest in a particular online purchase; the GPS that pinpoints your exact location on a map to help you find a particular store or to book a taxi; watches with sensors that register your movements and even your body temperature… Faced with this new reality, professionals from the cultural and creative sectors should begin to consider how artificial intelligence will affect their work and what steps they should take in the short and medium term, because the speed of these changes is likely to take us by surprise. Over the next few years, cultural content recommendations will largely come from artificial intelligence and big data technologies. That does not mean that booksellers and book and film reviewers will disappear altogether, but they will have to find new ways to add value for consumers of cultural content.
In just a few years, artificial intelligence has become part of our lives. Apple’s Siri and Google Now are just two examples, and although they may not yet be perfect, their development has been remarkable. We are also seeing the surprising results of the first big data forays into the creation of cultural content. Netflix analysed the actions and preferences of its subscribers to determine the focus and pace of the plot of its series House of Cards. For months, the team behind the platform studied the behaviour and decisions of its users: which films they chose and which they watched through to the end, their votes and comments, their favourite characters, their interactions with each other, and so on. The aim of this big data analysis was to discover user preferences and take them into account when producing new episodes. Similarly, Marvel Studios also tracks and analyses user-generated big data in order to detect what characters and relationships will maximise the audience’s interest in its films and to generate plot ideas and archetypes for its characters in future projects.
“Smarter” book recommendations
The advice of booksellers and librarians is already supplemented by new recommendation systems produced by software tools such as Tekstum and Komilibro. Like Amazon, these technologies generate personalised recommendations targeted at individual readers by analysing purchasing habits, but they also carry out more detailed studies of our cultural consumption behaviour. Information on actual reader behaviour and other aspects such as moods and emotions will become the main asset and competitive edge for publishers, bookshops and libraries in the digital era.
In the analogue age, the main and almost sole indicator of a book’s performance on the market was its retail sale figures in bookstores. With the advent of electronic books and audiobooks, the publishing sector now has access to enormous amounts of data on reader behaviour. The new artificial intelligence algorithms track much more than bestseller lists, publisher metadata or star ratings, which do not take into account possible nuances that set particular books apart from each other. Algorithms based on big data and artificial intelligence, like the one used by Tekstum, help publishers, booksellers and librarians find out what feelings, sensations and emotions a book transmits to its readers, and offer a scientific, real time analysis of reader opinions, comments, and reviews on literature platforms, blogs, and social media. These new algorithms give book industry professionals real-time information on aspects such as the average amount of time that readers spend on a book, whether or not they finish it, peak reading days and times, what paragraphs have are underlined most, and how readers interact with books from certain genres.
These types of data give publishers, booksellers and librarians an insight into reader behaviour. On one hand, this helps them to publish of books that attract readers on a more personal level, and on the other it allows them to offer more personalised recommendations of existing books that reflect what readers want and expect. The benefits of this data are truly positive for readers and for industry professionals. Readers can discover more books that closely reflect and meet their reading preferences, while publishers can minimise the risks involved in supporting new authors and works because they have more information on supposed market demand. This data, combined with the information provided by social media, provides a clearer idea of the market. Industry professionals can then use this valuable information to shape artistic creation according to different cultural preferences, and to create marketing and sales strategies that are more likely to succeed because they take into account user history data.
Nonetheless, one of the main risks of the misuse of data tracking is that all cultural production could end up catering exclusively to supposed market demands. Although many of us consider that much of today’s artistic work is already governed by those interests, I think we should aim to strike a balance between the use of these new technologies and artistic production, in order to guarantee the bibliodiversity of voices, stories, point of views, and so on.
Robots will write stories too
We are getting closer to the day when artificial intelligence, robots, and algorithms will write their own stories, books and narratives. There are already technologies that generate stories based on the style of particular writers, and algorithms that look for influences and connections between authors, so it would not be surprising if artificial intelligence soon starts to generate its own creative writing in the near future. In fact, several niche news platforms already use bots that read data-driven news items and use them to generate their own articles, even if they are then checked by human writers.
