Disclaimer: This is an assignment for the course “Digital Media – Critical Perspective” at Goldsmiths, University of London, from September 2023 to January 2024. This assignment is dedicated to answer the question shown in the title, and the examiner feedbacked with positive comments. I am aware the imperfections of this essay, but there is no revise plan for this this one at the moment.
“Do you have some feedback to share?” may be one of the most heard sentences when I was doing my previous job, which, from my perspective, is a process of encouraging team communications, building disciplinary surveillance, and reflecting on ourselves.
This essay considers feedback as self-reflecting, self-referencing, self-creating and self-connecting through discussions about Norbert Wiener and Gilbert Simondon. In addition, this essay investigates the popular online video-uploading platforms as examples and analyses multiple feedback mechanisms of multiple roles and elements that take place within the operation of the platform to investigate how the idea of feedback challenges the cultural creation with digital technology.
Cybernetics as feedback in machines
The concept of feedback is more than seeking advice from your colleagues in business sectors. It has a long history, applied to analogue or mechanical operators invented in the 18th century and digital objects blooming in recent decades.
Cybernetics is probability one of the most famous concepts that connect to feedback. It is not a modern invention that we may have imagined but instead has a root in ancient Greek – “kybernetics”- used to denote the steersman, governor, or pilot (Muri, 2008). Also, in the philosophical context, Plato picked the word in his work of Gorgias to referring skills of navigating and rhetoric, a technique of steering and control, in his view, which is to control and emphasise the concept of information feedback (Dicks, 2020).
The modern concept of cybernetics does not deviate from its meaning in Plato and is still related to the technique of steering and control but has expanded to a much larger context – the human-machine relationships in the context of both thermodynamic machines and digital-network machines. Norbert Wiener, one of the most profound figures in the history of cybernetics nowadays, learnt from James Watt, André-Marie Ampère, Ștefan Odobleja, Claude Shannon and many others. Francis Heylighen summarised the foundational principles of cybernetics – selective retention, autocatalytic growth, asymmetric transitions, blind variation, recursive systems construction, selective variety, requisite knowledge and incomplete knowledge (1991). From these principles, a cybernetical system is unlike a mechanical system, which operates under a trunk-branch structure, and the whole is equal to the sum of sections but owns a native structure that oversees the whole from the beginning, trying to reveal the mode at the structural whole level (Heylighen, 1991).
Influenced by neurobiology, the idea of feedback in cybernetics, as one of its critical methodologies, looks at the discrepancies between the expected movement and the actual movement when reaching something and takes the discrepancy as a new input to adjust the new output (Juan Manuel Espinosa-Sanchez et al., 2023). For instance, when we are trying to reach out our hands to grab a mug on a table but fail for the first time, the distance between our hand and the mug will be taken into consideration as the input, and we adjust our movement next time when we reach out our hands.
Having the feedback as a methodology and the overall supervision ability, cybernetics as a mode of operating thus be able to intake the result generated from itself as a new input and keep involving to achieve the designed output. All that matters is that cybernetics is not a physical matter; rather, it is a mode of operating embedded into complex machines that have a physical form and enable autonomous and self-adjustment towards the pre-set goals for the machine.
Associated milieu as which feedback creates
Cybernetics give us a straightforward understanding of feedback in the context of control and reaching goals. However, feedback could have a more subtle interpretation beyond a controlling mechanism. This essay would like to bring out the French philosopher Gilbert Simondon as another example explaining the idea of feedback, for which the feedback is a kind of recurrent causality (Hui, 2016), enabling the possibilities of creating the very own environment for the object that has the ability of feedback.
Simondon uses the guimbal turbine invented by a French inventor, Jean-Claude Guimbal, in the 1950s as an example (Iliadis, 2015). The guimbal turbine is used to generate electricity, but it is the first machine of its kind that could be installed underwater due to a crankcase and the pressure oil that covers the turbine (Iliadis, 2015).
In addition, and the most unique is that, unlike other turbines that could be malfunctioned due to the heating generated during the operation, the guimbal turbine utilises the water to that employed to generate electricity to cool itself down. The more water flows, the more power it generates; the more heat it generates, the more heat is taken away by water – it could always be in a stable status. In this occasion, the turbine is incorporating water as a geographical environment into its operation not only to complete the designed work but also maintains the conditions of working, thus creating what Simondon has described as an “associated milieu”, combining the natural environment as well as the technological environment of itself, which in Simondon’s process of individuation, becomes the technical individuals from isolated technical elements (Hui, 2016).
