At any time prior to the end of Week 10, you may submit an essay of between 1500 and 2100 words (approximately 5-7 double-spaced pages) on a topic broadly related to the themes of algorithms, equity, and justice. Your essay should address a concrete problem; assess the current state of affairs; and suggest one or more paths forward.

Examples of suitable topics related to algorithms and justice include but are no means limited to bias in facial recognition; the impact of automation on social services for the poor; and the role of technology in combatting climate change. Please note that questions of equity and justice are more specific than the general impact of social media or tech companies on society as a whole. Essays on such general topics are likely to receive only partial credit.

Somewhat more generally, a successful essay will discuss:

  • Concrete benefits or harms. Examples of concrete benefits or harms include changes to one’s likelihood of being arrested, granted health services, approved for a loan, or granted reasonably-priced health insurance.
  • Disparate impacts. You should show that the system in question harms or benefits certain groups more than others.

A bad example: arguing that social media is driving political polarization (this is a good topic, but not for this prompt).

A better example: arguing that polarization on social media is leading to an increase in hate crimes toward specific marginalized groups.

You are welcome to argue any angle on your topic, provided that you can offer sound sources and argumentation. For example, you may argue that bias in facial recognition is a major problem in need of solution, or you may argue that no substantial problem exists. You are welcome to ask me about your sources and proposed argument before you begin writing.

This essay is specs-graded, just like other assignments. If your essay doesn’t meet specs the first time, I’ll give you some feedback and you’ll have the chance to revise your essay prior to the end of the quarter.

Specifications

Sources

The essay should cite and elaborate on at least 10 distinct sources.

  • At least three of these sources must be peer-reviewed, scholarly articles or published scholarly books.
  • News articles, blog posts by accredited scholars of machine learning and data ethics, and documentary videos are acceptable for the remaining sources.

You should use a consistent citation convention. If you’re not sure which one to use, APA and MLA are both good choices.

Argument

The essay should have a thesis statement, which is supported by the discussion of the sources. Several examples of successful thesis statements: - Taken together, the evidence suggest that algorithmic surveillance of marginalized groups is only increasing with time. - Although studies have revealed significant biases in Google Search, more recent work suggests that these biases may be less severe than once believed. - While many proposals exist for how algorithms and computation can be used to combat climate change, these proposals often overlook the role of human interpretation in implementing computer-generating guidance.

Structure

  • The thesis statement should be stated at the end of either the first or second paragraph of the essay.
  • There should be at least one paragraph, at the end of the essay, which synthesizes the discussion and discusses the future outlook.
  • The paragraphs in between (the “body paragraphs”) should have clear purposes which support the argument for the thesis statement.

Grammar and Clarity

I recognize that English language writing may pose greater challenges to some students than others. Provided that your meaning is clear, a reasonable number of grammar and spelling mistakes are acceptable. Please try your best to proofread your work prior to submission. The Undergraduate Writing Center is an excellent resource for helping you craft high-quality essays, no matter how much experience you have with English language writing.

Some Suggestions for Getting Started

You are free to use any sources to which you have access. If you are having trouble getting started, the following may be useful starting places.

  • The Algorithmic Justice League maintains a library of articles and announcements that may help you find suitable material. Especially relevant for topics at the intersection of algorithms and racial justice.
  • A list of articles on bias in algorithms. This list is curated by Professor Safiya Umoja Noble, author of Algorithms of Oppression: How Search Engines Reinforce Racism and faculty here at UCLA. Note: Algorithms of Oppression is available through the UCLA library.
  • Cathy O’Neil, author of the excellent book Weapons of Math Destruction (available through the UCLA library), has a blog. Many of her posts are opinion pieces (and therefore unsuitable for citation), but she always provides links to primary material, which you can use.
  • The book Automating Inequality by Virginia Eubanks is an excellent discussion of the impact of automation on marginalized populations in the US. It is available through the UCLA library UCLA library.
  • This Twitter thread by Professor Emily M. Bender at the University of Washington includes a number of great sources to help you get started.