Creative writing is turning out to be fertile ground for artificial intelligence. Last year a “manuscript” written by a computer was longlisted for a literary award in Japan (a fact that, incidentally, did not surprise some of the more exacting literary critics). Irony aside, “artificial creativity” is progressing more quickly than we had imagined. In the near future, writers will use interactive technologies to produce parts of their work. New digital narratives in which reader decisions and interaction determine the course of the story are appearing all the time. AdrenaLivre, Blackbar, Adela and Seen are excellent examples of this new way of telling stories. Reader interaction is maximised through the collection of user behaviour data, taking the personalisation of the reading experience to unprecedented levels. These new technologies raise new questions: in these kinds of partly automated digital works, should the author also receive royalties for the “artificial” part? Or should it be exempt, given as it is open access technology? Should publishers have to purchase the rights to these technologies in order to be able to distribute royalties accordingly? How will publishers and developers share rights? We need to consider all these possible future human-machine collaboration scenarios within the creative process, as they will become more common than we think.
Some readers may think that this is still science fiction, but we are clearly not the only ones taking it seriously: the European parliament is considering the need to equip robots with rights and responsibilities in the future, given how fast research and development is progressing in this area.
Virtual Reality will become part of our “analogue” lives
All civilisations have their own forms of expression. Throughout history, some civilisations have expressed themselves orally and some in writing, some developed hieroglyphs, others used printing technologies, and so on. Now, everything indicates that the prevailing language in the digital era will be intelligently visual.
Experts in the field tell us that there are currently two different types of virtual reality. On one hand, there is “immersive” virtual reality, in which users become part of the story and experience it real time. Rather than learning or reading references to the past, students and readers can now use virtual reality to experience it first hand. This technology opens up a new world of opportunities for the cultural, publishing, and education sectors. On the other hand, there is “augmented reality”, which enhances the process of discovering and purchasing products and services, as well as other areas such as labour relations. This type of VR will allow you to “try on” a suit without getting undressed, check whether a hairstyle suits you before taking the plunge, redecorate your house without spending a cent, wander through a hotel room simulation before booking, etc. All of these will be common “experiences” earlier than you think. By 2025, we will no longer be able to tell the difference between analogue and virtual reality. We will live in both realities simultaneously, and it will be as “normal” to us as our interactions on social networks are now. It is increasingly clear that virtual reality will change the way we discover, buy, and use all kinds of products, and cultural products are no exception.
Bookshops and libraries will become even “smarter”
Technologies such as artificial intelligence, virtual reality and big data will also have an enormous impact on the way we discover traditional books in physical spaces. For now, library and bookshop users live in two different worlds: traditional physical spaces and new online services accessed on screen (online purchasing and borrowing of books, audiobooks and newspapers and magazines, customer support on WhatsApp…) The challenge is to combine these two worlds and particularly to add value to visits to libraries and bookshops now that users can largely meet their needs online. As a sign of the times, mobile phones will be key to the fusion of these two worlds. The very name “mobile” suggests what makes them so favourable to the complementarity of the physical and online aspects of libraries and bookshops.
Fortunately, the concept of libraries and bookshops will expand beyond physical space in the twenty-first century, because new technologies will allow users to access their services from any part of the city: bus shelters, parks, lampposts… Libraries and bookshops will be able to place smart sensors in any of these places and offer readers relevant content based on geolocalized context-sensitive information thanks to their mobile phones. More and more libraries and bookshops are incorporating proximity technologies such as beacons and NFC sensors to enhance the experience of discovering and purchasing print books. Bookshops that fail to recognises that their competitive edge lies in providing an “experience”, beyond the book, will see their customers crossing over to the online world…
In this new fusion of the physical and virtual worlds, “bots” will improve the way we interact with bookshops and libraries. Our smartphones will be full of chatbot apps that will help answer all kinds of questions on matters relating to specific books we are looking for, or to particular aspects of a certain work or author. The use of these chatbots is by no means incompatible with seeking the help of human experts to find out about subjects that require more specialised information. The ASK Brookly Museum app, for example, allows users to ask questions that are answered by the museum’s curators and experts.
