DataDownload: The true colors of the political spectrum

8 min readSep 12, 2020

DataDownload: The true colors of the political spectrum A weekly summary of all things Media, Data, Emerging Tech View this email in your browser

Hi everyone.

Welcome back. It’s fall — which feels strangely a lot like summer, but a bit chillier.

We’re gearing up for the NYC Media Lab Summit. If you haven’t grabbed a ticket, there’s a free ticket waiting for you HERE (or for $25 we’ll ship you a modest swag box and our appreciation.)

Super excited that Tristan Harris will be kicking us off with a powerful Keynote on October 7th. You won’t want to miss it, or his brand new film…The Social Dilemma.

Plenty more to read this week. So, we won’t keep you.

Please reach out with ideas, suggestions, or feedback always welcome. Steve@nycmedialab.org.

Steve

Steven Rosenbaum
Managing Director
The NYC Media Lab
Steve@NYCMediaLab.org Must-Read The True Colors of America’s Political Spectrum Are Gray and Green

Can palettes of grassland-green or business-tower-grey reveal a political affiliation? NY Times found that — to a degree — it can. “Think of it as the aerial-image version of those.” And while a single pane won’t reveal much, in the collage above a noticeable pattern starts to surface: open spaces lean politically right, city streets lean left. This is still an insignificant sample: “less than one-tenth of 1 percent of the land cover of the country.” So the paper used powerful resources to process every square meter in precincts where there were votes cast. Here, the pattern emerges more prominently.

These are fascinating to look at. So, one more — Landscapes Across the Political Spectrum:

6 min read

Two-year-old TikTok’s stats are staggering: 100M MAUs and 50M DAUs in the US (up 800% from January 2018), downloaded 2B times… and it’s challenging the US government in a lawsuit. As VP Michael Beckerman noted to reporters this week, “we’re a 2-year-old company operating with the expectations of a 10-year-old company…. We grew up in the techlash age, where there’s a lot of skepticism of platforms, how they moderate content and how their algorithms work.”

This skepticism has prompted ByteDance to design a very literal transparency center in LA — “the center will have areas for people to demo computer modules that showcase how TikTok’s algorithms and data practices work.” Reporters — including Axios — got a virtual tour of the center, learning more about how TikTok’s algorithm works:

When users open TikTok for the first time, they are shown 8 popular videos featuring different trends, music, and topics. After that, the algorithm will continue to serve the user new iterations of 8 videos based on which videos the user engages with and what the user does.
The algorithm identifies similar videos to those that have engaged a user based on video information, which could include details like captions, hashtags or sounds. Recommendations also take into account user device and account settings, which include data like language preference, country setting, and device type.
Once TikTok collects enough data about the user, the app is able to map a user’s preferences in relation to similar users and group them into “clusters.” Simultaneously, it also groups videos into “clusters” based on similar themes, like “basketball” or “bunnies.”
Using machine learning, the algorithm serves videos to users based on their proximity to other clusters of users and content that they like.
TikTok’s logic aims to avoid redundancies that could bore the user, like seeing multiple videos with the same music or from the same creator.

4 min read

It’s no secret that viral media — no matter context or veracity — is self-perpetuation and self-fulfilling. The latest content to stir up intense emotions is riot porn — snippets of rioting that have prompted groups to self-mobilize (via social media) and engage in violence against protestors. Strong emotions ignited by social media can create veritable change, like the BLM and Occupy protests, and at the same time incite violence against these changes. Joan Donovan, research director at Harvard Kennedy’s Shorenstein Center, noticed a reactionary split between how left- and right-wing citizens dealt with these emotions:

“Because right-wing reactionaries do not have the same kind of experience in organizing street protests as the left, they instead rely heavily on social media — particularly Facebook, YouTube, and Twitter — to mobilize crowds…. Given the way media accounts shape public perceptions about protest and define who has recourse to the “legitimate use of violence,” the kinds of content shared within these hyperpartisan media systems play a powerful yet often invisible role in mobilizing white vigilante groups. If social-media companies do not act swiftly to stop calls for violence against protesters, the situation can only get worse.”

