Essay Example on Interaction Biases in Technology Systems

Most people tend to think that machines, particularly smart ones, are unbiased. The evolvement of technology in the 21st century has created notions that self-driven cars are likely to have no preference in life and death decisions between a driver and a random person on the road. Besides, one is expected to assume that the smart systems involved in regular assessment of credits are likely to ignore everything else unrelated to the essential metrics that include income and the credit scores. The machines involved in learning systems are expected to yield the truth due to the unbiased algorithms used as most people think. However, these are just thoughts and not actual since machines are common to biases. The sources of biases in the technology systems are intense and could include the data used to train the systems, the daily interactions with the devices, similarity biases, and the bias those results due to conflicting goals. Although some of the sources of biases may not be noticed, the process of building and deploying intelligent systems will require an understanding on the different forms and sources of biases to allow for the designation process with awareness aimed at avoiding similar potential issues. Therefore, this paper focuses on some of the bothering interaction biases, the elements that cause the bias, and ways that a person could use to transform their reactions in actions. Interaction biases are commonly experienced in the form of presenting bias, machine bias and ranking bias, which could be evaded through the tools that prevent the use of algorithms by selfish advertisers and users but rather encourage the society to avoid any false impression or information that is common in social media platforms and web search engines.

Is your time best spent reading someone else’s essay? Get a 100% original essay FROM A CERTIFIED WRITER!

Order now

Importantly, some systems tend to learn by analyzing the existing examples in bulk and depth as well. However, it is worth noting that some systems tend to learn through the interaction processes. According to Rosenblatt (2016), bias could arise from the preferences of the users that drive the interaction process. One of the interaction biases that I experience on my day-to-day interaction with technology systems is the way I share content on the social media streams. I have had a significant preference for amusing memes and birthday messages. Thus, this has increased my likelihood of seeing such content in my social media sites. Mainly, this implies that an individual's daily interaction can cause the network to become bias in influencing the content that appears first when one opens the site (Rosenblatt, 2016). The second form of prejudice that has proved common is the ranking bias. Mainly, I have witnessed this in a web search engine where the result pages are listed in order of their relevance from top to bottom. In most cases, when in need of urgent solutions to a query, I will use the top-ranked result, which attracts more clicks as opposed to others since it is ranked top implying that it could be the most relevant, which is not usually the case in some cases.

The third form of bias that I have witnessed in my interaction processes with the technology systems is the presentation bias. As Rosenblatt (2016) indicates, this form of bias presents itself in instances where everything that a user sees can get clicks while everything else fails to get any clicks. I have witnessed this when using the recommendation systems. For instance, in a service offering of video streaming to individuals, users are likely to have different recommendations they can browse. However, the number of people that could potentially benefit from the offering is large compared to the hundreds of suggestions offered in the service. Such a bias is likely to affect the new items that users have not seen before, and there is expected to be any usage of data to browse the new sites since the people may not be sure of their effectiveness. Therefore, this shows three of the most common interaction biases that one is likely to experience when handling technology systems in day-to-day life.

Social media is the primary source of news in the globe today. By using social media, the users are exposed to content whose accuracy can be questioned outrightly. Users can access various theories of conspiracy via social media, pseudoscience, and news reports that have been fabricated, as Rosenblatt (2016) asserts. The government and propagators of political propaganda have used this platform for their selfish gain. However, the fact that content with low credibility spreads fast in social media sites implies that people and algorithms behind the platform and vulnerable to any form of manipulation (Baeza-Yates, 2018). The social media is one of the platforms for interaction biases described above especially the bias described that relates to the content that appears first on social media pages, which is dependent on what a user views the most.

The brain is also another platform for interaction bias witnessed in technology systems. The biases resulting from brain processes are usually referred to as cognitive biases. Most people are likely to encounter this form of bias every day. It is worth noting that the brain can deal with the processing of a finite amount of data. Too much incoming stimuli, as Baeza-Yates (2018) indicates, can lead to overload. Therefore, this poses severe implications on the quality of information available in the social media platforms. Consequently, there exists an intense competition on the limited attention of the users, which means that some ideas are likely to go viral regardless of their low credibility even if the people who share such information have a preference for contents of high quality. In an attempt to overcome this form of bias, I use some tricks that have proved useful but could become a source of bias when misapplied. In my daily search for credible and relevant information, I use Fakey, which is a literacy game for smartphones. The application contains social media news and news articles from both low and highly credible sources. A user who shares reliable information gets points. Mainly, I use this app to ensure that the content I share on the social media pages is credible since it fosters me to conduct further research on the information before making any judgments and before sharing it as well.

