Proxy Culture

From SI410
Revision as of 11:31, 23 April 2019 by Hbaccouc (Talk | contribs) (Bias)

Jump to: navigation, search

Proxy culture refers to the reliance on documented experiences (signifiers) in place of first-hand experiences (the signified). Proxy culture has roots in the intersection of philosophy and technology and may manifest in many ways, including through the use of Yelp business reviews, Amazon product reviews, and Google Maps to understand businesses, services, products, and more.[1] Proxies in themselves allow individuals to have experiences and interactions that would be otherwise difficult or impossible, although accuracy of these representations through signifiers can be disputed. [1] Proxy culture is interacting with the internet for advice that in the past, we could only get from fellow humans.

Wikipedia is a classic example of proxy culture in practice.

History

Proxy culture is a relatively new and modern term. It was first used widespread by Luciano Floridi in his paper titled "A Proxy Culture"[1] in 2015 to describe the phenomena where people are increasingly more likely to depend on the infosphere to make decisions. Common examples of proxy culture are friending someone on Facebook you have never met, trusting other TripAdvisor strangers for recommendations and using a electronic map to get to your destination.

Luciano Floridi

Luciano Floridi is a Professor of Philosophy and Ethics of Information at the University of Oxford. He has several recognitions and awards for his work relating to information ethics.[2] Floridi approaches the philosophy of Information from a multifaceted point of view. In his studies he takes into account logic, epistemology, computer science, IT, and humanities computing.

Luciano Floridi

Definition and Etymology

Proxy by itself is primarily defined as "the agency, function, or office of a deputy who acts as a substitute for another."[3] The first known use of the term was in the 15th century rooted in Latin cultures.

According to Floridi in his aforementioned paper, the concept and idea of a proxy is not quite the same but has many similarities. The concept is rooted religiously, while the idea is political.

Today, a proxy itself is mostly related to technology instead. The purely technical term is related to systems where different signals and requests are used, such as a website interacting with the internet.[1] Proxy culture looks at the human component and how they interact with signifiers to better understand the signified.

Examples in Popular Culture

Influencers are people who have a large following on social media and other internet platforms. People often form parasocial relationships with these influencers, where they come to see these public figures as personal friends. In order to exploit these relationships for monetary value, companies often offer social media influencers deals and sponsorships in order to gain favor with their audience. Though there is growing policy about transparency with sponsored content on social media platforms, sponsored content is still a way in which companies take advantage of proxy culture for advertisement. If someone's favorite YouTuber or Instagrammer is singing the praises of a product, viewers will be very likely to give it a try, too.
College Confidential is a website where people ask advice about their academic records in regards to applying to colleges, graduate schools and other various things within the academic world. People are trusting strangers with extremely personal information, and entrusting their future academic careers in college confidential users. People ask questions like "am I qualified" or "is this school a good fit for me" to people that know nothing about them besides the information provided. Many students and parents turn to college confidential, and more recently college admission subreddits for information instead of an actual college guidance counselor.

Degenerate Proxies

Degenerate proxies are epistemologically separate from proxies because the interactions with a proxy is similar to interacting with the signified. Degenerate proxies fulfill some, but not all, definitions of a full proxy. These degenerate proxies can be categorized in three forms: icons, indexes, and symbols. They are not full proxies because they are merely a standby.[1]

Icons Icons aesthetically resemble what they represent. For example, a photograph of a dog is an icon for a dog. This is not a full proxy because what happens to the photograph of the dog doesn't directly effect the dog.

Indexes Indexes are correlational, meaning a given index is often associated with what they represent. An example of this is how dark clouds are an index for rain because those often follow.

Symbols Symbols represent a category they're typically associated with. A drop of water is a symbol for liquids, for example.

Ethics

Monopolies

Powerful entities who are in charge of information dissemination have control over proxy culture by curating content. Thus, these entities could unfairly demote or promote products that solely benefit themselves. For example, Microsoft was accused and prosecuted by the U.S. Justice Department of creating a monopoly for internet browsers by not allowing its operating system users to download the competing internet browsers, Netscape and Opera.[4] Due to being rules as breaking antitrust law, this resulted in Microsoft being broken up into separate software and hardware units. Recently, there has been much controversy over weather Amazon is a monopoly as well, and if a similar action should be taken against them.

