Desktop version

Home arrow Computer Science

  • Increase font
  • Decrease font

<<   CONTENTS   >>

What is artificial intelligence in popular culture?

Artificial intelligence is not a new concept, and the origin of Al goes as far back as Greek antiquity. Nevertheless, it was less than 100 years ago that the technological revolution took off and Al went from fiction to plausible reality (Shani, 2015). Due to its complexity, there are various definitions of Al, which means that no single definition encompassing what this new digital technology means was made. In fact, the notion of Al still evolves, and therefore, I do not seek to lay down a general, all-purpose definition of Al that can be applied in any context in this book. Instead, this book attempts to arrive at a definition that is suited to the cultural sector that uses Al, big data, and algorithms to produce, distribute, and consume cultural content. While historicizing the concept of Al, it provides the best relevant notion of Al in the media and cultural sectors—in particular, in conjunction with cultural production. Al, machine learning (ML), and deep learning (DL) are each a subset of the previous field. Al is the overarching category for ML, and ML is the overarching category for DL. Al can be “applied to fields as wide as computer audio or visual recognition, self-driving vehicles, robots that can respond autonomously to their environments, recommendations of films via Netflix, and financial analysis” (New European Media, 2018,5) (See Table 2.1).

In our modern era, the definition of Al was originally developed by John McCarthy, an American computer scientist, pioneer, and inventor, who was

Machine Learning

A subset of A that inc udes

abstruse statistical techniques that

enable macchmes to improve at

tasks with experience.

Deep Learning

The subset of ML composed of

algorithms that permit software to

tram itself to perform tasks, like

speech and image recognition,

bu exposing multilayered neutral

networks to vast amounts of data

Table 2.1 Relationships of Al, ML, and DL Source: Moore (2019, 8).

Artificial Intelligence Any technique that enables computers to mimic human intelligence, using logic, if-then rules, machine learning (including DL).

known as the father of Al after playing a seminal role in defining the field devoted to the development of intelligent machines. He coined the term “Al” during the 1956 Dartmouth Conference, the first Al conference (Gunkel, 2020). As is well documented (Childs, 2011) in the Independent of the U.K., “The objective was to explore ways to make a machine that could reason like a human, was capable of abstract thought, problem-solving and selfimprovement.” He believed that “every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it” (Childs, 2011). Since then, many computer scientists, science and technology study (STS) scholars, and sociologists have developed the concept of AL

Through the 1960s and 1970s, the excitement further grew as computers became more accessible. Nevertheless,

it became apparent that Al was not living up to the hype. This led to what has been called the Al winter, a period through the late 1980s and 1990s when investment in Al research and commercial efforts slowed significantly. In recent years, excitement about Al has increased due to some high-profile successes.

(Moore, 2019, 6)

Al has been mainly the preserve of computer science, information science, mathematics, linguistics, psychology, and neuroscience. However, it took several more decades for people to acknowledge the true power of AL As the Go match between AlphaGo and Lee Sedol explained in Chapter 1 proves, Al has finally become one of the most popular and influential terms that people remember in recent years.

While there are several major differences, one of the major variances of Al from other digital technologies is its capacity to replace human beings. The word “artificial” “has come to denote that machines can be made to replicate or simulate human intelligence, and hence the affinity between artificial intelligence and digital technologies (including advanced robotics and accelerating automation)” (Elliott, 2019, 2). Of course, this notion itself is controversial because Al cannot replace humans entirely, but partially. In the media sector, for example, journalists are concerned about the increasing role of robot journalism because they believe that robots may replace human journalists in writing and editing processes. However, as is fully discussed in Chapter 7, it is premature to confirm this kind of new trend or phenomenon, as humans are still major actors in media journalism with the help of robots.

