Tag Archive for: automation

Public media and automation

I’m super happy to announce a book chapter, co-authored with my colleague Jannick Sørensen, in The Values of Public Service Media in the Internet Society. Our chapter is titled Can Automated Strategies Work for PSM in a Network Society? Engaging Digital Intermediation for Informed Citizenry.

2020 was a tough year for everyone, all round. It was also tough in the research output space as reviews slowed, research focus was redirected, conferences stopped, and the overall productivity of our research space grinded to a turtle pace – at times driven by an increased demand on our skills in the teaching space.

What I think we will see is a slowing of research output in the next few years as we all took a hit in research access, fieldwork and overall ability to keep researching during 2020. But it is nice to see colleagues still publishing for the moment and getting back on track in 2021.

One of those outputs for 2021 is our co-authored chapter that explores the role automation plays in public service media. To approach this we have used the lens of digital intermediation to understand how user visibility plays into the overall strategies of increasing uses of automation within public service media.

As always, please get in touch if you have issues with access to the book chapter.

hutchinson_predictive_media

The following passage is a thought moment, and by no means exhaustive of placing the idea within existing theories/fields. It would be interesting, and probably the published version of this will do so, to align it with media and cultural studies, queer theory or perhaps discrimination studies. That said, here is a thought process…

I have been undertaking substantial research into artificial intelligence (AI) and automation since arriving here at Hans Bredow. I am beginning to think that perhaps automation/AI isn’t the best or most appropriate way to frame our contemporary media lives. Those concepts certainly are a part of our media lives, but there may be a better way to describe the entire environment or ecosystem as I have previously written.

What I do understand at this point is that media curation/recommendation is responding to us as humans, but we are also responding to how that technology is responding and adapting to us. This is a human/technology relationship, and one that is constantly being refined, modified, adapted and changed – not by either agent alone, but collectively as any two agents would negotiate a relationship.

This type of framing, then, suggests we should no longer be thinking about algorithmic media, or automated media alone. Perhaps what we should be thinking about is the relationship of processed and calculated digital media with its consumers – for this I will use the term predictive media.

I will attempt to explain how I have arrived at predictive media.

Artificial Intelligence (AI) Media

Media certainly isn’t in an AI moment – I’m not entirely sure I align with AI to be honest (or at least I am still working through the science/concept and implications). Beyond its actual meaning, it feels as thought it is the new business catch phrase – “and put some AI in there with our big data and machine learning things”. If artificial intelligence is based on machine learning, the machine requires three phases of data to process: to interpret external data, to learn from those data, and to achieve a specific goal from those learnings. This implies that the machine has the capacity for cognitive processing, much like a human brain.

AI is completely reliant on data processing to produce a baseline, incorporate constant feedback data after the decisions have been made, and the recalculation of information to continue to improve its understanding of the data. Often, there is a human touch during many of these points placing a cloud of doubt over the entire machine learning capacity. While this iterative process is very impressive when done well, there will always be data points that are indistinguishable to a computer.

We should instead be thinking about these processes as a series of decision points, of which we also have input data.

Say for example, you are making a decision to board a bus to travel into town. AI would process data like distance, timetable, the number of people on the bus, for example and recommend which bus you should catch. What it can’t tell is if the bus driver is drunk and is driving erratically, or that the bus has advertising that you fundamentally disagree with, or that you have 10 students travelling with you. In that scenario, it will be the combination of AI processes along with your human decision making that prove to be the best interpretation of which will be the best bus to catch in to town.

As I see it, we are not in a pure Algorithmic Media moment – and this will be a long way away, if it manifests at all.

Algorithmic Media

We have also seen the rise of algorithmic media, which often presents itself as recommender systems or the like, which essentially suggests you should consume a particular type of media based on your past viewing habits or because of your demographic data.

Algorithmic media can be very useful, given our media saturated lives that have Netflix, Spotify, blogs, journalism, Medium, TikTok, and whatever else makes up our daily consumption habits. We need some help to sort, organise and curate our media lives to make the process possible (efficient).

