Tony's Thesis Deed(s) of the Day

[Tony H S Chu]

Keeping my thoughts in one thread to a) keep them in the same place and b) make it easy for you to ignore if it's distracting.

Deed(s):

Watched this TED video of Clay Shirky: "How the Internet Will (One Day) Transform Government" which talks about Git as a new paradigm for argumentation, which we should press into the service of legislation.

Discussed with Barbara about mapping the typographic hierarchy of a bill. She suggested I go to NYPL and look at other complex pieces of printed text and observe how they are type set.

Thought(s):

"What is the minimal meaningful set of shared annotations which can be machine aggregated?" 

"How to preserve clause completeness and nest hierarchy in plain-text format without using tags" => Pythonic use of white space, Markdown-esque syntax and a corresponding interpreter.

[Tony H S Chu]

... what should have been Wednesday's log.

Deeds:

Met with Chrys Wu, a former journalist at WaPo, NYTimes and LATimes and community strategist. Learned a bunch about a journalist's process, and some of the nuance of political reporting. Also got a list of names to reach out to.

Thoughts: 

"How do I want people to feel when they use Project L? What emotional/visceral need does it satisfy?"

"Check out Summly, a Betaworks investment that automatically summarizes long text."

[Tony H S Chu]

... and Thursday and Friday compressed into one day due to the Feast..

Deeds:

Talked to a number of people at the Feast about Project L. Receptions were generally, "That would be really cool if you could do it. But it sounds really hard." i.e. most people were a bit skeptical.

Wrote thesis blogpost on how I'm thinking about thesis lately: http://blog.tonyhschu.ca/post/32918375933/ricocheting-thoughts-of-thesis.

[Tony H S Chu]

Today, which is Saturday.

Deeds

Started to read bits of The Elements of Typographic Style, as per Barbara's recommendation.

Read and attempted to map structure of a House Resolution (H.R. 4010)

Researched people I should talk to, in several categories:

• journalist and journalism professors
  
• law and politics professors
  
• law and politics bloggers
  
• civic hackers and Sunlight foundation
  
• Congressional Offices
  

... and just a lot of basic secondary in the space.

Thoughts

"Shit, a lot of the ideas I want to do is being done already by various projects in the Sunlight foundation. What makes what I am proposing better? And it better be an order of magnitude better."

On Tuesday, October 2, 2012 11:55:42 PM UTC-4, Tony H S Chu wrote:

[Nicole Sylianteng]

I think I sent this to Barbara a while back but it might apply to you too: 

Elements of Typographic Style applied to the web

http://webtypography.net/

--
 
 

--

Nikki Sylianteng

http://nikkisylianteng.com

[Tony H S Chu]

Sunday..

Deeds:

Worked on Thesis Artifact. Typeset (if what I am doing can be called that) the first three sections of the DISCLOSE Act.

Thoughts:

"It's nice to be writing code again."

"Oh shit I have to start emailing some of these people tomorrow. Argh. Talking to strangers is hard."

On Tuesday, October 2, 2012 11:55:42 PM UTC-4, Tony H S Chu wrote:

[Tony H S Chu]

Monday

Deeds:

Chatted with Gary Chou about thesis. Gary thinks that it can be simpler, and closer to the user (i.e. less abstract.) Echoes the discussion I had at the Feast with Amit around finding the right level of abstraction.

Wrote reflection blogpost from today's crit. Will post tomorrow.

Thoughts:

"This is a social engagement and social engineering problem, not a UI and API problem."

"Don't substitute the problem you ought to solve with a problem you can solve."

"Gotta get in touch with Christina Cacioppo about the senate people she talked to at USV."

[Tony H S Chu]

Tuesday...

Deeds

Published blog post reflecting on yesterday's crit: http://blog.tonyhschu.ca/post/33232429968/t-45-gut-churn-in-public

Attended Barbara's type class in the morning, and gain perspective on type hierarchy. Maybe I should go to every class and just audit, it'll get me to school early.