The society is also involved in the biases witnessed in the technology systems in one way or another. The direct connection of individuals with peers leads to social biases that guide their selection of friends and also influence the information seen on social media sites. For instance, when selecting the best app or platform that offers the best video-streaming services, a user is likely to read the reviews and choose the platform that has at least one of their friends who has rated the platform and even commented about it. Similarly, it is possible to determine the partisan preferences of a social media user by looking at the friends that the user connects with (Baeza-Yates, 2018). Such a tendency to evaluate information favorably if it comes from one's social circles creates what Baeza-Yates (2018) terms as an echo-chamber, which attract manipulation of different sorts. One way that has proved effective in helping me to maneuver through this form of bias is my need to learn various new things such as applications and platforms that my social circle may not be aware of and my desire to do things differently. In some way, this has helped prevent society from influencing my interaction processes with social media sites.

Machine bias is another common source of interaction bias in the world today. The interaction biases tend to arise from the algorithms that are used to determine what people see online. Social media platforms and internet search engines often use machine biases. In most cases, the personalization technologies are usually designed to select the engaging and relevant content for each of the individual user. However, it could reinforce cognitive and bias from society. Most of the advertisement tools in social media, for example, are usually designed in a way that seeks to exploit confirmation bias by ensuring that the messages are channeled to individuals already inclined to them. The social media platforms and most of the web search engines often expose users to less diverse sources that they could use for their desired uses. Computer programs commonly referred to as social bots interact with people through different social media platforms. Some social bots ate used for disinformation and creating false appearances and propaganda.

The digital platform comprises of algorithms that are used to execute the variety of software hence allowing people to interact and work together (Baeza-Yates, 2018). The operating system of Microsoft is one of the platforms that will enable the designers of the computers to build machines through the platform that would have a role to play in interfacing with software created by the developers. The application software allows most users to carry out a variety of useful functions. The 21st century has witnessed the growth of online platforms that include Facebook, Google, LinkedIn, Uber, and Amazon. Such software has enabled different markets to interact. The platforms have some linking functions that they have to perform. Mainly, this implies that they have the ability to structure the activity carried out by a human being. In addition to that, these platforms are not neutral, and the decisions are usually made by the designers for their selfish gain as had earlier been mentioned. They are generally designed to undertake different activities that are specifically designed to target some users.

A variety of theories exist on the correspondence between people and things built into the infrastructure of algorithms and the different online platforms. The arguments are usually used as filters in most of the online platforms. As a result, they are likely to shift an individual's perception of the phenomena or occurrence described. For instances, the algorithmic choices for special presentations in any advertisement or promotion have a role to play in influencing the decisions of the users and limits them by validating them as chosen scientifically. I have learned to avoid machine biases by identifying a possible reason behind any information present on social media platforms and web search engines. As Baeza-Yates (2018) indicates, if an idea or decision lacks an explanation, chances for bias is high and makes it difficult to prove any form of bias allegations.

Further, one standard solution that could help in solving any form of interaction bias is exploration and exploitation. As Baeza-Yates (2018) asserts, interaction bias tends to affect internet users in the information they receive and new entrants for any applications or other service offering software. By conducting intensive exploration and exploitation, a user is likely to identify that user traffic has been used and intermingled with top recommendations for them to explore. However, this solution has an underlying paradox. For instance, exploration could mean a loss or a cost of opportunity for exploiting the already known information. In other cases, there could be a loss of revenue, especially for digital advertisements. The only way to learn and discover new information present on the web is through exploration and exploitation. Consequently, technology designers ought to come up with tools that are unbiased and seek to remove any form of interaction biases by allowing a user to explore and exploit the web search engines without any interference from designers with ads that have a selfish motive.