Monopolies in proxy culture are ethically concerning because they do not promote healthy competition, innovation, and fairness. In the extreme case where there is only one entity in control of an entire market having no competitors would effectively allow that entity to raise their prices and lower the quality of their products without any threats. The producer has no incentive to innovate and this forces consumers who rely on these monopolistic market to likely use or pay more for subpar products. Without a monopoly, the consumer would have more choice to pick the best product and entities continually compete to create quality products.

Mitigation

Despite how monopolies may utilize proxy culture to damage rivals, studies have shown that this may be difficult due to the collective action nature of proxy culture. Any user may contribute to proxy culture, yet more credible and frequent editors often reduce potential damage caused by negative competition. [5]

Credibility

Bias

People who have particularly negative or positive experiences are inherently more inclined to review a product, meaning a product that is otherwise fine may be skewed to be seen too negatively or positively due to proxy culture contributions. Feedback is also skewed by the demographic offering it. This creates barriers for alternative opinions or information to contribute to proxy culture's representations. For example, there is an extreme gender bias in Wikipedia, with only about 10-15% of Wikipedia contributions coming from women.[6] Despite men and women having approximately the same number of articles on Wikipedia, typically either men write about women or a small group of women overstretch themselves to compensate for their lack of representation. This means that female-representative information is more likely to carry men's imposed perspectives. The bias takes place in a positive feedback cycle, as the lack of gender representation on Wikipedia may deter gender minorities who would otherwise edit on the site from offering their unbiased perspectives.

Anonymity

Reviewers are often anonymous, allowing participants to be anonymized and feel more comfortable--they feel less obligations to use a filter on their opinions because their words stand on their own, rather than being attached to their gender, race, and income level.[5] This is positive because people with limited information on others may use those indicators to determine trustworthiness.[7] It may also facilitate more vulnerability and honesty since participants would be difficult to track and thus can worry less about repercussions.

This can be detrimental when that anonymity is abused to harass others or feign credibility. Because participants' identities are difficult to discover, they have less fears about verbally abusing or threatening others. This can create apathy within a community and scare or hurt people unfairly. The quality of the content they contribute to proxy culture doesn't have large stakes, so they may give inaccurate representations that do not represent their experiences.

Falsification

Competitors may often sabotage others in order to gain an upper hand. For example, mobile developers post falsified negative ratings on rival apps to damage competitor reputation. This works the opposite way as well, with many committing fraud by inflating app sales or posting falsified app ratings for themselves.[8] To try to pin blame on competitors, a group may post several positive reviews on a competitor's product to get them flagged for fraudulent behavior.

Reddit often employs bots to moderate their forums, and users can develop their own for other purposes.

The scope of this behavior isn't limited to apps, as Amazon has come under fire for allowing false reviews and shady brands to be present on their website with little to no regulation. Likewise, Yelp is notorious for having review wars amongst competitors to drive down ratings of competing restaurants or services.[9] This has led to the rise of websites and tools such as fakespot.com, an artificial intelligence tool that utilizes linguistic analysis and comparisons of similar items to identify fake reviews on sites such as Amazon, Yelp, Walmart, and Steam, and offer consumers adjusted review grades that reflect those deemed as genuine.[10]

Ethically, this poses the issue of competitors abusing review systems to gain unfair advantages in a market. Furthermore, having dishonest reviews in systems hinders consumer trust even if the majority are honest. This damages an otherwise effective way to use proxy culture to understand a given product.

Bots

Proxy culture involves a ton of data that needs to be vetted, which is often difficult to do with humans alone. On open-source knowledge bases, they're often employed to combat an influx of malicious attacks--using machine learning and artificial intelligence have been developed to monitor quality of crowd sourced content or remove questionable posts. More convincing but still questionable content may pass a bot, but still need to bypass a human moderator to become published. These bots are often non-intrusive, and also serve other functions such as language translation[11]. Overall, they reduce the burden on manual labor and improve efficiency for large-scale sites.

