DataDownload: Your Personal Data Is Super Valuable. Time to Cash In?
DataDownload: Your Personal Data Is Super Valuable. Time to Cash In? A weekly summary of all things Media, Data, Emerging Tech View this email in your browser
Dear Readers — Today, I have to tell you I’m sure of one thing. DATA is fast becoming super sexy. UC Berkley’s Dawn Song is building apps so that you can get paid by companies that want to use it. Data journalism is quickly becoming more than just a buzzword. AI is growing as a tool for journalism, with lots of juicy details being published next week at Hack/Hackers London. And our friend David Carroll argues that microtargeted political ads pervert the democratic process. His argument is political messaging is targeted on… data. Again, data privacy as a solution? Ok, let’s just say yes to that.
So — before Skynet becomes self-aware, it’s time to take control of the data, before it takes control of us. Unless it’s already happened.
Thoughtful feedback always welcome: Steve@NYCMediaLab.org.
NYC Media Lab
Building a World Where Data Privacy Exists Online
Many experts have proposed ways we could own and even get paid for all the personal data we kindly donate to tech giants (Jaron Lanier’s plan is particularly awesome… also see the NYT opinion piece on a public option for AI). But few are actually building something to support the concept.
UC Berkley professor Dawn Song (Twitter), one of the first researchers to demonstrate real-world adversarial examples that could fool AI systems, recently founded Oasis Labs, a platform that represents one of the first steps towards a future where users can generate revenue streams from their data.
Oasis gives developers the ability to build applications that grant users control of their data and the ability to audit how that data is used. Song has already initiated a pilot with Nebula Genomics, a direct-to-consumer gene-sequencing company that allows customers to retain ownership over their genomic data.
6 min read
In Data Journalism, Tech Matters Less Than the People
Ben Casselman’s LinkedIn description notes that the NYT journalist has a “specialty in data-driven storytelling” (a statement cemented by the fact that he was previously chief economics writer at FiveThirtyEight).
With a monitor littered with “charts, spreadsheets and way more Chrome tabs than any sane person would consider reasonable,” Casselman has a firm grasp of what it means to be a data journalist… and what it doesn’t.
In this interview, Casselman explains why data journalism isn’t just “staring at spreadsheets until a story magically appears,” the software he uses on a daily basis, and the minimum technical skills reporters today need.
6 min read Read More
For the Media
Why It’s About Time We Talk Seriously About AI and Journalism
This Monday Journalism AI will publish their report at Hacks/Hackers London. The project — a collaboration between Polis and Google — involved seven months of talks with 71 news organizations on the role of AI in journalism.
Ahead of the report, social media editor at The Telegraph Francesco Zaffarano spoke with Polis founding director Charlie Beckett, who tackled some tough questions such as what the current state of AI in the industry is, how to prevent the gap between small and large media companies from widening, how to deal with cultural resistance, and what AI’s role in misinformation is.
4 min read Read More
Wikipedia Co-Founder Wants to Give You an Alternative to Facebook and Twitter
Wikipedia co-founder Jimmy Wales launched Wikitribune to help fight fake news with original stories and citizen journalism. The publication floundered, so Wales pivoted the project to a “news focused social network” called WT:Social. Now for the vital differences: WT:Social will be funded by donations instead of ad dollars, will show the newest links first (no algorithmic curation, with a possible upvote button added in the future), and will support small niche communities.
How’s the social network faring so far? In the past month, 50k have signed up, but Wales is aiming for 500M.
2 min read Read More
There Could Be a Simple Solution to Facebook’s Political Ad Woes
Data-based renegade and The New School associate professor of media design David Carroll (for the uninitiated, read this) thinks Zuckerberg’s argument that political ads are expressions of free speech is fundamentally flawed (meanwhile, Twitter recently banned political ads).
The argument goes like this: Zuckerberg said users should debate among themselves to expose misinformation in political ads, while Carroll notes that “false ads are being targeted to only select voters that they can predict won’t challenge them.” He suggests Facebook continue allowing political ads, but eliminate the microtargeting that’s killing “informed conversation [between] neighbors.”
4 min read Read More
What We’re Watching
This AI Clones Your Voice After Listening for 5 Seconds
If you haven’t heard of Two Minute Papers, it’s a succinct YouTube channel that summarizes the latest in AI research for the less technical crowd.
In a recent video, the host summarized a 2018 paper that described a text-to-speech system that only needs seconds of input to recreate your voice. Here’s an unofficial implementation of the system, and you can play around with the audio sample page from the video here.
5 min watch Watch Now
Events & Announcements
Event: 2019 Nonprofit TechCon
Date: December 5, 9AM-4:30PM
Hosted by NYN Media, this is the go-to conference for information on technology tools, services, and innovation at New York’s nonprofits. Attendees will explore the best methods and biggest challenges to keeping up in a quickly changing world. This event will cater to everything from entry-level learning about the current state of technology to more advanced discussions. Topics to include tech advances, using data effectively, cybersecurity, integrating tech into organizations, and more. Use discount code NYCMEDIA for 50% off tickets. Register Here.
Event: Future Forum
Date: November 17–18
An all day event for people who work with AI and those who will be impacted by it. The event will convene tech industry gatekeepers, researchers, early adopters, regulators, journalists, activists, and skeptics who will work together to demystify AI. RSVP Here.
Event: Walking Tour of NYC Internet Infrastructure
Date: November 23, 2PM-5PM
What does the Internet look like? In her book Networks of New York, Ingrid Burrington lifts our eyes from our screens to the streets, showing us that the Internet is everywhere around us, all the time — we just have to know where to look. She will be bringing the book to life in a walking tour that begins in NYC’s Lower East Side. RSVP Here. A Deeper Look A Deep Dive Into BERT: How BERT Launched a Rocket Into Natural Language Understanding
Search Engine Land presents the clearest (and one of the lengthiest) elucidations of BERT we’ve seen. Why is an SEO publication diving into an NLP framework? Simply, “Google describes BERT as the largest change to its search system since the company introduced RankBrain, almost five years ago, and probably one of the largest changes in search ever.”
Besides succinctly describing the framework, SEL explores why BERT was such an important step forward, summarizing NLP’s issues with lexical ambiguity (ex. polysemy, homonymy, homographs, homophones, coreferences, etc.), the shortcomings of language learning methods like Word2Vec and Google Pygmalion, and how BERT’s transformer architecture changed the game.
49 min read Read More
A Visual Timeline of AI Predictions in Sci-Fi
Ever read an older sci-fi book and stumbled on an uncannily accurate description of a modern technology (or, remember that scene where 1991 Arnold says his brain runs on a neural net processor)? Even The Jetsons (and over a hundred years of concept art) got video calls right. Tech fantasies are coming into existence at an increasingly rapid pace — Noodle.ai’s infographic shows us the concepts that became reality from films and TV shows.
4 min read Read More