The feedback here is bilateral and circulative, just as Simondon has indicated: recurrent. At the most obvious level of feedback, the input is the kinetic energy potentially contained in the flowing water, and the output is the electricity generated by the spinning turbine. By arguing the bilateral here, this essay suggests that the turbine does not take everything from the water for granted, absorbing everything that feeds into it and doing its best to complete what it is designed to do. Instead, it generates a sort of surplus from its operation. The most intriguing thing here is that the turbine, or the “technical individuals” in Simondon’s words, is not adding entropy into its own mechanism but utilises the surrounding environment, the environment that itself relied on to digest this entropy in order to maintain its normal operation. The water and the turbine here are both the source of output and the recipient of input, creating a higher feedback level. Moreover, the whole process starts from the very first moment it starts working and never ends unless it is taken out of the water. Also, no more interventions or manipulations are required to unfold this ability – it is already built into the structure of the turbine, which is just waiting to be activated by the working conditions it relies on – it is a kind of potential to emerge.
From Web 1.0 to Web 2.0
In China, a widely-used video platform among younger generations is called Bilibili (https://www.bilibili.com/). It has functionality similar to YouTube, enabling users to upload videos they make and get paid from the platform or third-party companies based on followers, clicks, likes and saves. Such platforms have been widely studied, stating that they have created a new model of the online economy.
These platforms gradually evolved from Web 2.0, which emerged at the beginning of this century, transforming the traditional webpage from static to updatable, interactive webpages and putting the customers from the content consumer into the position of the content creator and further using their creations as an entertainment product to draw attention for the wider public to gain growing web traffic, therefore promising a more wide-spread and effective advertising and more gaining more revenues from companies that wish to promote them.
This mode of making profits by integrating content creators, users, and other companies within an online platform forms a complex feedback system that combines all kinds of feedback I discussed above, and this essay will discuss it later as a whole.
The role of content creators in the video platform
For now, let us focus on the content creators’ side as they provide videos as input for the platform[1]. They expect to be rewarded with financial incentives and receive communal encouragement or any other comments as output for their work, which is the starting point for the feedback that follows the structures of enhancing the existing standards, which in one way is restraining the videos created by uploaders but also contributing to the dominance of the platform.
However, before diving into the analysis, it is crucial to clarify what “content creators” or “uploaders” are when referring to them. This essay denotes those people who regard making videos as if this is his/her career and take the standard criteria of being successful, meaning to have as many followers as possible and make as much profit as possible, which is not following what Bernard Stigler has said about amateurs (Bernard et al., 2016), a kind of contributional economy.
Take the idea of feedback in neurobiology. Suppose the outcome of the first upload did not meet the expectations of an ambitious content creator. For example, the numbers of clicks, likes, and subscription rates are low, people watching videos from the start to the end are rare, comments are saying that the video is boring or does not make any sense and so on. Given the outcomes, the uploader might want to adjust their videos in terms of time durations, editing skills and topic choices based on either positive or negative output generated from the previous input as a new series of inputs for the next time they upload new videos.
In this scenario, the feedback takes place within the web-host, updatable, and interactive sphere that the online video platform provides, where it serves as a massive reference database where already successful content providers also upload their works. For newcomers, they may be the golden examples of how popular videos should look like, along with the feedback that is given from the video viewers mentioned above, forming a mechanism that how the feedback is challenging the cultural creation in such platforms – in one way, they are imposing a protocol to successive new content creators, shaping and regulating the forms and contents of their creations, but also transforming people that use the platform into an organic cybernetic.
The role of the video platform itself
At this stage, we have to turn to the platform itself. This essay argues that the platform is creating an associated milieu through multiple feedback mechanisms operated within it. To begin with, simply analogise it as a profit generator for its shareholders and compare it to what Simondon has described for the guimbal turbine. In such an analogy, the water that pushes the turbine to spin has become the newly created videos made by uploaders; the platform relies on those flows of videos to attract as many public masses to spend their time watching videos as possible and hence be able to promise better advertising effects for other companies and thus generate as much revenue as possible.
Similar to turbines that may malfunction due to overheating, the platforms may also face problems such that there are too many videos uploaded into the platform, leading to not only homogenisation but also expending too many hardware and network resources that merely add up the cost and do not generate revenues. The guimbal turbine resolves the overheating problem by adding a layer of pressure oil and crankcase and putting itself in water. At the same time, the platform equivalently incorporates the power of personalised algorithms under the structure of protocol to fix the problem, further creating multiple feedbacks within the system to suffice its operation.
By referring to the protocol, this essay takes the ideas from Alexander Galloway (Galloway, 2006), meaning that the protocol is made for Deleuze’s society of control (Deleuze, 1992) by not imposing corporal violence nor disciplinary constraints, but the ability of productive computations, licensing the access of information and data in a distributed and seemly-nonhierarchical environment. More importantly, the protocol is an apparatus that runs on the level of desire (Galloway, 2006) that gives guidance.
