Mini-project 2: Infrastructural design

Goal

This project has two goals:

• To give you hands-on practice in making design interventions at the level of infrastructure (code and algorithms that structure many people's practices).
• To develop skills in expressing ideas about social impact in the form of an op-ed which can have wide influence.

MATERIALS FROM SECTION

Here are some materials that we will be using to help shape MP2:

1. What is an oped made of?
2. Sample oped.
3. Oped creation worksheet.
4. Oped creation worksheet_online.

Overview

Contemporary life in the US and many other nations is deeply influenced by computational platforms: code and algorithms that shape what we do, how we communicate, and what we know. In this miniproject, you will make an argument for a change to the design of a contemporary computational platforms with widespread effect, which will result (according to your argument) in an improvement in the social impact of that platform. The change will be at the level of code, i.e. a technical change to the algorithms on which the system relies, rather than, for example, a change to how the system is advertised and promoted.

You will make this argument in the form of an op-ed, i.e. a brief, structured argument that is accessible to, and influential for, a wide audience.

Skills for this miniproject will be covered in section on Oct. 8 (writing an op-ed) and Oct. 15 (making a technical change to an infrastructure). Please review also your lecture notes from Sept. 30, Oct. 5, Oct. 7, and the associated readings for more background.

Instructions

1. Brainstorm a list of computational platforms which you feel are having an impact on a significant group in contemporary society.
2. Ask yourself, "If I had the power, how would I change the impact of this platform?"
3. Comparing your desired impact to what you believe the platform currently does, brainstorm possible design changes that would alter that impact. For example, I might decide that Facebook is causing political polarization by catching people in filter bubbles where they only interact with people of similar political viewpoints. If so, I could change the design of Facebook so that it automatically identifies the political viewpoint of its users, and then makes sure half of my feed had posts from people whose viewpoint is significantly different from mine.
4. Sleep on your initial ideas.
5. Review your ideas, selecting a few which you feel are the most compelling.
6. Refine your proposed ideas, with an eye to making sure (1) your idea is original, (2) an average reader would consider it technically plausible, and (3) it would be likely to have the impact you want. In the previous example, I might decide that Facebook would calculate user political viewpoints by checking the sources of the news articles that the user posts, and that I would reduce the amount of the feed artificially taken up by people with different viewpoints to 10 per cent so that users would not become alienated from the platform.
7. Select your best idea, and write an op-ed that advocates for your design change. Your op-ed must make a simple, clear, coherent argument for the value of your design change, i.e. why this change would be effective in bringing about a better social impact. Your argument should be compelling for a broad audience.

Please study carefully the appendix below for more information about how to write op-eds. This will provide more details on what we will be looking for in your assignment.

Submission

Deliverables:

1. Your op-ed. It must include the following components:
   1. A lede or opener which draws the reader into your argument
   2. A clear, easy to understand description of the original design change you are advocating for.
   3. 3 distinct points which support your argument that this design change would create the particular positive impact on society that you intend. Each point is undergirded with concrete evidence.
   4. Acknowledgement of 1 or more obvious counterarguments, with effective argumentation disarming each.
   5. A conclusion which includes a clear recommendation for what should be done differently.
   Your op-ed has a strict word limit of 700 words. Any additional words will not be graded. You can check how many words we will think your op-ed has by using the word counter in Google Docs, under Tools->Word Count. Your project is due electronically via submission through Gradescope by 11:59pm on Sun, Oct 17. Your submission must be in *.pdf format* - you may not use image submission. In line with standard newspaper submission practices, for this miniproject your document must be plain text (no images and no handwriting).

   Appendix I: Op-ed writing

   (Adapted from: The OpEd Project; this is a great resource!)

   An op-ed is a short argument to a public audience, that newspapers typically printed "opposite the editorial page" (hence: op-ed). Op-eds can be influential in that they can sway readers to support a position that you care about. To do so, they need to be carefully crafted to speak understandably to a wide audience.The rhetorical strategies of an op-ed are useful for many kinds of persuasive public writing.