More specifically, in the 2010s, Al was defined, in general, as the “simulation of human intelligence through computers, mainly referring to machine learning. Put simply, machine learning is a form of data” (Asia Pacific Foundation of Canada, 2019, 6). This general definition seems to identify three major areas—Al, machine learning, and big data—involved in Al interchangeably. From a slightly different tone, Al is also defined “as a set of algorithms that is able to cope with unforeseen circumstances” (Berend-sen, n.d.).1 Here, Al is compared with algorithms in that Al involves analyzing big data and looking for patterns.

On the one hand, an algorithm is a set of instructions—a preset, rigid, coded recipe that gets executed when it encounters a trigger. On the other hand, Al—which is an extremely broad term covering a myriad of Al specializations and subsets—is a group of algorithms that

modify its algorithms and create new algorithms in response to learned inputs and data as opposed to relying merely on the inputs it was designed to recognize as triggers. This ability to change, adapt and grow based on new data, is labeled as intelligence.

(Ismail, 2018)

Al systems especially learn from experience and over time get a better understanding of what people, in particular, managers, operators, and designers need for their operations. When these algorithms are automated, it is called Al (Lengnick-Hall et al., 2018). As Broussard (2018, 94) also points out, “An algorithm is a series of steps or procedures the computer is instructed to follow.”2

Gonfalonieri (2019) also claims that an algorithm is a “process or set of rules to be followed in calculations or other problem-solving operations, especially by a computer.” Here “the goal of an algorithm is to solve a specific problem,” defined by someone as a sequence of steps. An algorithm is a shortcut that helps people give instructions to computers. An algorithm simply “tells a computer what to do next with an ‘and,’ ‘or,’ or ‘not’ statement.” However, traditional algorithms have an issue. Indeed, people have to tell to create a step-by-step process to reach their objective. Rather than following explicitly programmed instructions, some computer algorithms are designed to allow computers to learn on their own. Putting these definitions together, UNESCO (2020) states,

Al involves using computers to classify, analyze, and draw predictions from data sets, using a set of rules called algorithms. Al algorithms are trained using large datasets so that they can identify patterns, make predictions, recommend actions, and figure out what to do in unfamiliar situations, learning from new data and thus improving over time. The ability of an Al system to improve automatically through experience is known as ML.


Another major element that characterizes Al is whether Al can learn automatically, and it is considered that machine learning is a set of algorithms that arguably enable the software to update and “learn” from previous outcomes without the need for programmer intervention, although the process is not entirely automatic. It is fed with structured data in order to complete a task without being programmed how to do so. Machine learning is made up of a series of algorithms. Basically, Al is designed to learn in the same way as a child. Thanks to a data set, an Al can find patterns and build assumptions based on those findings (Gonfalonieri, 2019). In this light, Niranjan Krishnan, head of data science at Tiger Analytics said (Ismail, 2018),

Al is like a gear system with three interlocking wheels: data processing, machine learning and business action. It operates in an automated mode without any human intervention. Data is created, transformed and moved without data engineers. Business actions or decisions are implemented without any operators or agents.

He continued to state that the system learns endlessly from the accumulating data and business activities and outcomes are getting better with time.

More narrowly, Al could refer to

a branch of computer science focused on simulating human intelligence, one that recently has been especially engaged in the subfield of machine learning: the training of a machine to learn from data, recognize patterns, and make subsequent judgments, with little to no human intervention.

(Lewis, 2019, 673)

In a similar vein, Elliott (2019, 4) defines Al as “encompassing any computational system that can sense its relevant context and react intelligently to data.” Al and machine learning often seem to be used interchangeably. However, they are not quite the same thing, but the perception that they are can sometimes lead to some confusion. “Both terms crop up very frequently when the topic is big data, analytics, and the broader waves of technological change which are sweeping through our world” (Marr, 2016). Simply speaking, “Al is the broader concept of machines being able to carry out tasks in a way that we would consider smart,” while “machine learning is a current application of Al based around the idea that we should really just be able to give machines access to data and let them learn for themselves” (Marr, 2016).