Think of a Google search. It is often the case we search for specific information based on our needs. Google knows the sorts of information we are interested in and will attempt to return information that is relevant and useful. Of course this information result has a number of levers in operation behind the mechanics of results, for example commercial priorities, legislation, trends, etc., Further, we have also seen how algorithms can be incredibly racist, selective, indeed chauvinistic.

In some areas, developers have started addressing these areas, given the algorithms are developed by humans. But there is still a long way to go with this work.

So in that sense, I’m not algorithmic media makes a whole lot of sense due to the problems associated with it. It could be that by the time the algorithmic issues are entirely addressed, we will have moved on to our next media distribution and consumption phenomena.

Predictive Media

So if this is our background (and I understand I have raced through media and technology history, and critical studies here – I will flesh this out in an upcoming article), humans have altered their relationship with technology.

Heather Ford and I are about to (hopefully!!) have an article published that explores the human/technology relationship in detail through newsbots, but I think it is broader than bot conversations alone.

Indeed, content producers adapt and shift their relationship with algorithms daily to ensure their content remains visible. But I think consumers are now beginning to shift their relationship with how technology displays information. If not shift, we are definitely recognising these digital intermediary artefacts that impact, suspend, redirect, or omit our access to information.

Last week, Jessa Lingel published this cracking article on Culture Digitally, The gentrification of the internet. She likened our current internet to urbanisation, and made the argument that the process of gentrification is clearly in operation:

an economic and social process whereby private capital (real estate firms, developers) and individual homeowners and renters reinvest in fiscally neglected neighborhoods through housing rehabilitation, loft conversions, and the construction of new housing stock. Unlike urban renewal, gentrification is a gradual process, occurring one building or block at a time, slowly reconfiguring the neighborhood landscape of consumption and residence by displacing poor and working-class residents unable to afford to live in ‘revitalized’ neighborhoods with rising rents, property taxes, and new businesses catering to an upscale clientele

Perez, 2004, p.139

In her closing paragraphs, Jessa made a recommendation that is so obvious and excellent, why haven’t we done this before?

Be your own algorithm. Rather than passively accepting the networks and content that platforms feed us, we need to take more ownership over what our networks look like so that we can diversify that content that comes our way. 

Lingel, 2019, n.p.

It made me think about food and supermarkets – certainly in Sydney we have two (maybe three) major supermarkets. But there is a growing trend to avoid them and shop local, shop in food co-ops, join food co-ops, and change our food consumption habits entirely. If those major chains want to push inferior products and punish their suppliers to increase the bottom line, as consumers we (in the privileged Australian context) have the option to purchase our food elsewhere.

Why wouldn’t we do the same with our digital internet services? Is this a solution to bias, mismatched, commercially oriented media algorithms and the so-called AI?

Is Predictive Media the Solution?

I think we can apply the same approach towards predictive media.

We cannot consume the amount of media that is produced, suggesting we may be missing crucial information. We cannot trust automated media because it has proven to be incredibly bias. But perhaps it is in changing our relationship with technology and understanding how they work a little better, we might find a satisfactory medium.

It is not only greater transparency that is required to address our problems of automated and algorithmic media, but it is a proactive engagement with those machines to train the programs to understand us better. But changing that relationship is difficult if you don’t know that is an option. So perhaps the real call here is to establish alternative and transparent digital communication protocols that are easily accessible and decipherable for users. In education, change is possible, and this may be a defence against the current trajectory for digital media.

The combination of both increased understanding/transparency and more active engagement with training ‘our’ algorithms could be the basis for predictive media, where predictive media helps us beyond a commercial platform’s profit lines, and exposes us to more important and critical public affairs.

Original Image by Hello I’m Nik.

Jonathon_Hutchinson_Internet_Research

I’m lucky enough to be the Program Chair for the 2019 Association of Internet Researchers Conference, to be held in Brisbane in October. During the last week, I have engaged in the next task as Program Chair and gone through each individual submissions as I assign them to reviewers. This process involves reviewing the title, the abstract and then matching those papers to most suitable experts within the Association.

For those non-academic folk reading this, the conference process usually involves responding to a conference theme as designed by the conference and organisation committees, where potential delegates submit a proposal of anywhere between 500 and 1200 words addressing that theme. This proposal is then sent to a number of reviewers who conduct a blind review (blind meaning they do not know who the author(s) is/are), and then the paper is returned to the program chair with a review and overall score. The papers that receive a suitable score are invited to submit their paper to the conference, while the others are rejected.