   An op-ed typically consists of the following components:

   1. A lede

   This is the opener of the op-ed. It needs to be short and punchy to grab the reader’s attention, for example through a compelling example, a surprising statement, or a dramatic anecdote. The lede must draw in the reader, so they want to read the op-ed instead of moving on to the next article. Usually but not always ledes tie to contemporary events: something that happened recently, either in the news or (less frequently) in your own life.

   2. A thesis

   Op-eds are short pieces of accessible public writing; they are not the place to make complex, nuanced arguments. Instead, op-eds must be anchored in a clear, unequivocal position you are taking on an issue. This is what is called the ‘thesis’ – the single main point the argument is going to make. Your thesis may be written explicitly into the op-ed, or it may be the conclusion that the reader comes to after they read it. Whether or not it is explicitly written into the op-ed, it must be crystal-clear in the author’s head while writing it, and in the reader's head after reading it.

   3. An argument grounded in evidence

   The thesis’s validity will be undergirded by several points (in the real world, usually 2-4; for this assignment, 3). Each point must draw on credible evidence, such as: statistics, news, reports from credible organizations, expert quotes, scholarship, history, or first-hand experience. Each point must contribute directly towards supporting the overall thesis.

   As is true in all forms of writing in this course, text and ideas from others must be properly quoted and cited in your op-ed. However, op-eds do not use formal academic citation. They attribute sources by naming them within the text ("A recent study by researchers at Brigham Young University showed...") and/or including URLs. See below for an example of this kind of citation.

   4. “To be sure”

   The argument should pre-empt possible skeptics and critics by acknowledging any flaws or shortcomings in your argument, and addressing any obvious counterarguments. Effective strategies to do this include:
   • Acknowledge and dismiss: Acknowledge that a counterargument exists, but argue why that counterargument is not so important.
   • Validate and trump: Acknowledge that an important counterargument exists, and explain how your argument addresses it.
   • Personal caveat: Address a personal limitation that may come up (e.g. "As CEO of Google, it might seem that I am biased in my view on algorithmic fairness") and explain why you have credibility anyway

   The "to be sure" must be written in a tone that will speak to potential critics, i.e. by trying on the shoes of those who might disagree with you. If you want them to listen to you, you need to make clear you have heard them. The "to be sure" must communicate respect and empathy, based in an imagination of your opponent as intelligent and moral (whether or not this is what you actually believe).

   5. Conclusion

   The end of the op-ed must make a clear recommendation of what should happen differently based on the argument it makes. A weak conclusion argues for a general attitude change (e.g. "people should be aware that food waste is a big problem"). A strong conclusion recommends something concrete that should be done differently, ideally something your reader is in a position to implement or advocate for immediately (e.g. "Designate one day a week to use up leftovers in your refrigerator.")

   Appendix II: Example op-ed

   Note: this was written by Prof. Sengers for a "Bloomberg" type audience. It was not written for this assignment so it does not exactly fit the assignment requirements, but should give you a sense of how op-ed writing and referencing work.

   How much you pay for electricity could change radically under a new proposal in New York State. In the future, your electric bill may depend on how good you are at time management.

   This is because, as a recent Ithaca Journal report uncovered, there are plans to charge radically different prices for electricity depending on the time of day [1]. As part of an experimental ‘smart metering’ system, Avangrid - the parent company of New York State Electric & Gas and Rochester Gas & Electric - is proposing to change the delivery charge for electricity from current flat rates of 4-5 cents per KWH, to 3 cents in off-peak times and a whopping 18 cents in peak times. Under the proposed plan, electrical consumers who are able to reorganize their energy-intensive activities to occur at non-peak times – for example by running the dryer at 11pm – may decrease costs by up to 40 per cent. But consumers who cannot could see their bills double or triple.

   This policy addresses an important issue with our electrical infrastructure. Utilities need to generate enough electricity to meet demand at peak times. As peak demand rises, they have to build more plants, but that extra capacity often stands idle. The idea, then, is that charging people more for electricity when it is in demand will encourage people to shift electrical use to non-peak times. This will let us wring more out of our current infrastructure, rather than having to build more. In addition, shifting people to use electricity when it is in greater supply will help us manage the transition from fossil fuels, which are available 24 hours a day, to wind and solar energy, whose supply varies over the day. Avangrid’s experimental system is a test run of variable charging for electricity that is expected to be the future across the nation.