Thoughts

Lots,

Giving serious thought about this matching people to bills idea, tossing a couple ideas around:

1. A OkCupid-esque open question submission, open answering system, where people are algorithmically matched to bills based on answers.

2. A WhatTheBook style, 7-12 incisive questions that matched users to a dozen hand curated bills.

3. A pulse.me style category and subcategory selection.

Also thinking lots about models of keeping people informed. Perhaps a personal dashboard of bills that match a user's preference, with a weekly email summary? Perhaps I'm jumping to solutions.

Speaking of which, also giving serious thoughts to research. Gary Chou cautioned that users cannot tell me what to build (henry ford, faster horse, etc.) nonetheless I think talking to people will help me understand the needs better. Thinking about three types of stakeholders:

1. Political Insiders (Congressional aides, think tanks, policy makers)

2. Storytellers (Journalists, activists, bloggers)

3. Civilians (people who feel like they should pay attention, but haven't found a way in.)

Tomorrow will be cranking out emails day, trying to reach as many of the first two group as possible. Need to devise strategies for engaging "civilians" as well.

[Tony H S Chu]

Wednesday...

Deeds:

• Wrote blog post on law as abstraction and its parallels to programming: http://blog.tonyhschu.ca/post/33280812626/levels-of-abstraction
  
• Giant Brainstorming session re-examining the "Help People Engage with Congress" space

Thoughts:

• "What if I could better illustrate how law is like programming?"
  
• "The need is to bridge the abstract of bill text, to the visceral concerns of citizens, through multiple layers of abstraction."
  
• "The need is to help stakeholder { political insiders, advocates, citizens } adopt a common, shared language."
  
• "What does success look like? What is a measurement of success?"
  
• "This is getting way too big. What's a small piece I can bite off and work on?"
  

[Tony H S Chu]

Thursday ...

Deeds:

• Spoke with Amit to brainstorm a number of ideas.
  
• Wrote a blogpost about what metrics I should measure success by for my thesis.
  
• Spoke with Cooper, and got these gems: "Fantasy Congress ala Fantasy Football", "RapGenius for Bills" (See: Apple's Terms of Service on RapGenius)
• Looked up Votizen's API. Might be an interesting tool to integrate.

Thoughts:

• Q: How did Nate Silvers get people to care about polls? A: People already cared about polls, he just made them make more sense. That's what I want with policy.
  
• via Amit, "Interface/platform as campfire for a community to gather around."
  
• Survive Congress: Only a small number of bills becomes law. Reminded me of this grim XKCD comic: http://xkcd.com/931/ - can I generate/visualize where bills usually die in this process in a similar way?
  
• On the statistics note - what if there's a prediction market for bills going through Congress? Likelihood of bills passing and failing generated with weight predictions instead of simply votes.
  
• Am I writing too many blog posts?
  

[Tony H S Chu]

Friday...

Deeds: None.

Thoughts: How might I find/gather/crawl data to figure out where bills are most likely to fail?

[Tony H S Chu]

Saturday...

Deeds:

Did some serious research deep dive. (Thanks in part to suggestions from random twitter acquitance: http://notes.pinboard.in/u:tofias/722fbd1a36977dc4e782)

• Read thru OpenCongress' "How Bills Become Law." 
  
• Read though some social science papers on what makes bills pass or fail.
  
• Read a theory about the Logic of Congress: http://wikisum.com/w/Arnold:_The_logic_of_Congressional_action
  

Drew a bunch of potential data viz sketches.

Thoughts:

• "Maybe I can use Bayesian analysis to identity hot spots in the legislative process."
  
• "Most bills die in committee through inaction. Wonder if there's more granular information about why."
  
• "Not all bills are the same. Over 40% of bills are considered 'trivial'"
  
• "Food labeling as a model for informing citizens about bill content and implications." Tangent: Food label redesigns: http://well.blogs.nytimes.com/2011/07/28/designing-a-better-food-label/
  
• "Explaining committee reports might be more important than parsing bill text. Bill text is to source code as committee report is to code comments."
  