In the realms of media and culture, focusing on its relationship to communication, “communicative Al refers to Al technologies—such as conversational agents, social robots, and automated-writing software—that are designed to function as communicators, often in ways that confound traditional conceptions of communication theory and practice” (Guzman and Lewis, 2019; cited in Lewis, 2019, 673). As Andrejevic (2020) especially points out,

as automatically generated information comes to play a central role in the rationalization of production, distribution, and consumption, artificial intelligence robotizes mental labor: it promises to augment or displace the human role in communication, information processing, and decision-making.


What he emphasizes is that “Al resuscitates the promise of automation in the mental sphere: to be faster, more efficient, and more powerful than humans” (Andrejevic, 2020, 4).

Again, this does not mean that Al entirely replaces human beings. Rather, the focus should be the interaction between technologies and people, both creators and consumers, as they are the developers as well as the users of Al and algorithms, which create new forms of cultural content and consumption patterns. It must be clear from the outset that we use Al to make the work more innovative and workers’ role more rewarding, instead of simply cutting costs by replacing human workers (Britt, 2019). While admitting that various explanations exist, this book defines Al as the simulation of human intelligence through computers supported by and connected with big data and algorithms to not only “intermediate” humanmachine interactions but also “mediate” production and consumption of media and culture through the convergence of intelligent technology and human creativity.

What is significant is that Al symbolizes innovation in the early 21st century, and many countries around the globe have emphasized the increasing role of Al and invested in this particular area. As Robin Mansell (2017, 4286) points out,

Investment in the development and use of novel digital applications, including intelligent or social machines and robots, supported by algorithms and machine learning, is expected by many industry leaders to raise income levels and foster movement along a singular pathway through a fourth industrial revolution.

As Shah (2013, cited in Shorey and Howard, 2016) also argues, “What should be done with Al and big data, and what other kinds of relevant knowledges could it help produce” are among the major agendas for the government, industries, and cultural creators. What they commonly pursue is the enhancement of digital economy, driven by the use of Al and big data, as several digital platforms like Google and Facebook already prove.

Specialists such as digital platform workers, game designers, music composers, and webtoon creators have already utilized Al and big data to dramatically enhance the experience of enjoying and playing, “resulting in increased popularity and soaring profits” (Klinenberg and Benzecry, 2005, 9). Since Al has rapidly taken on a major role in the digital platform and cultural industries, which are very significant sectors for the national economy and society, both governments and corporations have enthusiastically invested in Al and big data. As usual, several Western countries lead the Al sector, followed by a few Asian countries. China and Korea are latecomers in the realm of Al; however, they have already advanced their Al-related platform and cultural industries to the level of potentially leading contenders globally (Walch, 2019). The use of various digital technologies, such as high-speed internet and smartphones, proves that these two major actors show sometimes cooperative and at other times conflicting relationships in the development of an Al-saturated industry structure.

On the one hand, governments have acknowledged the significance of Al and vehemently developed relevant policies, in both financial and regulatory standards. As a policy matter,

Al operates at the intersection of big data and automation. On one side, machine learning—an algorithm that improves through experience, and the kind of Al most talked about in many countries—requires massive amounts of training data to optimize its algorithms, a privacy issue. Once trained, Al requires proper implementation, and should be used only when experts deem it acceptable.

(McKelvey and MacDonald, 2019, 44)

On the other hand, major platform and cultural firms have invested heavily in developing synergistic relationships between several media holdings, “integrating their production processes into convergence systems that yield content for different outlets, cross-promoting programs in different media,

Al, digital platforms, and popular culture 23 and establishing lines of vertical and horizontal integration in production and distribution” (Klinenberg, 2000, cited in Klinenberg and Benzecry, 2005, 10). In other words, like other digital technologies, such as smartphones, games, and the internet, Al and big data have facilitated the growth of large media conglomerates. In The Guardian, Paul Nemitz, a senior European Commission official (Chadwick, 2018) argues, “We need a new culture of technology and business development for the age of Al which we call ‘rule of law, democracy, and human rights by designs.’”