We are just about to send the papers out to the reviewers after they have been assigned, which has provided me with some unique insights into the state of the field of internet research. Granted, the proposals are responding to the theme of Trust in the System, which will skew the submissions slightly, but typically academics will usually make their research align with any given conference theme as one’s field usually moves towards a common trajectory. The research that has been submitted can be read as a very strong overview and indicator of where the field is currently, and where it is heading.

Of course the items below are seen through my eyes, which is the first parse of the content coming through the submission portal – the final version of papers that will be accepted and presented will no doubt differ slightly from these initial observations.

What are the hot internet research topics?

As you would expect there is a growing number of research papers in the area of algorithms and platforms. The concept of automation and recommender systems has spread beyond Netflix and permeates in the areas of news and journalism, smart cities, politics, and healthcare.

Platform research continues to be incredibly important with work critically looking at YouTube, Instagram and Facebook as the most popular areas. It is interesting to see the rise of focus on emerging Chinese social media platforms – while I didn’t notice any on TikTok, there was a focus on WeChat and Weibo.

Other very popular areas of research interest include governance and regulation of internet and social media, news and journalism related to the internet, social media and politics, methodologies, labour and things/bots. There is also a group of researchers interested in Blockchain.

Who are internet researchers?

One of the core roles of the review assignment was aligning the papers that were submitted with relative experts in the field. To assist in this process, members of the Association nominate the topics and methodologies of which they are experts. This information provides a unique insight into how we see ourselves as internet researchers.

I have not crunched hard data on this, and would not publish any sensitive data from the Association, so this is a broad observation of my aggregated insights. That is, these are the methods fields that kept popping up when I was assigning papers to reviewers.

One of the most popular internet researcher categories that was available from the pool was ethnographers for social media – participant observation across social media practices. I directly fit into this category and needless to say much of the work undertaken by these researchers could easily align with my own research endeavours.

An emerging category that aligns with the growing field is social media algorithm analysts. As humanities and social scientists become increasingly involved in data science alongside media and communication, the rise of algorithmic analysis has become not only popular, but essential to understand our field.

News and journalism experts are often coupled with social media experts, and the other interesting (and popular) couplings included discourse analysis with social media, and social media and textual analysis/content analysis.

There is a significant gap however, in those researching identities and activism – from what I can see across most of the communication infrastructure formats. A number of researchers are presenting work in this area, yet we still don’t see ourselves as a large cohort of experts in identity research – which seems odd. Perhaps this is just how the methodological categories appear in the conference system, or perhaps this is true of how we (don’t) identify as researchers?

So what does all this mean?

Well, these insights certainly won’t change the field’s direction but it does offer some insights into the gaps of internet research. I think we have platform research covered, while social media and ethnography is very strong. Social media and politics also has a very strong presence.

But there are areas that lack representation in internet research, that would be useful for researchers to pick up on in the next 12 months.

These include:

  • Ethics – in both use of internet and how to research the internet;
  • Algorithm analysis – the growing field here requires more people to apply data science to their existing work on platforms, social media etc.;
  • Geography and geolocation – I didn’t notice any human geographers (I might have missed this) conducting work in internet research in this sample. There is a small group of researchers undertaking geolocation specific work, but there is room for more;
  • Internet histories;
  • Labour;
  • Public sphere;
  • Surveillance;
  • Apps;
  • Conflict; and
  • Commerce.

For me, a light bulb just went on with how to personally align my research after attending conferences. I guess I always thought of conferences as a chance to present my current work alongside the field. But after having undertaken this Program Chair role, I find it is better to also analyse the gaps in the field to position your work for the next 12 months.

Perhaps scholars have always worked like this and I am just catching up with the game, but having these insights has been incredibly useful to shape my thinking. Hopefully they are useful to others in some capacity.

Original photo by 85Fifteen on Unsplash.

Data_Ethnography

I’m cooking up some ideas while I’m away on Sabbatical at the Hans Bredow Institute. My core focus at this stage is the ‘how to’ research automation and algorithms. My current approach is integrating retro engineering through ethnography and design thinking. At this stage, I’m calling it Data Ethnography and below sets out a guideline for what I think that should be.