   While the motivation for this idea is laudable, the likely social implications are not. This is because the cost of variable pricing will be carried by households that are unwilling or unable to systematically rethink the timing of their energy-intensive activities. These are often vulnerable households.

   For example, low-income households will likely take a hard hit. Rearranging one’s schedule of activities simply takes time and thought which people who are already struggling to meet their families’ daily needs may be unable to muster [2]. In addition, lower-income consumers may have less control over factors that influence their electricity use, as Tawanna Dillahunt has shown holds for renters who want to reduce energy use [3]. Finally, planning one’s future time use can be nearly impossible for workers subject to the constantly changing, just-in-time work scheduling practices that are becoming more common in lower-wage sectors [4]. Consumers in any of these situations will find their electrical bills skyrocketing, with little they can practically do about it.

   But the losers under this policy are not just low-income households. Plotting your personal activities over the day to optimize your benefit, as variable pricing encourage consumers to do, is a form of time management that comes naturally to white-collar professionals who plan their own schedules. But for other workers, these ways of thinking are not necessarily natural, or even desirable.

   For example, in my research into how rural Canadians organize their work, rural fisherman-farmers taught me that planning your time in advance doesn’t make sense when your work options are continuously altered by unpredictable weather conditions. If you wait to use your chainsaw until rates are down, it may be dark outside, or it may be raining sideways. Thriving in these conditions relies not on futile attempts to predict what will happen, but on savvy responses in the moment to conditions fundamentally outside your control. Fishermen-farmers are certainly able to plan out their activities in time when they have to, but it’s not their bread-and-butter. As a result, they are not particularly good at it; nor should they have to be.

   The same is true for sleep-deprived new parents, people struggling with health issues or addiction, and anyone else who finds rational control of their time beyond their current emotional, financial, or organizational means. The fundamental problem with variable pricing, as energy technology analyst Yolande Strengers points out [5], is that it assumes that electrical consumers should all be able to develop and execute a rational reorganization of their activities to optimize their energy use. And under the proposed plan, the cost for failing to live up to that moral demand is a bill up to 4 times higher.

   What should we do instead? If the goal is to even out energy demand without decimating the households who are least able to consciously manage it, there are several alternatives.

   As a start, if variable pricing does enter US homes, there should be income cut-offs under which participation would be opt-in only. But a better option would be to change the target audience for it. While variable pricing does not fit how many consumers organize their lives, it does fit with business’s conscious attention to strategy and pricing. It makes sense to concentrate variable pricing on the business context, which uses nearly twice as much energy as the residential sector [6].

   But this policy deserves more fundamental rethinking. As environmental researcher Max Liboiron argues, a more effective design strategy than trying to get consumers to consciously redesign their behavior is to change the factors that influence what people do unthinkingly [7]. Electrical policy should shift from forcing people to consciously restructure their behavior, to making non-peak electrical usage an automatic solution. For example, utilities could work with device manufacturers to make the default ‘on’ time for large appliances such as dishwashers and washing machines be when prices are low. If you need them to turn on right away, you would push an extra button. The expensive option would be the decision that requires extra thought. This form of restructuration would have the benefits of variable pricing without penalizing consumers who don’t manage their time like a white-collar professional.

   1. http://www.ithacajournal.com/story/news/local/watchdog/2017/08/10/adjustable-pricing-tripled-rates-next-step-electric-deregulation/528720001/
   2. http://www.sciencedaily.com/releases/2013/07/130702100757.htm
   3. http://dx.doi.org/10.1145/1620545.1620583
   4. https://www.nytimes.com/interactive/2014/08/13/us/starbucks-workers-scheduling-hours.html
   5. https://doi.org/10.1145/2621931
   6. https://www.eia.gov/consumption/
   7. https://discardstudies.com/2014/01/23/against-awareness-for-scale-garbage-is-infrastructure-not-behavior/