Machine Eatable

Illustration by Jonny Goldstein

Machine Eatable is a monthly lunch series I curate at Civic Hall, a candid discussion led by community leaders on the front lines of data science for civic good. I started the series with Jeanne Brooks and DataKind, and the Microsoft Tech & Civic Engagement team. The format is anti-panel: we get a few smart people up front, and 45-60 other smart people in the audience, and they get to talk to one another.

Previous #MachineEatable talks:

Illustrating Data Literacy
June 2017

Eyebeam Research Resident Mimi Onuoha (@thistimeitsmimi) and Data & Society Fellow Zara Rahman (@zararah) will share Inform/Transform, a zine about all of the ways that information is aggregated, controlled, and used. Their zine provides a helpful analog tool that can increase data literacy, complete with relatable examples.

Mimi and Zara will give a reading, discuss the thinking behind the project. Then, you’ll flex your creative muscles to illustrate your own experiences in data literacy.

When Recommendation Systems Go Bad
March 2017

Machine learning and recommendations systems have changed how we experience the internet, and by extension, many of the products and services we use in our civic and private lives. While the reach and impact of big data and algorithms will continue to grow, how do we ensure that people are treated justly?

As the people that build these systems, we have a social responsibility to consider their effect on humanity, and we should do whatever we can to prevent these models from perpetuating some of the prejudice and bias that exist in our society today.

Evan Estola (@estola), the Data Team Lead at Meetup, will share examples of recommendation systems that have gone wrong across various industries, and what can be done about them. Importantly, Evan will offer concrete technical approaches that can be employed. Evan will also offer arguments data scientists can use to justify making ethical decisions in a field that’s obsessed with optimization.

The journey from analytics to editorial
February 2017

Samarth Bhaskar (@samarthbhaskar) is an editor for Digital Transition at the NYTimes. He works with editors, journalists and the masthead in the newsroom to support editorial and operational decision making. He was previously a newsroom analytics manager at the Times, and an analyst at Etsy and Obama for America.

Hannah Poferl (@hannahpoferl) is a newsroom strategy editor at the NYTimes. She supports newsroom leadership through data and strategy projects. She has played a key role in the introduction and adoption of data in the newsroom since 2014. Prior to to The Times Hannah worked as an analytics lead for digital agencies including Brooklyn-based Huge and Threespot in Washington DC.

Journalism’s adjustment to the digital age has required a crash course in data and analytics. The editor’s hunch now sits next to realtime metrics on content performance across a dizzying array of channels.

How do leading publications like The New York Times adjust to a wide range of new competitors? How has data been introduced to newsroom staff and leadership? How has it affected newsroom decision making? What can be done to ensure that we protect the craft of investigative journalism while ensuring it finds and resonates with a meaningful audience?

Rescue Government Data
January 2017

Liz Barry, Sarah Lamdan, Bronwen Densmore, Bethany Wiggins, Laurie Allen, Jerome Whitington, Brendan O’Brian

A decentralized network of civic hackers, scientists, librarians, and archivists have come together to protect at-risk US government data. A war on empiricism threatens climate science and access to the important government data that makes it possible. This has inspired the formation of the Environmental Data & Governance Initiative, an international network of academics and non-profits dedicated to monitoring and preserving environmental evidence.

We’ll talk about the organizing work they’re doing to protect these data, and most importantly, how you can help fortify and rescue government data.

Best Practices in Pro Bono Data Science Project Scoping and Execution
October 2016

Raluca Dragusanu, MicroCred DataCorps Volunteer, Susan Sun, VOTO Mobile DataCorps Volunteer

It’s International Pro Bono Week and we’re excited to celebrate by showcasing our volunteers that have recently completed projects applying data science to tough social challenges.

