Skip to main content
Exploring what’s next in tech – Insights, information, and ideas for today’s IT and business leaders

AI-driven inclusivity: The next frontier of art and the data economy

New episodes of The Element explore how tech innovation is busting Hollywood stereotypes and biases and ushering in an era of greater data democratization.

In new episodes of The Element, host Janice Zdankus, vice president of innovation for social impact at Hewlett Packard Enterprise, leads discussions on a range of topics related to digital equity and inclusion. More specifically, Zdankus and guests explore how technology innovation is driving changes across media, the data economy, and more.

Take the business of filmmaking for example. Hollywood is increasingly taking diversity, equity, and inclusivity seriously, using artificial intelligence and machine learning tools to better understand—and change—the biases and stereotypes that have long played out on the big screen. Applying AI analysis to the treatment of race, ethnicity, and gender in new scripts ensures movies are designed around DEI principles from the start, the experts say.

Please read: The rise of artificial intelligence and machine learning

In this season's first episode of The Element, Yves Bergquist, director of the AI & Neuroscience in Media Project at the University of Southern California's Entertainment Technology Center and CEO of AI engineering firm Novamente, details the tech behind this new way of storytelling.

It's science, not magic

"We're starting to see some interesting experiments around using natural language processing to understand emotions and the archetypes of minority characters. The industry is very focused on this and making big investments," Bergquist says. "Increasingly, studios want to have data-driven signals that can help them accelerate the conversation around what story is, because right now … there's a lot of frustration around this very human, very biased process of assessing narrative."

At a deeper level, Bergquist's team looks at patterns among clusters of characters in a script, using data to surface stereotypes.

"To do that, you have to dig into more semantic dimensions of content," he says. "We look at the emotions behind the language, and we're able to create these representations of emotional tonalities, emotional journeys, emotional arcs, emotions archetypes of various character sets, and we're able to tell you how close or distant a certain minority is from the pool of minority characters that have been tagged as stereotypical."

Please read: 3 steps toward a more digitally inclusive world

Bergquist and his team applied these analytic techniques to recent news coverage as well. "We found some real interesting stuff, like the coverage of the Black Lives Matter movement," he says. "In terms of what emotions were dominant in that coverage, it pretty much tracks with what you would think with Fox News vs. Al Jazeera vs. MSNBC."

Bergquist stresses, however, that his job isn't to push organizations toward specific content decisions. Rather, he says, "we give them signals that we hope are helpful to augment their human decisions with more context and more data."

"My passion is to better integrate this data-driven world with the world of humans and the world of creativity and nuance in the world of art," Bergquist says.

Data for all

In episode 2 of The Element, Zdankus turns to the concept of data democratization, whereby digital data is decentralized and open to all. Joining her is Matt Armstrong-Barnes, chief technologist at HPE, and Nathan Schneider, assistant professor of media studies at the University of Colorado, Boulder.

The guests look at how the control of most digital data by a handful of companies limits innovation and the ability of some groups to benefit from the value of that data. They also detail ways companies can develop equitable data strategies, including swarm learning, where data insights can be shared among many without compromising the security of the data itself.

Please read: What's all the buzz about swarm learning?

With data volumes rising to unprecedented levels, traditional ways of processing that data—and gaining value from it—are "running out of steam," Armstrong-Barnes points out. That means use of advanced analytics, AI, and other technologies is critical to organizations' ability to generate data-driven insights and apply those to their decision-making processes.

But it's not just about business, says Schneider, author of "Everything for Everyone: The Radical Tradition that Is Shaping the Next Economy." When it comes to the data universe, "we're talking about something public, and these decisions about the future of data are decisions that we should make … with a sense of the public good, the common good, at the forefront of our minds."

Please read: With decentralized clouds, the world moves closer to an all-inclusive digital economy

"This economy is producing new forms of centralization, producing some of the largest, the most powerful, and in many ways unaccountable companies the world has ever seen," he says. "We need a new social contract, a new sense of the expectations around data collection, data use, data processing, and the algorithms that are involved in turning that data into information that people are using to make decisions."

Breaking down silos

The experts discuss various approaches to this, including developing public, nonprofit, mission-based platforms as alternatives to profit-based corporate data sources, as well as cooperatives whereby ownership of data is shared.

Swarm learning is another way data can be used for the benefit of many, they explain—for example, in medical research, whereby hospitals can share insights from protected patient information without exposing the patient records themselves. "Swarm learning is critically designed to bring artificial intelligence into this space," Armstrong-Barnes says.

Please read: The rapid transformation of healthcare

Ultimately, the experts agree a future that ensures a more equitable data economy depends on better education and policy-based guidance on data usage and practices, corporate data strategy that reflects organizational values, and technology that securely opens data to the benefit of society as a whole.

This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.