DataDownload: Our future — and global power dynamics — may depend on AI ethics

8 min readNov 23, 2019

DataDownload: Our future — and global power dynamics — may depend on AI ethics A weekly summary of all things Media, Data, Emerging Tech View this email in your browser

So — are you feeling all warm and fuzzy about the future of News and status of Data Privacy? If so, I’d like you to share your magic reality distortion field with me.

In my reading, the Gannett story is bad for Local News. The Polis / Google report is equally grim. And AI innovations are happening in a mysterious hidden land, with even logging out unable to protect you.

But, Medium’s decision to pay based on attention rather than “claps” is a good thing. And the NY Times letting folks inside their R&D world is also a good thing. Disclosure, we’re working with the Times on a Spacial Computing challenge right now, so expect cool things in the months ahead.

So — dig in, chow down, and remember that Turkey has tryptophans, so you’re excused if you’re a bit sleepy next week.

Ping me if you have thoughts or feedback about the Newsletter. Steve@NYCMediaLab.org.

Steven Rosenbaum
Managing Director
The NYC Media Lab Must-Read The Most Important Supreme Court Decision for Data Science and Machine Learning

With more companies incorporating AI into their strategy and algorithmic ethics top-of-mind (thanks to a string of controversies these last three years), it’s worth revisiting the legality of data usage. The impression one would take away from scouring media pieces on AI is that the tech is moving way too fast for regulators to keep up. And it is — but the algorithmic law field isn’t without precedent.

In this Towards Data Science feature, Harvard environmental and data science PhD candidate Matthew Stewart reviews the ramifications of the Author’s Guild v. Google case. The gist is in 2005 the Guild sued Google for copyright infringement because they used copyrighted books for training a book search algorithm. The case wrapped up around a decade later in Google’s favor, setting a precedent for the use of discriminative and generative ML algorithms, something that will likely be revisited in the age of AI.

9 min read

Read More Why Is Google Slow-Walking Its Breakthroughs in AI? Google is an AI powerhouse, but in recent years — contrary to its MO — it’s been keeping Pandora’s box away from its users for fear of abuse. (Last year, the company went as far as to say it won’t offer facial recognition services.) The powerful AI software it does release is sometimes leashed with restrictions, as with its recent Celebrity Recognition service.

Google offers the service to media companies who genuinely work with celebrity content, but any celebrity can opt out via web form, and the list is limited to thousands of famous people. Meanwhile, “Amazon and Microsoft have said their own services recognize hundreds of thousands of public figures.”

6 min read Read More For the Media New Powers, New Responsibilities. A Global Survey of Journalism and Artificial Intelligence

In last week’s newsletter, we noted that Journalism AI — a collaboration between Polis and Google — would release their report after seven months of talks with 71 news organizations on the role of AI in journalism. Well, the report is out — you can read all 111 pages of it here and you can check out a great wrap-up over at Nieman Lab. The primary lesson learned, according to project leader Charlie Beckett, is actually an urgent call to attention:

“Perhaps the biggest message we should take from this report is that we are at another critical historical moment. If we value journalism as a social good, provided by humans for humans, then we have a window of perhaps 2–5 years, when news organisations must get across this technology.”

111 pages

Read More The Future for Local News In Danger The “new Gannet” — as New Media Investment Group (NMIG) CEO Michael Reed called the merger of Gannet and NMIG’s GateHouse Media — is meant to create a “more viable” print and digital news empire. But the rosy sentiment is a mask for the industry’s difficult reality: analyst Ken Doctor notes that there hasn’t been a single year of revenue gain in the newspaper industry as a whole since 2008; GateHouse laid off over 60 employees this year; and Gannet several hundred.

Doctor somberly asks, “after two more years of this, what kind of products are going to be left in local communities that readers feel are worth a digital subscription? What’s your strategy after all the cost-cutting?” Besides the introduction of AI into newsrooms (see above), and experimentation with new ways of storytelling (see the What We’re Watching section), Ev Williams might serve as inspiration for papers struggling with this question: the Medium founder recently changed the platform’s payment model, basing it on reading time instead of claps (Medium’s version of “likes”) — read about it in this Trading News For Noise piece.

