Notes

5.3 Bias Notes

  • Bias is all about breaking down our opinions
  • There is much bias implemented by most huge companies, even if some goes unnoticed or is unintnetional.

5.4 Crowdsourcing Notes

  • Wikipedia is not actually bad, there is just a negative connotation of it because all our teachers tell us that. Anyone can edit it, so it is not a reliable source (or at least they think it isnt). Textbooks are “way more reliable” but really textbooks are only written from the perspective of a few people who wrote it.
  • Wiki is actually a good thing because it utilizes crowdsourcing. Anyone can list their perspective on a subject.
  • Textbooks are bias based on the location and ideals at which they are written, so they are technically bias.

Discuss APIs I have Used

  • The most prevalent API I have used is a Yahoo Weather API that took in the data from millions of cities in order to output the temp, weather, humidity, etc of all these cities. This crowdsources the weather inputs of a diverse range of people and areas.
  • The greatest discovery I have found on github is the ability to connect your repositories to vscode and invite collaborators through github to code on your vscode. This integration is extremely interesting to me. This utilizes crowdsourcing because it allows a vast diversification of members to help you code.

Hacks

5.3 Bias Hacks

  • Facebook vs TikTok:
    • Facebook was the original popular social media platform, but now all its users are grown up, so at this point it is viewed as an old people social media. On the other hand, younger generations prevalently use tiktok, because it is newer and adjusts to their depleting attention spans with endless entertainment.
    • Both Facebook and TikTok have faced criticism for their algorithmic biases and for excluding certain content or demographics. Whether this exclusion is purposeful or not is a matter of debate. However, it is generally considered harmful as it can lead to unequal representation and opportunities for certain groups. The correction of algorithmic biases is important for promoting fairness and inclusion. Whether it is good business or not depends on the company’s priorities and values. I personally think it is though, because they make billions in ad revenue for it.
  • Amazon, Alexa, Google, Apple Siri:
    • The difficulty in detecting accents or young voices in these virtual assistants has been widely reported and has led to criticism of the companies. This could be due to a variety of factors, including the limited diversity of the data used to train the systems or a lack of attention to the issue. Female voices do evidently have a softer tone and connotation in their voice, so it makes sense that they use a female voice everyday. This can be harmful as it can limit diversity for certain groups. Additionally this could demean females themselves because the female voice is viewed as that of an assistant. I think that there is no need to be corrected however, because each company has numerous options for your voice assistant. This is good business and is not harmful at all.
  • Netflix, Google Algorithm:
    • First of all, when you search something up on google, there are only so many links that are put in front of you. Due to this, there is potential for a coder at google to implement their bias into the search results that pop up. There have been allegations that Netflix’s algorithm biases certain content to certain viewers. The purpose of this is to improve the viewing experience for each user. However, this can also result in unequal representation and opportunities for certain content creators. Whether this is harmful or good business is a matter of debate, but it is important for companies to be transparent about their algorithmic processes and ensure that they are fair and inclusive.
  • HP Racism video questions
    • The situation is that the face tracking capability of the HP computer webcam works perfectly on a white woman but malfunctions and does not work with black people.
    • The owner definately does not believe that this malfunction was intentional, you can tell because he clearly jokes around about it being racist.
    • I think this occured due to the failure to use black subjects when testing this technology.
    • I do not believe that this is harmful in any way, because there were no intentions to be racist. HP obviously simply made the mistake of not using colored subjects when testing their facial tracking.
    • This issue should most definately be corrected, and it probably has since this video is from 2009.
    • I think you should have a diverse range of test subjects when creating a technology based on the outside looks of an individual. If we do this everytime without fail, we will never have an issue like this again.

Overall Takeaway

  • Though much of it is certainly not intentional, it is evident that there is much bias in Americas largest companies. This teaches us the lesson to ensure that there is no bias in our code or algorithms.

5.4 Crowdsource Hacks

  • To initiate crowd-sourcing for a class of 150 computer science students we can
    • Divide the class into groups: Depending on the size of the task, divide the class into smaller groups of 4-5 students each and define the problem or task: Clearly state the problem or task you want the students to solve or contribute to
    • We can also Set up a platform and Choose an appropriate platform, such as GitHub, Google Drive or a project management tool, for the students to collaborate and submit their work, furthermore we will Assign each student a specific role such as a project manager, researcher, developer, or designer.
    • Provide guidelines: Provide clear guidelines and instructions for the students on how to collaborate, communicate, and submit their work
    • Regularly monitoring the progress of the students and provide feedback and guidance where necessary. Another important task we can do is evaluate the submissions based on the guidelines and criteria provided, and provide feedback to the students.
  • Crowd-Sourcing for our fitness program. To implement crowd sourcing in a fitness tracking web app we can:
    • Encourage user engagement: Provide incentives for users to share their fitness data, such as challenges, rewards, or recognition on a leaderboard.
    • Collect and store user data by Creating a database or use an API to collect and store user-generated fitness data, such as workout routines, progress, and other metrics.
    • Implement data sharing features: Allow users to share their fitness data with others, either through social media integration, direct sharing with friends, or by making it public to all users.
    • Analyze the crowd-sourced data to provide insights, generate personalized recommendations, or create a community-wide analysis of trends and patterns.
    • Foster a community by encouraging users to engage with one another and provide resources such as forums, discussion boards, or messaging systems to facilitate collaboration and support.