9 min read Read More Would You Spend $10,000 on a Virtual Dress? Gucci Is Betting on It

A quick catch-up on virtual clothing milestones: Scandinavian retailer Carlings successfully marketed digital clothing — a total change of pace for the company; The Fabricant sold a $9,500 dress… on the blockchain; and lest we forget, Fortnite rakes in billions — partly from skins. More recently, Ralph Lauren teamed up Snapchat to advertise digital fashion accessories for bitmoji, and some people have spent thousands on Aglet, the “Pokemon GO for sneakerheads.”

It’s not so surprising that Gucci wants in on the $100B virtual goods market. Gucci has already made digital clothes for Drest (the “first interactive luxury styling game”) and athletic wear for Tennis Clash. Gucci CMO Robert Triefus says that the “worlds of fashion and gaming are colliding. We’re approaching gaming with a sense of experimentation, because this will put us in a good position to be ahead of the trends when they become ingrained.”

6 min read Read More This Know-It-All AI Learns by Reading the Entire Web Nonstop

Google’s knowledge graph, which you can see in action by typing in known companies and celebrities, is essentially a network of billions of interconnected facts (read the company’s definition here). Google does this for more popular terms, but AI company Diffbot wants to do this for everything.

The company’s system has been ingesting 100M -150M entities (people, places, companies, things, etc.) per month, and rebuilds its knowledge graph every four or five days. It uses that data to relate one thing to another via subject, verb, and object.

This type of network can do some pretty impressive things, such as being able to tell Nike if someone is selling fake shoes (vs. just talking about them on a forum). Ultimately, the company wants to build a “universal factoid question answering system, an AI that could answer almost anything you asked it, with sources to back up its response.”

6 min read

Read More What We’re Watching The Social Dilemma

Tristan Harris was an employee at Google when he began to get a disturbing feeling that technology was pulling apart the threads of society. He shifted his concern into action. Today he’s described as “the closest thing Silicon Valley has to a conscience.” He now runs the Center for Humane Technology — and he’s made a movie that aims to scare Americans into action. It’s called The Social Dilemma, a film that launched on Netflix this week. Tristan will keynote the NYC Media Lab Summit on October 7th. He’s got some important insights to share. Watch it on Netflix here.

94 min watch

Watch Now What We’re Listening To Podcast: ‎TESLA: The Inventor in Old New York

“The Serbian immigrant Nikola Tesla was among the Gilded Age’s brightest minds, a visionary thinker and inventor who gave the world innovations in electricity, radio and wireless communication. So why has Tesla garnered the mantle of cult status among many?” Find out on this podcast!

52 min listen

Listen Now Virtual Events Virtual Event: The NYC Media Lab SUMMIT 2020. Building the Future Together
Date: October 7–9
We’ll bring together 1,000+ virtual attendees from NYC Media Lab’s core community — including media and tech executives, university faculty, students, investors, and entrepreneurs — to explore the future of media and tech in New York City and beyond. Register Here.

Virtual Event: Shift Dev 2020
Date: September 14–15
A two-day open air and hybrid Developer conference specifically created to continue the tradition of delivering amazing educational and inspirational talks that the developer community has come accustomed to but in a format fit for the world we live in today. Register Here.

Virtual Event: Terms of Usage, a Data Science Day 2020 Virtual Event
Date: September 14
As part of the Data Science Institute’s flagship annual event, this virtual session will bring together thought leaders who are driving discussions on ethics and privacy in data science and engineering. Register Here. A Deeper Look RoboRXN: Automating Chemical Synthesis

IBM has built RoboRXN — a drug-making lab housed completely in the cloud, with the aim of helping WFH scientists design and create new molecules at an accelerated pace. The process is a human-machine loop where the scientist sketches the skeletal structure of the compounds they want to make in a web browser interface, then sends off the blueprint to RoboRXN, which uses machine learning to predict the compound’s ingredients and how they should be mixed. Finally, the structured recipe is sent to a remote lab robot, which conducts the experiment.