However, they can also make mistakes if they aren't trained for a certain case. Proxy culture involves using a corpus of information to create representations, but machine learning may generate inaccurate ones; this happens when representations are created from biased data and causes ethical issues. For example, Amazon attempted to create a recruiting tool to rate employee applicants using old data to create proxy culture representations of these people.[12] This tool ended up unfairly flagging certain characteristics on resumes as less employable, such as enrollment at a female university or being involved in all female activities, despite unfound correlations.

Virtual Reality

Inside of his most popular work, "The Handbook of Information and Computer Ethics", Phillip Brey dedicates an entire chapter to a number of emerging ethical issues that have accompanied rising virtual reality technologies. [13] The definition that Brey sets aside for virtual reality has numerous overlaps with the characteristics found in proxy culture, including artificial representations, detachment, and disinvestment in real-world events. Brey uses altered digital media such as photos and videos as an example of the ways in which technology can be used to fabricate nearly indistinguishable representations of reality. These representations as a substitute for reality may lead people to over-engage in virtual environments and consequently disengage in real life interactions. The ability to garner control over virtual reality representations may win users over in place of real-world interactions.[13]

See Also

References

  1. 1.0 1.1 1.2 1.3 1.4 Floridi, Luciano. “A Proxy Culture.” SpringerLink, Springer Netherlands, 21 Oct. 2015, link.springer.com/article/10.1007/s13347-015-0209-8.
  2. Oxford Internet Institute. “Professor Luciano Floridi — Oxford Internet Institute”, https://www.oii.ox.ac.uk/people/luciano-floridi/.
  3. “Proxy.” Merriam-Webster, Merriam-Webster, www.merriam-webster.com/dictionary/proxy.
  4. “Digital wars: Apple, Google, Microsoft and the battle for the Internet..” Arthur, Charles. Kogan Page Publishers, 2014. https://books.google.com/books?hl=en&lr=&id=YlZpAwAAQBAJ&oi=fnd&pg=PR5&dq=internet+monopoly+google&ots=AomNG3vOL8&sig=wKAqj5MjQriX4raGRhwvIAjKwZQ#v=onepage&q=microsoft%20monopoly&f=false.
  5. 5.0 5.1 Priedhorsky, Reid, et al. 2007. "Creating, Destroying, and Restoring Value in Wikipedia." Group’07 November 4-7, Sanibel Island, Florida.
  6. Torres, Nicole. "Why Do So Few Women Edit Wikipedia". 02 Jun 2016. Harvard Business Review. https://hbr.org/2016/06/why-do-so-few-women-edit-wikipedia
  7. Stanley, Damian A., et al. "Race and reputation: perceived racial group trustworthiness influences the neural correlates of trust decisions." Philosophical Transactions of the Royal Society B: Biological Sciences 367.1589 (2012): 744-753.
  8. Zhu, Hengshu, et al. "Discovery of ranking fraud for mobile apps." IEEE Transactions on knowledge and data engineering 27.1 (2015): 74-87.
  9. Penaflorida, Rexly. "Everything You Need to Know About Fake Yelp Reviews." ReviewTrackers. 10 January 2019. https://www.reviewtrackers.com/fake-yelp-reviews/
  10. "Fakespot | Analyze and identify fake reviews." fakespot.com
  11. Nasaw, Daniel. [www.bbc.com/news/magazine-18892510/ “Meet the 'Bots' That Edit Wikipedia.”] BBC News, BBC, 25 July 2012, www.bbc.com/news/magazine-18892510.
  12. Dastin, Jeffrey. "Amazon scraps AI recruiting tool that showed bias against women". 09 Oct 2018. Reuters. https://www.reuters.com/article/us-amazon-com-jobs-automation-insight/amazon-scraps-secret-ai-recruiting-tool-that-showed-bias-against-women-idUSKCN1MK08G
  13. 13.0 13.1 Brey, Philip. “Disclosive Computer Ethics.” ACM SIGCAS Computers and Society, vol. 30, no. 4, 2000, p. 10., doi:10.1145/572260.572264.