Heidi Boisvert
1,651,532 views • 7:49

For the past 15 years I've been trying to change your mind. In my work I harness pop culture and emerging technology to shift cultural norms. I've made video games to promote human rights, I've made animations to raise awareness about unfair immigration laws and I've even made location-based augmented reality apps to change perceptions around homelessness well before Pokémon Go.

(Laughter)

But then I began to wonder whether a game or an app can really change attitudes and behaviors, and if so, can I measure that change? What's the science behind that process? So I shifted my focus from making media and technology to measuring their neurobiological effects.

Here's what I discovered. The web, mobile devices, virtual and augmented reality were rescripting our nervous systems. And they were literally changing the structure of our brain. The very technologies I had been using to positively influence hearts and minds were actually eroding functions in the brain necessary for empathy and decision-making. In fact, our dependence upon the web and mobile devices might be taking over our cognitive and affective faculties, rendering us socially and emotionally incompetent, and I felt complicit in this dehumanization.

I realized that before I could continue making media about social issues, I needed to reverse engineer the harmful effects of technology. To tackle this I asked myself, "How can I translate the mechanisms of empathy, the cognitive, affective and motivational aspects, into an engine that simulates the narrative ingredients that move us to act?" To answer this, I had to build a machine.

(Laughter)

I've been developing an open-source biometric lab, an AI system which I call the Limbic Lab. The lab not only captures the brain and body's unconscious response to media and technology but also uses machine learning to adapt content based on these biological responses.

My goal is to find out what combination of narrative ingredients are the most appealing and galvanizing to specific target audiences to enable social justice, cultural and educational organizations to create more effective media.

The Limbic Lab consists of two components: a narrative engine and a media machine. While a subject is viewing or interacting with media content, the narrative engine takes in and syncs real-time data from brain waves, biophysical data like heart rate, blood flow, body temperature and muscle contraction, as well as eye-tracking and facial expressions. Data is captured at key places where critical plot points, character interaction or unusual camera angles occur. Like the final scene in "Game of Thrones, Red Wedding," when shockingly, everybody dies.

(Laughter)

Survey data on that person's political beliefs, along with their psychographic and demographic data, are integrated into the system to gain a deeper understanding of the individual.

Let me give you an example. Matching people's TV preferences with their views on social justice issues reveals that Americans who rank immigration among their top three concerns are more likely to be fans of "The Walking Dead," and they often watch for the adrenaline boost, which is measurable.

A person's biological signature and their survey response combines into a database to create their unique media imprint. Then our predictive model finds patterns between media imprints and tells me which narrative ingredients are more likely to lead to engagement in altruistic behavior rather than distress and apathy. The more imprints added to the database across mediums from episodic television to games, the better the predictive models become. In short, I am mapping the first media genome.

(Applause and cheers)

Whereas the human genome identifies all genes involved in sequencing human DNA, the growing database of media imprints will eventually allow me to determine the media DNA for a specific person.

Already the Limbic Lab's narrative engine helps content creators refine their storytelling, so that it resonates with their target audiences on an individual level.

The Limbic Lab's other component, the media machine, will assess how media elicits an emotional and physiological response, then pulls scenes from a content library targeted to person-specific media DNA.

Applying artificial intelligence to biometric data creates a truly personalized experience. One that adapts content based on real-time unconscious responses. Imagine if nonprofits and media makers were able to measure how audiences feel as they experience it and alter content on the fly. I believe this is the future of media.

To date, most media and social-change strategies have attempted to appeal to mass audiences, but the future is media customized for each person. As real-time measurement of media consumption and automated media production becomes the norm, we will soon be consuming media tailored directly to our cravings using a blend of psychographics, biometrics and AI. It's like personalized medicine based on our DNA. I call it "biomedia."

I am currently testing the Limbic Lab in a pilot study with the Norman Lear Center, which looks at the top 50 episodic television shows. But I am grappling with an ethical dilemma. If I design a tool that can be turned into a weapon, should I build it? By open-sourcing the lab to encourage access and inclusivity, I also run the risk of enabling powerful governments and profit-driven companies to appropriate the platform for fake news, marketing or other forms of mass persuasion. For me, therefore, it is critical to make my research as transparent to lay audiences as GMO labels. However, this is not enough. As creative technologists, we have a responsibility not only to reflect upon how present technology shapes our cultural values and social behavior, but also to actively challenge the trajectory of future technology. It is my hope that we make an ethical commitment to harvesting the body's intelligence for the creation of authentic and just stories that transform media and technology from harmful weapons into narrative medicine.

Thank you.

(Applause and cheers)