No doubt this is the skeleton for a journal article, but what is here is the early developing of a new method I am currently working on.

If you think this methodology could be useful, or you have any feedback or suggestions, please leave them below in the comments.

Why Data Ethnography?

Humanities and social science digital research methods have been interrupted due to the prominence of privacy and surveillance concerns of platform interoperability that produces large quantities of personified data. The Facebook Cambridge Analytica scandal, especially the revelation of its ability to construct predictive models of its user’s behaviors, brought to the public interest concerns over how platform user data is harvested, shared and manipulated by third party providers. The global pushback against the platform provider’s use of these data resulted in platforms closing down some access to application programming interfaces (APIs) to inhibit data manipulation. However, these restrictions also impact on how public benefit research is conducted, providing a useful prompt to rethink how humanities, social scientists and human computer interaction scholars research the digital.

While the datafication of our digital lives has provided us with new insights, the digital methods that enable us to research our digital selves have always been mixed to understand the field of enquiry, along with its surrounding political, cultural and economic constructs.Increased digital practices built on sophisticated calculations, for example the use of algorithmic recommendations, connected devices, internet of things, and the like, have impacted on our research environments, prompting the question, how do we research what we can’t see?This article provides evidence from investigating the visual cultures that surround YouTube that a new methodology is required to research the apparent ‘black boxes’ that operate alongside our digital selves through data ethnography. Data ethnography is the combination of stakeholder consultation, top level data analysis, persona construction, fine data analysis and finally topic or genre analysis. Data ethnography enables not only what we cannot see, but provides a useful way to understand government interoperability mandates and inform appropriate policy development.

Overview of Data Ethnography

The Five-Stage Process of Data Ethnography

Consultation

This methodology emerged from asking the question, what does the Australian YouTube visual culture look like? Building on the long-term participant observation that is synonymous with ethnography, a researcher is able to understand the norms, cultural affordances, communication practices, and so on. The researcher is required to both produce and consume videos on the platform to understand how users will create content to suit the platform constraints. Simultaneously, viewing the content provides insights into how viewing publics are constructed, how they communicate, what is considered important, norms and languages. In the context of YouTube, this included the platform, but also the intermediaries such as digital agencies, multichannel networks and other digital intermediaries such as influencers to highlight publication strategies. The combination of this ethnographic data provides a compelling starting point for the additional methods that emerge.

The video content was analysed using discourse analysis reflective of Jakobson (1960) to understand the video language function as referential, poetic, emotive, conative, phatic, and/or metalingual. As such the discourse in one of four ways: contact enunciation – looking into the camera & addressing the audience; emotive enunciation which is the expressive or affective relating to the style of the YouTuber; genre including thematic content, style and compositional structure; enunciative contract which is the reading contract (Véron, 1985) between the enunciator (addressor) and enunciatee (addressee). The discourse analysis enabled the vast amounts of YouTubers to be categorised into a smaller, more manageable group of users.

Building on the discourse analysis, I asked the users of the platform the following questions:

  1. What is your gender?
  2. What is your age?
  3. How often do you use YouTube in a week?
  4. What is your favourite category of YouTube video?
  5. Are you likely to watch the next video that YouTube suggests for you?
  6. Do you ever watch the trending videos?
  7. When you enter this address into your web browser, what is the URL of the “up next” video that it suggests for you: https://youtu.be/4_HssY_Y9Qs

The results of these several questions then guided the following snowballing process of the additional methods.

Top Level Data Analysis

Before undertaking comprehensive data scraping processes that rely on platform data availability, it is useful to observe how various incidental metrics are available. In the case of YouTube, this related to likes, comments, views, and the like that provide insights into what people are watching, how they engage with the content, and how they talk about the content. These top level metric data observations enable the researcher to direct the research or focus on areas of interest that are not otherwise obvious through the consultation phase of data ethnography. The top level metrics further support the user practices on how content is produced, published, shared, and consumed amongst a wide variety of users. Finally, the top level data analysis enables the researcher to ask questions such as what data are available, which processes might be automated, and how might these data be repurposed for other sorts of measurements.