On Tuesday, October 2, 2012 11:55:42 PM UTC-4, Tony H S Chu wrote:

[Tony H S Chu]

Sunday... sick, but...

Deeds:

• Brief blogpost on recalibrating thesis core, apathy, and death-rays: http://blog.tonyhschu.ca/post/33616238242/t-44-fighting-apathy-with-death-rays
  
• Listened to the first three episodes of DecodeDC: http://www.decodedc.com/
  

Thoughts:

• What if we frame disrupting politics as lower the barrier-to-entry to the lobbying industry? Can you be your own lobbyist? Kickstart your own lobby?
  
• How can visualization help contextualize numbers in politics? e.g. PBS’s subsidy being 0.0014% of Federal Budget?
  
• Can we analyse legislative process for bellweathers & ROI akin to @fivethirtyeight's work for elections & polls?
  

[Tony H S Chu]

Monday ...

Deeds:

• Spoke with Sarah, and sketched a bunch of diagrams to illustrate thesis ideas. (Need to actually put those diagrams together properly)
  
• Spoke with Nick Grossman, advocate-in-residence at Union Square Ventures. He's excited about my ideas (almost too excited) and gave me a lot of references to look at.
  

Thoughts:

• My audience is the interested, inclined, but disempowered
  
• Political apathy is learned helplessness
  
• Wouldn't it be cool if there's a way to bring up every relevant Jon Stewart video clip for any given policy issue?
• How can I make politics less serious?
• I am not sure I am comfortable with the idea of building a political attention death-ray.

[Tony H S Chu]

Tuesday ...

Deeds:

• First blog post trying to imagine different constraints on thesis to see if something more concrete shakes out: http://blog.tonyhschu.ca/post/33710679409/playing-with-time-limits ... it makes me realize I do want to make something about clarity and education, and not necessarily about empowerment, which is odd.
  
• Second blog post is about rhythm and iteration: http://blog.tonyhschu.ca/post/33753890765/always-be-iterating-maybe

Thoughts:

• I don't want "sexy" - I want something useful and illuminating. I want people to look at what I made and feel smarter.
  
• ... that said, I have no idea what I am doing.
• What does it mean to push the envelop of interaction design?
• What does it mean to build a following for an idea?

[Tony H S Chu]

Wednesday ...

Deeds:

none related to thesis =(

Thoughts:

• Comparing metaphors: Wikipedia for Politics vs. Buzzfeed for Politics
  
• What attracts good writers and thinkers? They are key to a healthy community
  
• How does communities that can respectfully disagree form? What kind of culture is required?
  
• Something something empathize with moral intuitions instead of adversarial reasoning. (via Jonathan Haidt, see:
  
• "The moral roots of liberals and conservatives" - http://www.ted.com/talks/jonathan_haidt_on_the_moral_mind.html
  
• "Why won't they listen?" - http://www.nytimes.com/2012/03/25/books/review/the-righteous-mind-by-jonathan-haidt.html?pagewanted=all&_r=0

[Tony H S Chu]

Thursday ...

Deeds:

• Spoke with Amit to consolidate ideas about prototypes for thesis. Small Gem.
  

Thoughts:

• Build something small, beautiful, intuitive, and modular.
  
• All my thesis ideas can be modulars that build upon each other. Which means I should start on the smallest.
  

[Tony Chu]

Friday ...

Deeds:

• Compiling sort of a mood board for visually engaging dashboards: Looking at theverge.com, kickstarter.com, chartbeat.com
• Messed with Google Alerts to generate feeds for keywords, like this search for "DISCLOSE act": http://www.google.com/alerts/feeds/07052118064226548677/8489315057844182288
• Listen to the This American Life episode on Congress and Lobbying: http://www.thisamericanlife.org/radio-archives/episode/461/take-the-money-and-run-for-office

Thoughts:

• "What are the components I need for a tiny dashboard?"
• "How can I bootstrap some of this data?"
• "Where can I find bayesian probabilities for stages a bill goes through?"
• "I better start prototyping tomorrow."

