The (Big) Data Economy: Inclusion and Fairness

This panel was part of The Marconi Society’s 2022 Decade of Digital Inclusion Symposium. Our expert group included moderator Danielle Davis of the Multicultural Media, Telecom and Internet Council (MMTC), Jordana Barton-Garcia of Connect Humanity, Sarah E. Chasins of Berkeley Engineering, Laura Chioda of UC Berkeley’s Institute for Business and Social Impact and Tiffany Deng of Google.

A replay of this conversation is available here.

Our expert panel shared their viewpoints on the topic of how we can harness big data to serve society.  

Danielle Davis, Tech and Telecom Policy Council for the Multicultural Media Telecom and Internet Council (MMTC) – moderator

We’re going to discuss how digital transformation, automation and globalization have sparked radical shifts in society and have given rise to a new economy, driven by big data and the Internet of Things. The expansion of this digital economy has placed big data, machine learning (ML), artificial intelligence (AI), and data science at the center of the debate about the future of digital inclusion 

Internet access is poised to reach hundreds of millions of new individuals over the next decade, bringing services and opportunities to historically excluded populations. Big data offers great potential for data-driven decision making to inform individuals, businesses, and governments, but it will be crucial to ensure transparency, equity, and trust. 

Algorithmic decision-making increasingly affects everyday life and the benefits must be weighed against the potential to codify and amplify existing biases, as well as the potential for fraud or the invasion of privacy.  This raises the question that we’re going to be discussing today. How can we properly harness data and its value to maximize individuals’ and society’s welfare?

Jordana Barton-Garcia, Senior Fellow, Connect Humanity

We only get to digital equity in the big data economy when we have equity at the most basic level with fiber-based networks for all, including rural, BIPOC and low income communities.  Fiber-based broadband enables 4G and the real 5G – intermittent signals / low latency – which enables the Internet of Things that support big data and AI for all. We cannot miss this moment. We must engage everyone in key decisions and not lose precious time focusing on things like minimum outdated speeds. We have the local ISPs and community partners who are rolling up their sleeves and are ready to get underserved communities to where they really need to be.  

Broadband is critical because it is an intersecting issue – it provides access to healthcare and education, the ability to start and grow a business and the opportunity to be part of the labor market in the digital economy.  Broadband is key to upward mobility.  We used to be able to enter the middle class with jobs in manufacturing and the retail industry.  Now our entry-level jobs are more digitally focused and require efforts like training young people in designing, building and maintaining networks, coding, IT and the like. Broadband also helps us with both the supply and demand side of workforce development.  For example, we cannot attract business to the border area in Texas because we do not have the fiber infrastructure.  At the same time, our country faces a shrinking middle class because we are not preparing people for the labor market.

My hope for the future of digital equity is that we have ubiquitous fiber-based networks for all communities and that communities have brought solutions to the table to help them create the right types of partnerships and networks for their needs.  When we make these kinds of investments in local communities, we create an inclusive economy and opportunities for underrepresented groups to use, create, and own assets in the digital economy. That is how we address wealth gap and wipe out persistent poverty.  

Sarah E. Chasins, Assistant Professor of Electrical Engineering and Computer Sciences (EECS) at UC Berkeley

My work is about ensuring that the power of programming computers and automating tasks is available to everyone, particularly teams that are doing very important work with low resources.  Being able to program computers lets people very quickly do some tests that would be tedious and time consuming.  I’m often working with teams who are trying really hard to change society for the better, but they have to do these automatable tasks very slowly and often manually with volunteer labor or using folks who are already stretched thin with their other responsibilities. I’d love for these teams to have the same access to programming and automation that high resource teams already have today. 

For example, we can help defense attorneys sift through police misconduct data in order to identify people who have previously lied on the stand and be able to make the case that those individuals should not be testifying against their clients. 

My hope for the future of digital equity is that we can give tools to the people who are already working to support vulnerable communities.

Laura Chioda, Director of Research at the Institute for Business and Social Impact (IBSI), at UC, Berkeley

Billions of people interact with the digital economy everyday.  Yesterday’s AI was prescriptive, like computers playing chess.  Today’s AI is interactive, accounting for nearly every variation in human behavior.  The digital economy has created huge amounts of information and we need to be able to process that data in a productive way that produces insights without jeopardizing privacy and security. 

AI can produce social benefits and improve lives in areas like weather tracking, determining where ICU beds are available and improving supply chain logistics.  AI can improve lives and financial inclusion. For example, I came to the US with no credit history.  AI can help us create gender-oriented models to account for the fact that women typically repay debt at higher levels than men.  This is a great example of the power of data to level the playing field. If we do not assess enough data for women, we will have inaccurate information about them and deny them opportunities that should be theirs.

My hope for the future of digital equity is that we use AI and machine learning to provide more opportunity and better options for people who are under-resourced and under-priviledged.

Tiffany Deng, Chief of Staff and Program Management Lead for Google’s Research Center for Responsible AI and Human-Centered Technology

I focus on technology from a responsibility perspective.  Just think about how ubiquitous things like AI are today and how we interact with it in so many different facets of our everyday life. It’s so important to ensure that there’s balance and that we’re thinking about the power that AI holds and the responsibility we have to understand how it affects different communities disproportionately.  We need to understand the toll it can take and have safeguards and guide rails there to ensure that technology is not creating an outsized burden on under-represented communities.  

AI is about using systems to mimic human behavior.  Machine learning (ML) is a subfield of AI that involves having systems that collect lots of data that is translated into models to predict behaviors.  This is where the term big data comes in and it’s important because it informs everything from credit worthiness to recommendations to college admissions and screening for job applications.  Because of this pervasive reach, people need control over their personal data and how it is used in all aspects of their lives.

The representation in the room where decisions are being made and models are being built is critical in bringing in diverse perspectives.  For example, if a company is deciding where to find new employees and there is not base equity in the room, AI can perpetuate skewed data.  

My hope for the future of digital inclusion is that we are successful with Google’s unified speech model, which will give people everywhere the opportunity to hold the power of the Internet in their hands.