For YouTube, the top level data analysis translated to the following areas of interest:

  1. Views
  2. Likes
  3. Dislikes
  4. Published On
  5. Comment Numbers
  6. Reaction to those comments
  7. Comments on comments

On the YouTube platform, these are the metrics that are available to the non-social science data scraping process. Researchers with no data programming skills are able to extract these data.

Persona Construction

Persona construction is a research approach that is based in human-computer interaction (HCI), user-centred design (UCD) and user-experience (UX). Emerging from the Design Thinking field which is human-centred to solve problems, persona construction is useful to understand how problems can be addressed between human and machine interaction. “Design Thinking is an iterative process in which knowledge is constantly being questioned and acquired so it can help us redefine a problem in an attempt to identify alternative strategies and solutions that might not be instantly apparent with our initial level of understanding” (Interaction Design, n.p.). It can have between 3 and seven stages, but these stages are not sequential or hierarchical, but rather iterative and the process typically does not abide to the dominant or common approaches of problem solving methods.

There are 5 phases in Design Thinking:

  1. Empathise – with your users
  2. Define – your user’s needs, their problem, and your insights
  3. Ideate – by challenging assumptions and creating ideas for innovative solutions
  4. Prototype – to start creating solutions
  5. Test – solutions

Persona Construction in Design Thinking is in the second phase of the process, which enables the researcher to define user needs and problems alongside one’s insights. There are four types of personas: Goal-directed, Role-based, Engaging, and Fictional personas. The data ethnography methodology uses Fictional Personas which “The personas in the fiction-based perspective are often used to explore design and generate discussion and insights in the field” (Nielsen, 2013, p.16). In this environment, a persona “is represented through a fictional individual, who in turn represents a group of real consumers with similar characteristics” (Miaskiewicz & Kozar, 2011, p. 419). Secondly, and similarly to ethnography, a persona is described in narrative form. This narrative has two goals: (1) to make the persona seem like a real person, and (2) to provide a vivid story concerning the needs of the persona in the context of the product being designed.

In the context of YouTube research, the key criteria for the fictional personas were:

  1. Name
  2. Age, gender
  3. Marital status
  4. Occupation
  5. Hobbies
  6. Technology familiarity
  7. Devices used

To ensure the accuracy of the process, the research was conducted behind the university wall which has a large range of IP addresses. The research was conducted using Firefox under a new username for each persona, the researcher was not in a signed in state for Google or YouTube, a new Google account was created for each persona and the location of user was set by suggesting a phone area code as per their country. Their interests (Hobbies) became the search terms and the algorithmically generated results were recorded in a pre-trained and post-trained state.

Fine Grained Data Scrape

By engaging the persona construction method which reveals insights into how an algorithm will treat its users, or within the context of this research the sorts of results it will recommend, it is then possible to engage in a fine-grained data scrape. A fine grained data scrape is defined as ….[ref]. In this research, it become possible to understand which were the top related videos, which channels were the most viewed, and sorts of networks that emerge around those videos. This process is most useful for not only identifying specific nodes or videos, but also clusters which can be translated into thematic areas, issue publics (Burgess and Matamoros-Fernández, 2016), and audience clusters. I have previously written about the specific social network analysis (SNA) method so I will not go into that detail here, but in order to find these thematic clusters there is a process of data extraction, cleaning and processing which can be followed. SNA is defined as a computational practice “which draws on computational technologies to help identify, aggregate, visualise and interpret people’s social networking and social media data traces” (p.1). In the first instance, I engaged the YouTube Network Analysis Tool (Ref) to extract the network data of related videos to those which returned as popular in the persona construction method – a post trained algorithm state. This digital method tool extracts the data as a Gephi file which can then be manipulated to provide a social network analysis (SNA) across the dataset.

Topic Modelling

The final method to understand how users congregate around popular content on YouTube, and how they communicate about the material, was to engage in topic modelling.

Topic Modelling is the final method which attempts to understand how users talk about certain things in particular ways. Specifically, I was trying to understand how certain topics emerged in relationship to other topics, which can be understood through the Latent Dirichlet Allocation topic modelling approach. Smith and Graham note, “Informally, LDA represents a set of text documents in terms of a mixture of topics that generate words with particular probabilities” through a predetermined number of topics. This provides the researchers with a “heuristic approach that aim[s] to maximise the interpretability and usefulness of the topics”.

