How Your Data is Being Used to Shape the Future of Media & Entertainment

A look at AI and the Data Production Landscape, with speakers from the Data Science Salon in NYC.

When it comes to creative production, the speakers of Data Science Salon NYC agree that AI isn’t going to completely replace humans anytime soon.

But: there are a number of ways that AI has changed the data production landscape. “Sometimes it’s in small ways — the algorithms powering the tools in your Adobe creative suite,” offered Justin Hendrix, Executive Director at NYC Media Lab. “Sometimes in larger ways, such as tools that automate text generation or the production of video clips from a corpus of footage.” And when coupled with data science, it’s providing “a deeper understanding of what drives consumer engagement with creative video/media, and providing and recommending desired content to them on a personalized basis,” said Chris Whitely, Senior Director of Applied Analytics at Comcast.

“People not only expect amazing content — they expect it to be tailored to their specific tastes. In television, we’re a far cry from the days of only having a handful of channels that had programs appealing to most of the country,” said Josh Miller, Director of Data Analytics at Samba.tv. Now, we have the opposite problem. “It’s becoming a chore to search through it all to find what we want to watch. That’s where data science and machine learning can step in to help surface some of the best content for us.” For example, one application for media companies is “creating metadata to content and compare scanned image recognition and use AI to do mass personalization,” according to Ayan Bhattacharya, Advanced Analytics Specialist Leader at Deloitte Consulting.

Looking back, we’ve come a long way. “In the old model, you would first have to map viewership trends to a demographic group, then map your purpose to that group,” said Miller. “With large TV viewership data sets, I don’t need to know your age, gender, or ethnicity — if you like sci-fi shows, then I know you like sci-fi shows.” And that means better, more personalized content for each audience thanks to the help of AI and data science.

Protecting the data that drives innovation

When it comes to data privacy and governance, our speakers agree that there’s no compromising on what is ethical. “Data privacy is about being very clear with consumers about what data is collected and why, giving them an option to participate, and then making sure that data is protected,” offered Whitely. One way to make sure consumers understand how their data is being used is “by always being opt-in, not opt-out,” said Miller. “I also think it’s important that companies write language around their opt-in process in plain English that describes what data is being collected and why as well as what users get out of it in return.”

And that helps to protect the company’s relationship with the people who give their data. “Our first-party data comes from people who actively decide to join our panels or respond to our surveys,” said Lauren Lombardo, Senior Data Scientist at Nielsen. “These respondents join our family because they want to contribute to market measurement. We realize how special this relationship is and how important it is to protect the privacy of those who trust us with their information.” Consumers are fickle, and won’t hesitate to jump if they feel threatened. “There’s a lot of competition in this space,” siad Miller. “Some companies may leverage people’s personal data for large short-term profits, but the only companies that will survive in the long-term are the ones that customers feel they can trust continuing sharing their data with.

And once you do have the data, there are important steps to take to make sure it’s used for its intended purpose. Harini Kannan, Data Scientist at Capsule8, ensures “user data (labels) are anonymized. We are also in the process of adding more layers of data encryption as our research develops. I think it’s the responsibility of everyone in data science to make sure there is a rigorous process around data collection / feature engineering.” If you protect the consumer where you can and make transparency a priority, they will reward you with their loyalty.

The future of AI in Media & Entertainment

It’s never easy to predict the future, and the speakers of Data Science Salon New York have somewhat divergent views on how we might expect ML and AI to be applied to the field of Media and Entertainment in the next five to ten years. But there’s one clear area of agreement: “We’re going to continue to see the implementation (and improvement) of personalized recommendation algorithms that are based on person-level data.” said Lombardo. “It will vastly improve the overall relevance of the content and advertisements served across premium video, and make content exploration much simpler and more enjoyable,” suggests Whitely.

But the recommendation engines of tomorrow will be even more robust. “We’ve barely scratched the surface of utilizing recommendation engines for generating new content,” said Miller. “We’ve already seen some shows like Netflix’s Black Mirror: Bandersnatch utilize the idea of letting viewers make their own decisions to choose their own storylines. What if the viewer didn’t have to make the choice, but instead AI could generate which story suits you best?” Lombardo agrees, “soon we’ll expect to have content that fits into the tiny sliver of a specific sub-genre we crave at that moment hand delivered to us through a personalized recommendation that is unlike anyone else’s.”

Outside of recommendation engines, the future of AI & ML in Media and Entertainment is more nebulous. “In five years I definitely see more engagement with 3D settings and audio visual interactions with personalized home assistants,” said Bhattacharya. “In 10 years, I foresee AI in media and entertainment profiling individuals’ social aspirations between viewing offerings and experiences in a more prescriptive manner.” Hendrix goes even further, “We will see the emergence of synthetic celebrities, and narrative experiences generated by the output of models. This will both drive down the cost of content production at the low end and drive up the value of highly produced, artisanal content at the high end.”

But there’s one additional area where all of our speakers agree: “Humans will continue to have the upper hand; most estimates suggest job growth in the media and entertainment sectors will continue to be strong over the next few years,” said Hendrix. “Perhaps all the automation taking place in other sectors of the economy will create more opportunity for media overall!”

Don’t miss your chance to see amazing speakers at our virtual events→ Data Science Virtual Salons 2020:

- DSS Elevate Virtual | Women, Data, Tech (July 30, 2020 // Online) #DSSElevate

- DSS Elevate Virtual | Women, Data, Tech (August 27, 2020 // Online) #DSSElevate

- DSS Virtual Salon | Media, Advertising, Entertainment (September 22–25, 2020)

- DSS Virtual Salon | Retail & Ecommerce (Nov 17–20, 2020)

- DSS Virtual Salon | Travel, Finance, Technology (Dec 8–11, 2020)