Cody Romano codyromano

Untitled

Untitled is a social experiment that challenges people of different faiths & political backgrounds to explore — and in some cases embrace — each other's beliefs. It will launch as a web-based game in mid-2021.

Motivation

The game aims to tackle two common problems in debates about politics & religion:

Labeling. Labels like "conservative", "liberal", "Christian" or "Muslim" oversimplify people's beliefs. Though labeling is a natural human tendency, participants in a debate must be willing and able to unpack labels to understand what each other actually thinks.
Unconscious bias. Implicit or explicit biases kick in when someone sees a stranger's basic profile: picture, name, job title, hometown. Identity characteristics warp the lens through which we perceive strangers' beliefs.

Living persona project (working title)

Problem

For-profit companies define user archetypes called “personas” before developing a product. A typical persona includes a face, a name and a problem description. “I’m Jamal, a premed student, and I need books for next semester. Can I find them on Amazon.com?”

In combination with quantitative research, personas offer a window into customers’ thoughts and feelings. It’s frighteningly easy to build something useless when you haven’t fleshed out the underlying problem.

Most nonprofits don’t have evidence-based personas for the communities they serve because the research is expensive and time consuming. This creates a chicken-and-egg problem: it costs money to gather stories about people who need your help, yet the narratives are essential to grant writing.

	// Let's say we have coordinates in a game. Managing the location of objects can be tricky,
	// and we want to validate that nothing flows off the screen

	export type PositionDataType = {
	x: number;
	y: number;
	};

	// Here's a utility for creating a position data type with runtime validation
	export const positionDataType = createDataType<PositionDataType>('position')

	/*
	Using Google Chrome on your desktop (not mobile):

	1. Copy all of this
	2. Go to a different webpage (any webpage)
	3. In the top bar, click "View" > "Developer" > "JavaScript Console"

	A prompt will appear asking you to choose filler words and how long you want to speak.
	*/
	(function countHowManyTimesIUsedFillerWordsDuringAMeeting(fillerWords, secondsToRecord) {

	import jwksClient from 'jwks-rsa';
	import jsonWebToken from 'jsonwebtoken';

	const jwksUri = 'https://appleid.apple.com/auth/keys';
	const client = jwksClient({
	jwksUri
	});

	function getKey(header, callback) {
	client.getSigningKey(header.kid, function(err, key) {


	const GET_CONVERSATION = gql`
	query GetConversation($request: GetConversationInput!) {
	getConversation(request: $request) {
	conversation {
	id
	}
	}
	}
	`;

	// Options available when creating the HOC
	export type WithFocusQueryOptions<ComponentProps, QueryVars> = {
	query: DocumentNode;
	getQueryVars?: (props: ComponentProps) => QueryVars;
	};

	// Props provided to the wrapped component
	export type WithFocusQueryProps<QueryResponse, QueryVars> = {
	data: QueryResponse;
	refetch: (options?: QueryVars) => void;

	const findWordsInMatrix = (matrix, dictionary) => {
	// Convert dictionary to map for O(1) lookup
	// We could also use a trie here (more efficient, but verbose implementation)
	const dictionary = dictionary.reduce((set, word) => {
	set.add(word);
	return set;
	}, new Set());

	const wordsFound = new Set();
	const visited = new Map();

	/**
	* Let's say we want to find the shortest path among all pairs
	* of U.S. cities. For example, let's assume (for the sake of argument)
	* that Seattle is 200 miles from Chicago, 300 miles from Miami, etc.
	*/
	const cities = [
	{
	name: 'Seattle',
	edges: [
	['Chicago', 200],

	/*
	Initial questions:
	- How large is the grid?
	- What is the format of the input I'm given?
	- Is the player guaranteed to exist on the grid?
	- Is there always going to be a blocking tile?

	Proposal: dynamic programming

	- If adjacent tile is out of bounds (e.g. we're at