For example, if we wanted to find out what are the popular topics that are discussed by a 14 year old Australian boy, we would construct the persona with interests, which in turn become search terms of, bike riding, Lego, Playstation, and Phil and Dan. The top YouTube Channel recommendations for this user before the algorithm training were:

  1. Family Guy
  2. Talk Shows
  3. Trailers
  4. Gordon Ramsey
  5. Joe Rogan
Jonathon_Hutchinson_Digital_Intermediation

Social media audiences consume approximately three percent of the entire amount of content published across platforms (Bärtl, 2018). Of this three percent, a small number of popular digital influencers create that content, for example Casey Neistat, Logan Paul, or Zoella that, arguably, leads to media homogenisation through the limited focus of popular themes and topics. Moreover, platform providers, such as YouTube and Instagram, operate on algorithmic recommender systems such as ‘trending’ and ‘up next’ mechanisms to ensure popular content remains highly visible. While platforms in the digital era exercise a social and political influence, they are largely free from the social, political and cultural constraints applied by regulators on the mass media. Beyond vague community guidelines, there remains very little media policy to ensure that the content produced by digital influencers and amplified by platforms is accurate, diverse to include public interest, or are indeed beneficial. 

This project will research the content production process of automated media systems that engage digital influencers, or leading social media users, who interact with extraordinarily large and commercially oriented audiences. The evidence base will assist in developing theory on contemporary digital media and society, which will consequently shape how communities access public information. Instead of harnessing this knowledge for commercial imperatives, this research project will examine the findings in the context of socially aware digital influencers who occupy similar roles to those found in traditional media organisations. Further, this project will examine how algorithms are making decisions for media consumers based on commercial executions, which are often void of the social awareness associated with public affairs and issues.  

At a time when mass media comes under scrutiny for its involvement in perpetuating misinformation around public issues, accurate media becomes increasingly crucial to the provision of educative material, journalistic independence, media pluralism, and universal access for citizens. At present, media organisations are attempting to repurpose traditional broadcast content on new media platforms, including social media, through automation built on somewhat experimental algorithms. In many cases, these organisations are failing in this new environment, with many automated media attempts appearing more as ‘experimental’. This should be an opportunity for media organisations to rethink how they produce content, and how new informed publics might be brought into being around that content. 

Instead of thinking of automation as a solution to their increasing media environmental pressures, media organisations should be looking toward algorithms to curate and publish informative media for its audiences. This moment provides a unique opportunity to research the contemporary social media environment as media organisations experiment with automated media processes. It also challenges our understanding of automated media through popular vanity metrics such as likes and shares, in what Cunningham and Craig (2017) are calling ‘social media entertainment’. Under this moniker, these scholars highlight the intersection point of social media platforms, content production, and entrepreneurial influencers who commercialise their presence to develop their own self-branded existence. Abidin (2016) refers to these users as digital influencers, to include YouTube and Instagram superstars who demonstrate an unprecedented capacity to manifest new commercially oriented publics. Digital influencers are typically young social media users who commercially create content across a host of social media platforms, which is liked, commented on and shared by millions of fans. It is estimated the top ten 18-24 year old YouTubers are worth $104.3 million collectively (Leather, 2016), indicating a burgeoning new media market. This model of exercising digital influence within automated media systems has potential to translate into the support of an informed public sphere amid a chorus of social media communication noise.  

The research is innovative in a number of ways. Firstly, it is groundbreaking through its approach of collecting and comparing datasets of contemporary social media practice from within the commercial and non-commercial media sectors. Secondly, it theoretically combines media studies, science and technology studies, sociology and internet studies to bolster the emerging field of contemporary life online: an interdisciplinary approach to everyday social media. Thirdly, methodologically it combines traditional qualitative methods such as interviews and focus groups, and blends these with contemporary digital ethnography techniques and emerging social network analysis. Fourth, this research contributes to the emerging field of automation and algorithmic culture, by providing a groundbreaking exploration of data science with traditional audience research: a field of particular importance for media organisations. Finally, the outcomes will provide innovative insights for digital agencies and leading media organisations. 