[Tony H S Chu]

Saturday ...

Deeds:

• Read a third of Republic, Lost. Felt infuriated, but that's why I picked the thesis I did.
  
• Watched the RSA video of Dan Ariely's lecture: "The Honest Truth about Dishonesty" http://www.youtube.com/watch?v=XBmJay_qdNc&feature=player_embedded

Thoughts: 

• "Crap, I didn't do my prototype."
  
• "What is the implication of Ariely's point about most dishonesty being the small, fudge a little kind."
  

 

[Tony H S Chu]

Sunday ... didn't think about thesis. Too busy with service design.

Monday ... didn't think about thesis, except in Amit's class. In which I felt calm about thesis. Which felt good.

[Tony H S Chu]

Tuesday ...

Deeds:

• Started to sketch what a nutritional label of a bill might look like, which led to;
  
• A skim through Tufte's "The Visual Display of Quantitive Information"
  
• Had a chat with Michael about the "embeddable label" distribution strategy
  
• Found "Manybills" by folks at an IBM research lab. http://manybills.researchlabs.ibm.com/
• The did machine learning classification of legislation section!!! Really cool CompSci work
• ... and really crappy UI design and social design front end...

Thoughts:

• "What is the absolute minimal people need to know about a bill to spark their curiosity?"
  
• Knee-jerk dimensions: Importance, Urgency, Timeline, Probability Prognosis, Topic Categorization
• via @erinmoore: "what would a political framework look like that allowed ppl to practice empathy instead of access info?"
• Re-read Clay Christensen's idea around "job-to-be-done" and asking, "What is the consumer *hiring* project L to do?"
• "I need to write a more thorough description of my strategy, now that I feel like I have one."

[Tony H S Chu]

Wow, completely fell off the wagon this week. I blame Service Design.

Wednesday ... don't think I did much around thesis because of the workshop.

Thursday ... same thing. No thesis because of workshop.

Friday ...

Deeds:

• Read through The Scope and Urgency of Legislation: Reconsidering Bill Success in the House of Representatives in full: http://congressionalbills.org/APSA%202005%20%203.pdf
• Wrote blog post clarifying the goal of the first prototype, and the thinking behind it: http://blog.tonyhschu.ca/post/34372571556/t-43-making-legislation-pointable
  

Thoughts:

• "How do I operationalize scope and urgency algorithmically?"
  
• "Oh look, TechCrunch is doing a politics thing: http://techcrunch.com/policy/"

Saturday ...

Deeds:

• Properly read through Tufte's "The Visual Display of Quantitive Information"
• Looked at examples of nutritional label design, dashboard design, and infographic design for research
  
• Drew ten different iterations of a potential widget graphic on post-it notes.

Thoughts:

• "Should I communicate to help understand? or communicate to perk interest?"
  

[Tony H S Chu]

Completely fell off the wagon again. This time because I fell off the internets due to Sandy.

Sunday ... was a write-off mostly running around town preparing for the hurricane.

Monday ...

Deeds

• Read a good chunk of "Analyzing Politics" by Kenneth Shepsle. Starting to get a better sense of how political scientists have systematically analysed group choice and democratic systems
• Did a first pass at the "nutritional label for law" ... before the power went out.

Thoughts I've forgotten. This is why I need to do this on a day-to-day basis.

Tuesday ...

Deeds

• Power was out, so I mostly read "Analyzing Politics" and "Republic, Lost" until it got dark

Thoughts

• "How do I explain to people about the distribution strategy of Project L?"
• "How do pace the introduction of complexity so people are not overwhelmed?"

Wednesday ... started to camp out at Tash's. Thanks Tash!