8 min read Read More You Can Log Out, but You Can’t Hide

Anti-tracking tool Ghostery released a new study that “shows that an overwhelming majority (79%) of websites globally are tracking visitors’ data — with 10% of these sites actually sending user data to 10 companies or more.” Axios rounds up key insights:

Google and Facebook collect more data than most other companies combined.
Too many trackers create slower browsing experiences. Some companies (WP, Bloomberg) are trying to reduce tracking partners to speed up their sites.
Almost one-third of tracked sites have a hidden Facebook tracker. (“Facebook won a critical privacy lawsuit in July over tracking users’ internet activity through ‘like’ button trackers even after they logged out.”)

1 min read

Read More What We’re Watching Documenting the World in 3D

The NY Times has a brand new website from its R&D team. And first up is “If-Then” — a new video series made by their R&D team. It’s an internal look at the process of using new technology to tell stories — and the challenges of going down uncharted paths.

In the first episode, If-Then explores how 5G can augment the reporting process, particularly how increased bandwidth can help capture objects of all sizes with photogrammetry — from the collection at the Museum of Trash to large-scale outdoor locations like Bannerman Castle on the Hudson River.

7 min watch Watch Now Events & Announcements Event: Citizens and Technology Summit 2019
Date: November 25, 6PM-8PM
The Citizens and Technology Summit brings together communities, researchers, and advocates to share, imagine, and design citizen-led agendas for technology and society. Register Here.

Event: Cloud Native Application Development with Kubernetes and OpenShift
Date: November 25
In this workshop, you’ll get an overview of Kubernetes, and move directly into leveraging OpenShift to what it provides for application development. You’ll then go through the process of building and deploying a microservice application in this cloud native way. Register Here.

Event: On AI ROI — The Questions You Need to Be Asking
Date: December 5, 12PM-12:45PM
Leaders across every industry are increasing investment in advanced analytics, data science, and AI, yet, many struggle to recognize a return. This talk will give leaders the tools to: identify and assess the possible impact of potential data science projects; support projects with high probability of success; and identify sunk costs and appropriate stopping points. Register Here. A Deeper Look AI Ethics Is All About Power

We’ve reached the age where popes warn against the dangers of AI. At the Common Good in the Digital Age conference, Pope Francis urged “Facebook executives, venture capitalists, and government regulators” to watch out for the impact of AI and other emerging tech. The warning klaxon has been sounding for years, and with good reason: “the chasm between AI haves and have-nots [is] so deep, that the global economic balance as we know it could be rocked by a series of catastrophic tectonic shifts,” reads a quote from a proposal by the Stanford Institute for Human-Centered AI.

The push for ethical AI, stresses senior VentureBeat AI author Khari Johnson, is not to have a “politically correct, corporate social responsibility” thing that’s nice to have (but that gets in the way of progress). Ethics are a way to recognize and break down the power dynamics at play that widen the said chasm. Johnson brilliantly breaks down the role of AI ethics in individual autonomy, government, society, and our future.

26 min read

Read More Intel Throws Down AI Gauntlet With Neural Network Chips Intel’s first generation of Neural Network Processors (NNP) are out. The two products are specialized for training (NNP-T) and inference (NNP-I), with initial customers including Facebook and Baidu. Naveen Rao, corporate VP and GM of the Artificial Intelligence Products Group at Intel, expects the company to generate $3.5B in revenue from their AI solutions in 2019 (“presumably [the Group] includes everything that has AI infused in the silicon”).

During the Intel AI Summit, Rao placed the need for specialized AI processors into perspective with a graph of AI complexity over time. Based on the number of model parameters, complexity has been 10X-ing every year: “to deal with that kind of growth, the users will not only have to rely on specialized processors that can perform the relevant calculations very quickly, but the ability to use them in scaled-out fashion.”