Aims and Outcomes 

The aims of the project are:  

  1. to understand how digital influencers operate across social media, in both commercial and non-commercial media environments;  
  2. to document how digital media agencies enable digital influencers to create large consumer based publics; 
  3. to examine and understand how algorithms are operating within large-scale media content production; 
  4. to identify how global media is incorporating digital influencer roles and automation (if at all) into their production methodologies; and 
  5. to provide a new theoretical framework, recommendations and a policy tool that enables media organisations to manifest and engage with its audiences on critical public issues.  

The aims will be met by engaging in digital ethnography methods that documents how digital influencers produce content and actively engage with their audiences in an online community. These users are responsible for creating discussion around a number of issues they deem to be important, yet are typically driven by commercial imperatives. These conversations inspired through influencer content production is then compounded by the digital agencies who operate as amplifying agents for those messages, by especially ‘gaming’ the exposure mechanisms of YouTube and Instagram. However, this research will seek to prove that if this model can work in the commercial media environment, can socially aware digital influencers adopt the same techniques. 

The primary research question is:  

  1. how do digital influencers operate to create large consumer based publics?  

The research subquestions are: 

  1. how does automation operate in media content production and distribution? 
  2. how do automated media systems make content distribution decisions based on behavioural assumptions? 
  3. how can media organisations incorporate the successful methods of automation and digital influencers in their publishing practice? 

Background 

Digital influencers are social media users, typically ‘vloggers’ or video bloggers, who create content about products or lifestyles on popular themes including toys, makeup, travel, food and health amongst other subject areas. Increasingly, digital influencers are using a number of social media platforms to build their brand and publish content to their niche and considerably large audiences. This process of content production and distribution is emblematic of digital intermediation through social media platforms that afford individuals to operate in a media ecology, while determined through algorithmic processes. As Gillespie (2014, p.167) notes, algorithms “provide a means to know what there is to know and how to know it, to participate in social and political discourse, and to familiarize ourselves with the publics in which we participate”. At the heart of these algorithmic platforms distributing trending and popular content are the digital influencers who are creating popular, entertaining media and represent the flow of traffic and information between increasingly large audiences. 

Media organisations have been experimenting with both digital influencers and automation to create and distribute its branded content. In many cases, commercial media have employed the services of digital influencers to boost their traditionally produced media content, while deploying, in many ways, crude experiments in automation. Media brands consistently send digital influencers products and services to integrate into their ‘lifestyle’ videos and images. Recommender systems (Striphas, 2015), such as those used for distribution platforms such as Netflix have proved most popular, where content is suggested based on an audience member’s past viewing habits. Recommendation systems have been adopted across a number of media services including Spotify, Apple iTunes, and most news and media websites. The integration of chatbots is also rising, where the most interesting experiment has emerged from the public media sector through the ABC News Chatbot. Ford and Hutchinson (forthcoming) note that the ABC News Chatbot is not only an experiment in automated media systems, but also a process of educating media consumers on how to access crucial information from within a cornucopia of media. 

The key theoretical problem demonstrated in these examples is an asymmetric distribution of agency when automated systems make ‘decisions’ that can be based on flawed normative or behavioural assumptions (Vedder 1999). At worst, there is no possibility to override the automated decision. That is why algorithmic recommendations are sensitive matters and should be explained to users (Tintarev & Masthoff 2015). But explaining and understanding recommendation systems requires deep technical knowledge as the results are produced by a series of complex and often counter-intuitive calculations (Koren et al 2009). Furthermore, recommendations are often the result of more than one algorithm applied in the online and offline processing of consumer behaviour data (Amatriain & Basilico 2015). The asymmetrical relationship this creates between users and media content providers is especially problematic due to the public complexion and social responsibility obligations that should be demonstrated by media organisations. 

Digital influencers as cultural intermediaries are tastemakers that operate across traditional media platforms such as television and radio, and have become more effective at their translation ability across social media platforms such as Instagram, Twitter and Vine for example. Digital intermediation is the next phase of this research, which builds on cultural intermediation, yet focuses on its relationship with automated media systems. 

Original by Ari He on Unsplash