The reality is highly controversial because the business practices supported by the government have not guaranteed diversity, democracy, and equality, unlike their expectations and promises, and instead they raise new levels of concern due to concentration of ownership and skills. Al and big data have even seemed to intensify asymmetrical power relationships between mega platforms like Facebook and Google, big venture capitals, and media giants, and small- and mid-sized cultural firms as well as cultural producers and cultural consumers. This book, therefore, offers several critical discussions on the use of Al in popular culture, which transforms entire cultural industries.

Digital platforms as mediators in the global cultural sphere

The increasing use of Al is visible in digital platforms. Digital platforms equipped with Al and big data have continued to grow and are becoming among the most significant players in the global cultural industries. A platform refers to “the online services of content intermediaries, both in their self-characterizations and in the broader public discourse of users, the press and commentaries” (Gillespie, 2010, 349). In other words, Gillespie seems to consider that platforms are and can be neutral. For example, he (2018, 41) argues that “social media platforms are ‘intermediaries,’ in the sense that they mediate between users who speak and users who might want to hear them, or speak back,” although social media platforms create several negative impacts like fake news and surveillance. Evens and Donders (2018, 4) also argue that “apart from the underlying software and algorithmic configuration, we put emphasis on the intermediary (and gatekeeping) position platforms have when connecting programming to consumers.” Digital platforms, therefore, could be defined as large-scale online systems premised on user interaction and user-generated content—including Facebook, Twitter, YouTube, OTT services, and smartphones (Jin, 2015; Lobato, 2019).

Digital platforms have different dimensions according to their primary areas and purposes. For example, OTT service platforms (e.g., Netflix) are different from social network site platforms (e.g., Facebook) and usergenerated content platforms (e.g., YouTube). It is indeed controversial to categorize OTTs, including Netflix and Amazon Prime, as digital platforms. In this regard, Lobato (2019) points out:

Netflix is not a platform in the same way as social media services like Facebook or Twitter are. Netflix is not open, social, or collaborative.

One cannot upload content to Netflix or design software applications to run within it. In this sense, it is fundamentally different from video sites containing both user-uploaded and professionally managed content (YouTube, Youku, etc.).. . . Netflix is closed, library-like, professional; a portal rather than a platform; a walled garden rather than an open marketplace.


However, due to several of their unique characteristics—data-driven, commercially oriented, and mediated—it is possible to define OTT services as digital platforms, at least as quasi digital platforms?

More specifically, Netflix is an on-demand video-streaming platform in that it connects content providers and final consumers. Although people or companies could not run their software programs on Netflix, this particular OTT service mediates content producers and consumers, and by providing their watching habits to Netflix, the customers fulfill their sociality on Netflix. Kulesz (2018b, 80), for example, emphasizes, “a platform facilitates interaction between users—buyers and sellers, creators and consumers, etc.—in a highly efficient way, and in this regard adds great dynamism to the cultural fabric,” instead of focusing on technological aspects. Netflix has continued to transform its own major characteristics, from a distribution channel to a production company to now a digital platform. Therefore, it is not dicey to discuss Netflix through digital platform approaches.

Due to the rapid growth and massive influence of digital platforms, academic discourses on digital platforms are widespread with three major focal elements. To begin with, as in the case of Al, the digital platform must be understood comprehensively, instead of narrowly focusing on technological aspects. Several scholars (Jin, 2015; Srnicek, 2016; van Dijck et al., 2018; Nieborg and Poell, 2018; Flew, 2018b; Mansell, 2021) argue that digital platforms encompass various unique characteristics. For many, a platform is a programmable architecture designed to organize interactions between platform users. Therefore, many people consider platforms simply as technological tools that allow the users to do things online: “sharing content, making connections, ranking cultural artifacts, and producing digital content” (Gehl, 2011, 1228). However, “these online activities hide a system whose logic and logistics are about more than facilitating: they actually shape the way we live and how society is organized” (van Dijck et al., 2018, 9).