Deeds

• Did a second pass on the nutritional label, based on Tash's suggestions
• Did a preliminary explanation of the label ... see: http://thesis.tonyhschu.ca/bill-label-explanation.jpg

Thoughts

• A big function of Project L is to provide context for the legislative process. Maybe it should be called PoliContext? ... not sure about it.
• One way to frame the participation problem is to look at feedback loops.
• The length of the feedback loop in participating in the civic process is LONG
• The signal of the feedback loop in participating is NOISY
• ... what if we can give people a different kind of feedback loop to play with?
• Got feedback from Michael Tofias about the label
• Became very clear that the graphic doesn't provide enough context around
• What it is for in the context of Project L
• What some of the metrics mean
• Got some new ideas around what to communicate, key ideas:
• Visualizing coalitions and how support for a bill gets built
• Communicating (or surfacing from other sources) policy implication
• Which means:
• I have to read Shepsle in more depth
• I have to explain Project L better, and more quickly.

Thursday ...

Deeds

• Investigated voteview.com, a site where political scientists are visualizing alliances and coalition patterns in Congress using Bayesian analysis
•  realized that I don't understand what is going on at all. Considering writing these people.

Thoughts

• How do I explain what I am doing graphically? Writing 5 paragraphs would be easy, but not the most effective. (h/t Tash Wong)

[Tony Chu]

Friday ...

Deeds

• Continued reading "Analyzing Politics"
  

Thoughts

• Still mulling over the idea of a short feedback cycle "game" for keeping up with politics. Cooper's idea of a "Fantasy Football" analog still seems interesting.
  
• Discussed ideas for next steps with Tash and Tom. Current thinking is around building the set of feedback systems I envisioned around just one bill. (H.R. 4010, which I have been using consistently.) ... and then ruthlessly growth-hacking that prototype over the rest of the semester.

--
 
 

[Tony H S Chu]

Saturday ...

Deeds

• Read even more of "Analyzing Politics" ... not sure what to do with all this new knowledge
• Wrote two halves of two different blog posts... but now I am stuck on both:
• "Thoughts on Political Futility" - trying to think through how shorter or alternative feedback mechanisms might invigorate political participation. Couldn't complete the thought though.
• "Pacing Complexity" - trying to think through the idea of progressive introduction of complexity. How do I pace the introduction of complexity so that it never at any one point becomes overwhelming, yet on the whole communicates the complexity of how politics in Congress works.
• Started thinking about the next iteration on the nutritional label prototype.

Thoughts

• ... see about I guess?

[Tony Chu]

Monday

Deeds

• Spoke with Amit about how to move forward with the first prototype.
• ... re-affirmed ideas for how to move forward.
• ... had design and metrics displayed ripped apart (in a good way.)

Thoughts

• Thinking about building a small site that allows people to embed stuff. Do it and keep it really, really small. Do the iframe thing. The obvious things. Cut as much out of the scope as possible
  
• Really liking the idea of allowing people to respond to a call-to-action within the widget.

--
 
 

[Tony H S Chu]

Tuesday ...

Deeds

• Met up with Clint Beharry, an alumni of the program, and his colleagues at Harmony Institute. They are all political scientists with statistics backgrounds, which made for some interesting conversations. Some good feedback, but a lot of what Amit said would happen too, i.e. ideas that might knock you off your core
  
• Started writing a blog post, didn't finish. That's three unfinished blogposts about thesis now

Thoughts

• I think I've arrived at a reasonable scope of a minimal viable product for end of term now. It has to do three things
• Be embeddable - I need to build an embeddable widget that add value to what people are writing on distribution networks.
• Provide context - really get to the point of why people should care, and then show people how Congress works if they are interest. Thinking the widget would provide a subset of all the metrics I provide in the "main" site
• Lead to action - Whether within the widget or in the site, the widget needs to provide people a clear call to action to participate and influence the system.
• ... some extension of these ideas (which I will NOT do this term.)
• Provide analytics - the widget has the potential to broker traffic, which can provide data of interest to:
• Bloggers using the widget to accompany their writing
• Congressional staffers who want to see public response to their work
• Researchers who want to see blogposts specifically about a certain bill
• Be a news aggregation hub with commentary (a la Reddit)