Looking a bit further into the future of libraries and bookshops, these bots are likely to continue to evolve with the implementation of natural language processing software, which includes aspects like moods, emotions, double meanings, and so on. It will also become common to see robots carrying out all kinds of tasks at libraries and bookshops. Humanoid robots like Pepper are already answering questions and helping customers with purchases at museum shops and bookshops. Libraries are once again spearheading experimentation in this field, using robots as a fun, hands-on means to teach young users to programme and use code. These robots are equipped with two cameras, four microphones, motion sensors, sound and facial recognition software, tactile sensing and echolocation to detect walls. Robots in the near future will also be able to converse with users in up to nineteen different languages, tell stories, and recite poems. These kinds of technologies will be implemented in bookshops and libraries, taking the interaction between humans and artificial intelligence to a more humanistic level.
The race to become the world’s smartest
Artificial intelligence is becoming one of the main interests of technology corporations. This new frontier is now the El Dorado for companies like Google, Amazon, Facebook and Apple, and it seems set to become a focus of technological development for the next few years. Facebook wants to use artificial intelligence to help blind people “see”, and is also working on inter-machine deep learning systems and facial recognition systems. Apple has acquired Emotient technology, which uses facial recognition systems to detect emotions, and is also continuing research and development of its existing Siri system. And, unsurprisingly, Google is still investing in improving its algorithm in all kinds of ways, trying to make it increasingly intelligent. One of its most creative recent inventions is DeepDream, a programme that can create and enhance new images based on existing ones.
If an activity can be tracked, it will be tracked in real time
Everything we do in both our analogue and digital lives will end up being tracked in real time. Artificial intelligence, virtual reality and big data will generate a whole lot of benefits as a result, but they are also taking us into new and potentially dangerous scenarios. Google, Amazon, Facebook, Apple and other tech companies will track and analyse data, supposedly in order to offer more targeted, personalised services. The fact that if something can be tracked, it will be tracked in real time is something inevitable that we will have to learn to live with in the digital era. But this does not mean we have to give up our rights as citizens and consumers.
Most online business models (search engines, social media, e-commerce platforms, recommendation systems, etc.) are based on personal data entered by users. Information on real customer behaviour and satisfaction, which was impossible to obtain in the analogue world, will become the main asset of artificial intelligence.
In the analogue world, laws prevent companies from using personal data for commercial purposes. But the new e-commerce platforms and social networks have enormous amounts of data generated by our online behaviour (what we browse, what we buy, what we read, what we don’t read – which is also important –, what we recommend, etc.). This user profile information has enormous commercial value that companies like Amazon, Apple, Facebook, and Google use to fine-tune their recommendation and sales algorithms.
Interestingly, the companies that encourage us to share every instant of our lives are the world’s least transparent. The NGO International Transparency rates Facebook, Google, Amazon and Apple lower than many banks. The secrecy of these companies is totally at odds with their push to bring everything into the public sphere. Amazon, for example, does not declare the results of its operations in the countries where it operates, including Spain.
Public authorities need to take further steps towards legislating what these companies can and cannot do with our data, in order to protect our rights as consumers and citizens. We need to guarantee that the public interests of the digital society of the future are valued above the interests of companies, regardless how legitimate these interests may be.
Conclusion: Turning data into a public service
As a society, we need to ask: Should we force companies operating in the digital economy to share the data tracked by their algorithms in order to make it public and guarantee access to it? If the aggregate data is in the public domain, public access will offset the competitive edge and no company will be able to take an overwhelmingly dominant position.
The fact that I raise this does not make me “digital dissident”. As an inveterate proponent of the digital age, this call for critical debate on the impact of artificial intelligence is not intended to detract from the importance or the value of its benefits to society. I only wish to encourage discussion on whether the future we are creating is better than the uncertain present.
We should not forget that we are now starting to define the model for the digital society of the future. And that what we do during these early stages, and how we do it, will have important repercussions on the future.