Among these, van Dijck et al. (2018, 9) especially point out that “a platform is fueled by data, automated and organized through algorithms and interfaces, formalized through ownership relations driven by business models, and governed through user agreements.” Prior to this, Jin also (2015) argues, digital platforms could not be fairly understood without contemplating three major areas: technological sphere, corporate sphere, and political sphere. While developing the notion of the platformization of cultural production, Nieborg and Poell (2018) emphasize the significance of the analysis

Al, digital platforms, and popular culture 25 of overall ecology relevant to digital platforms. They argue (2018, 4276) that platformization can be defined as

the penetration of economic, governmental, and infrastructural extensions of digital platforms into the web and app ecosystems, fundamentally affecting the operations of the cultural industries. So far, this process has been examined from three perspectives: business studies, political economy, and software studies.

As such, the digital platform shapes the communications, interactions, and consumption that it facilitates—through interface design, moderation policies, and terms of service (van Dijck, 2013). The platforms’ commercial interest in gathering user data implies that people cannot study a single layer but must acknowledge the complicated relationship between the technical affordances and the underlying commercial interests (Jorgensen, 2019).

Second, digital platforms must be treated as mediators, not as intermediaries. As discussed, Gillespie (2010) and Evens and Donders (2018) define platforms as content intermediaries. This certainly explains the partial nature of digital platforms; however, it does not focus on digital platforms’ true nature as one of the most powerful communication systems. In contrast to this, van Dijck (2013, 29) points out that a platform is a mediator rather than an intermediary, because it shapes the performance of social acts instead of merely facilitating them. Bratton (2015) especially considers the platform as “the third institutional form alongside nation-states and markets, mainly due to platforms’ increasing power in our contemporary society.” Compared with traditional media conglomerates, digital platforms are guided by business models that foreground connectivity between users and producers over the creation of products, not only by cultural creators but also general users, and therefore, they tactically mediate these two major components for their financial gains (Cunningham and Craig, 2019).

As such, digital platforms work not only as simple conveyors and/or distributors of value neutral technologies but also play a key role in manipulating and controlling the entire vicious circle of the entertainment industry for their own business successes. The ways in which digital platforms mediate access to cultural content through the use of Al algorithms ask us to question if and how these platforms are not only distributing but also reproducing media information and popular culture. What I am arguing here is that digital platforms supported by Al should be considered as powerful entities able to control the entire process of cultural production due to their role as mediators.

Third, digital platforms are global, and the process of globalization in the realm of platforms has been the fastest thus far. Several digital platforms, from search engines like Google to social media such as YouTube and Facebook, have clearly attempted to target global users, not only national users. By penetrating the global markets, these platforms have continued to gain huge profits from foreign countries, while dominating the flows of information and culture. Equipped with Al supported by big data gathered from global users, several major platforms, which were mostly developed in the U.S., dominate the global markets, as more than 90% of search engine users mainly rely on Google (StatCounter, 2019). Facebook and Netflix have also garnered more revenues in the global markets than the U.S. market due to the increasing number of international users since many years ago.

In fact, ever since it began streaming in the U.S. in 2007, Netflix has become one of the largest and most significant global OTT service platforms. Netflix, as an interesting example of a digital platform as well as television source due to its unique characteristics—“a computational, software-based system that can produce a television-like experience” (Lobato, 2019, 35)— has greatly reshaped both the global audio-visual industries and people’s habits in consuming cultural content. Since digital platforms have heavily relied on data that they garner from their users, a handful of mega platform giants, including Facebook, Google, and Netflix, have certainly controlled the global markets, which is also one of the major reasons why they are major actors in the age of Al, which also depends on the quantity and quality of data.

Digital platforms, including social media platforms, OTT platforms, and smartphones have greatly influenced people’s daily activities and cultural lives. Digital platforms as some of the most significant mediators rather than intermediaries should be understood from various perspectives, including technological, commercial, and global standards. In particular, it is critical to understand the significance of the nexus of Al and digital platforms—the two most important digital technologies in the early 21st century—in the vicious circle of cultural production.

<<   CONTENTS   >>

Related topics