The same algorithms and techniques that companies use to boost profits can be leveraged by mission-driven organizations to improve the world, but these projects must be responsibly scoped and executed to not only achieve maximum impact, but also mitigate potentailly damaging interpretations of the results. To explore this topic, we’ll learn about two recent DataKind projects focused on increasing accesibility to resources for underserved communities:

MicroCred: Financial Inclusion to Strengthen Communities Worldwide
VOTO Mobile: Giving Voice to the Voiceless Through Mobile Engagement
About Machine Eatable

Citizen-Led Social Experiments in Democratic Society
July 2016

J. Nathan Matias, Data Scientist, PhD student at the MIT Media Lab Center for Civic Media, a fellow at the Berkman Center at Harvard and a DERP Institute fellow (@natematias)

As data science is poised to offer substantial benefits to society, it faces deep public mistrust, especially large-scale social experimentation. Do experiments undermine democracy by advancing paternalism and favoring experts over citizen voices? What might an “experimenting society” look like if the public had an active role in the design and interpretation of experiments on important public questions? The roots of this debate go back to the 1930s, and they have a profound influence on how we do data science today.

In this talk, Nathan will share the unlikely history of arguments over the role of experimentation in democracies. He will also showcase early work on CivilServant, a system that supports online communities to conduct large-scale social experiments, developing mass-replicated open knowledge on community moderation questions, independently of internet platforms.

Discovering a path in data science and how machine learning can counter harmful speech online
April 2016

Warren Reed, Data Scientist, Office of Financial Research and DataKind volunteer (@JWarrenReed)

Moving from a career in finance to a career in data science was an easy step for Warren Reed. His discovery that he most enjoyed the quantitative aspects of working in the finance industry led him to a Masters degree in Data Science and in the first class at CUSP. Eventually, this lead to a job with the Obama Administration’s Office of Financial Research, he now works with any type of data that would benefit the entire financial community. With a strong sense of service Warren still wanted to give back and recently completed a project with DataKind and the Dangerous Speech Project. Hear Warren’s inspiring path into the data science field and the fascinating way he’s used machine learning to help counter harmful speech online.

Computation and/as Journalism
March 2016

Mark Hansen, Director, Brown Institute for Media Innovation at Columbia University (@cocteau)

In 1904 Joseph Pulitzer wrote at length about what should be taught in “The College of Journalism.” Data, or a turn of that century view of it, was on his list. “You want statistics to tell you the truth. You can find truth there if you know how to get at it, and romance, human interest, humor and fascinating revelations as well.” Fast forward 100 years and it seems that data and computation have lost their expressive voice. But data or computational journalism promises to change all that. These practices prioritize narrative and fill a gap (presumably opened by an over-emphasis on formal mathematics) between data and lived experience. In time, they will feedback to “official” data science, offering new tools and new methods endowed with the values and ethics of journalism. In this lunchtime discussion, Mark Hansen (@cocteau) will talk about the work he’s leading at Columbia’s School of Journalism, training new generations in a full-throated data practice.

What happens when you apply machine learning to social sciences?
February 2016

Hanna Wallach (@hannawallach), Senior Researcher at Microsoft

Computational social science is the use of computational and statistical techniques to study social processes. Hanna will open a discussion on the opportunities, challenges, and implications involved in developing and using machine learning methods to analyze real-world data about society. Among other things, Hanna will discuss fairness, accountability, and transparency.

Data, technology and refugees: the benefits and dangers of the new infrastructure for movement
January 2016

Mark Latonero, Fellow, Data & Society Research Institute (@latonero) and Paula Kift, PhD candidate at New York University (@paukif)

Designing data partnerships: how can big companies and researchers work together to share useful data?
December 2015

Mitul Desai, Director for Research Partnerships and Analytics at the Mastercard Center for Inclusive Growth (@CNTR4Growth) and Anoush Tatevossian (@artate), Strategic Communications & Partnerships Officer at UN Global Pulse (@UNGlobalPulse).

Data as a Process: How does our attitude towards data change if we see it as the result of a relationship rather than an end in itself?
November 2015

Mimi Onuoha (@thistimeitsmimi)

Interrogating Algorithms: how do we understand what’s going on inside the black box?
October 2015

Cathy O’Neil, Meredith Broussard, and Solon Barocas