In the case of online video platforms, the protocol could be analysed in two ways. The first one is kind of depicted in the previous argument, articulating that the content creators who create videos are influenced by financial incentives, viewing data and comments as outputs, and they intend to reference those successful uploaders to gain better performance. In the protocol aspect, the platform neither blocks videos made by certain uploaders from being viewed nor requires the content creators to make any specific type of video. All that drives the content creators is the “protocols” of making popular videos, which are contributed by those already popular video and their successful characteristics accompanying the statistics given by people, resulting in the new content makers that voluntarily take in this feed-back since they want to be a successful uploader as well.
The second is driven by the implementation of the algorithms within the platform. It is a key element that the platform is capable of dealing with the booming of uploaded videos and even using it to produce an associated milieu. As more videos feed in, without an algorithm system that personalises the feed page, every user sees the same flow of recommendations unless they subscribe to certain uploaders or have to use the search bar to find videos they are interested in, which is detrimental to users and platforms.
However, once the recommendation algorithm is adequately trained and applied, it could distribute videos to specific groups of viewers who are interested in particular themes of videos – the more videos it has, the more input resources it could be fed into the algorithm and consequently, the more specialised groups of users would be satisfied and further conducting to a growing number of users spending their time on the platform, leading to the more stable financial status of the platform and even more revenues it could make. Here, the algorithm is another protocol that has “virtues include robustness, contingency, inter-operability, flexibility, heterogeneity, and pantheism” (Galloway, 2006), thus being capable of operating within any video online platforms and tolerancing a variety of videos but utilising them for the sake of running the host it hosts on. It has the magic of compatibility and transformation.
It is more than that. The referencing desire and the personalised algorithm are also entangled and diffused, which is another loop of feedback. For one thing, the existing referencing pool provides a normalisation tool that trains the algorithm to filter out those inappropriate videos from being recommended to a greater public since they are not massively viewed or may not comply with the prevailing aesthetics or social norms. For another, the referencing pool is constructed by a considerable amount of data on clicks, likes, shares, subscriptions, time durations and comments of those top videos that were previously captured by algorithms and fed into itself by the code it depends on. Both of these processes do not happen just once but are always happening.
As a result, the “recommended for you” page shows a hybrid configuration of traces of referential feedback and a personalised algorithm. It is easy to admit that the algorithm is a kind of cybernetics machine due to its non-organic form and brain-like mechanism, but how do we find a place for the content makers and users on this platform? The answer is simple: They, as well as we, are becoming a kind of organic cybernetics that is handling out our sensibilities to cybernetic machines that “help” us define our lives and control what we see and how we perceive. This is a hidden aspect of cybernetics, in which Wiener is trying to replace the human brain with logical steps (Galloway, 2006); likewise, it is about what Bernard Stigler has said about symbolic misery – a mechanical turn of sensibility – the machine of media is designating the symbolics, thus re-directing our sensibilities into the media (Zhang, 2018), or the cybernetic control machines itself.
References
Bernard, Lu, X., & Xu, Y. (2016). 人类纪里的艺术 : 斯蒂格勒中国美院讲座 / Ren lei ji li de yi shu : Si di ge le zhong guo mei yuan jiang zuo. Zhong Qing Da Xue Chu Ban She.
Dicks, H. (2020). What Lies Concealed in the Roots of Cybernetics: The Renewal of Early Greek Thinking of Being as Physis in Martin Heidegger and Edgar Morin. Nottingham French Studies, 59(3), 384–398. https://doi.org/10.3366/nfs.2020.0297
Heylighen, F. (1991). Principles of Systems and Cybernetics: an evolutionary perspective. Cybernetics and Systems 92, 1992, 3–10. https://www.researchgate.net/publication/2441746_Principles_of_Systems_and_Cybernetics_an_evolutionary_perspective
Hui, Y. (2016). On the Existence of Digital Objects. U of Minnesota Press.
Iliadis, A. (2015). Two Examples of Concretization. Platform: Journal of Media and Communication, 6. https://www.academia.edu/12135436/Two_Examples_of_Concretization
Juan Manuel Espinosa-Sanchez, Gomez-Marin, A., & Fernando de Castro. (2023). The Importance of Cajal’s and Lorente de Nó’s Neuroscience to the Birth of Cybernetics. The Neuroscientist, 0(0). https://doi.org/10.1177/10738584231179932
Muri, A. (2008). Enlightenment Cybernetics: Communications and Control in the Man-Machine. The Eighteenth Century, 49(2), 141–163. https://muse.jhu.edu/article/254477
Zhang, Y. (2018). 心灵无产阶级化及其解放途径——斯蒂格勒对当代数字化资本主义的批判 /The Proletarianization of Mind and Its New Way of Liberation——Stiegler’s Critique of Contemporary Digital Capitalism. 探索与争鸣 /Exploration and Free Views, 1, 4-13+1+141. doi:CNKI:SUN:TSZM.0.2018-01-001.
[1] The content creators themselves are another layer of the feedback within this system as they perceive the world as input and make videos as output. Due to this narrative is beyond the scope of the platform, it is not being discussed in this essay.