AI use cases in media industry

The Future of Media Companies

What comprises the “Media” nowadays? Newspapers have gotten much thinner if they are still publishing paper editions, as advertising revenues have dried up. Some have hunkered down, cut costs, and tried to carry on without much adaptation. Some deliver content online for free to customers, relying on traditional advertising wanting to access their customer base. Some have closed their doors after a century or more.
Other organizations, be it radio, TV, or Internet personalities, have tightly refocused their content to their core users, raising prices and shedding a general audience in favor of a dedicated fan base that knows exactly what they want. It may not be a big audience, but these organizations bring in lots of money through self-promotion and being controversial (e.g., Breitbart).
One such company, Aspermont of Australia, spends time identifying a niche market. They create a product specifically for that niche, such as mining, agriculture, or energy industry news, which people are willing to pay for. They recently started a Canadian publication for a mere fifteen thousand dollar outlay that is doing quite well. They don’t need the massive infrastructure of a traditional publisher since they regard themselves as a technology company sharing valuable information.
Magazines are extant, but how long will that last? A collection of paper pages with targeted news that has no permanent value? It’s all available online, and magazines are now publishing there, sometimes in parallel with a printed issue. Many magazines are also available for tablets/e-readers by subscription, eliminating the bulk of carrying actual magazines around.

Cable TV is under siege. They are losing channel-subscriber customers. Those have evolved to just using a cheaper broadband datalink to watch a much wider variety of programs from around the world at a much lower cost. Smart TVs (or adaptor boxes for “dumb” TVs) make connecting to the internet rather easy.
Cable TV fights back by providing free TiVo units, time-shifting, the ability to watch programs away from home on tablets, phones, or laptops, along with customized stations, not packages, so you can pick and pay for only the channels you want to receive. They’re adding VoIP or digital telephone services, and providing faster internet connections.

Developing Insight

In each case, you can see that all of these groups are attempting to target a specific audience. The problem, as always, is the amount of data they possess, and an inability to sort it to develop insights into their customers’ needs. This is the ideal use for Artificial Intelligence (AI) today.

The problems with modern media are multifaceted. Just as an example, let’s consider television news. There used to be people with powerful, sonorous voices, filled with gravitas (think Morgan Freeman), and when they reported something, it was uncolored, factual, and you just felt that it was reliable.
These included names like Walter Cronkite, John Chancellor, Chet Huntley and David Brinkley. Whether it was astronauts walking on the Moon, or the latest casualties from the Vietnam War, there was no doubt that what you had heard was true, or at least all the information that was currently available.
The most respected newsmen of the day agreed; they encouraged people to still read the daily newspaper and the news magazines of the day like LIFE and TIME. It was just an inescapable fact that they simply couldn’t go into significant depth in a single story. They had so many things to report in a limited amount of time that they could only provide a sketch of events.

Not a Diatribe

Nowadays there is so much competition for the consumer’s eye that the public has responded with short attention spans, and it is thus hard to hold an audience for any media. The news has turned into more of a “show,” with adversarial panels taking contrary positions, so that they can argue.
Viewers can then root for the opinion that they want to be true. And far from being a “news hour,” it consists of a few stories carefully selected and then designed to stir up the most controversy.
Why did this happen? For identity; so that people would remember a newscast and return to it. The news people themselves didn’t want it for the most part. It was driven by rating-obsessed executives.
Coloring the news to gain an audience was never an issue for those execs; the news and the delivery system was a commodity to be sold. There is some evidence that conventional anchormen/women were eased out in favor of those that agreed to be controversial.
Now our news stations are categorized into left- or right-leaning, and some are just cartoonish with people yelling opinions, employing over the top sound effects, with rapid-fire graphics, before declaring their opinions as facts. The idea of presenting facts and unvarnished truth seems to have fallen by the wayside.

Veracity vs. Verisimilitude

Of course, that is but a single example. Artificial Intelligence (AI) can help to recover credibility in all areas of media where they have lost their luster. We’ve all experienced CGI (Computer Generated Imagery) replacement of deceased actors in serialized movies, such as with Peter Cushing in STAR WARS: ROGUE ONE, or Natalie Cole singing “Unforgettable” with her deceased father Nat King Cole, blended into one continuous B&W scene.
That tells us that we don’t need an attractive figurehead or a particularly broad audience for any specific media. You can potentially deliver customized content provided via AI avatar, at any time, through any channel, and in any place a customer chooses. That has the potential to allow us to go back to untainted, unsensationalized, and truthful information. The truth is only educationally disadvantaged people don’t prefer the truth; good decision-making requires good information. Intrinsically, we know that fiction belongs in books, not in the real world.
Artificial Intelligence can drive the delivery of media to the consumer under any guise they choose. It doesn’t have to be “tuned,” but rather, just presented in a format they enjoy. Just to counterpoint Marshall McLuhan, we could specifically say that by making any delivery method possible “The Medium isn’t the message,” or doesn’t need to be, anymore.

Real Financial Benefits

AIs can also deliver content, without necessitating human action to make it happen. Google, Yahoo, YouTube, and social media like Facebook have all been doing so for decades now. Their model is based on serving content fast and automatically.
What we’ve learned from them is how to mine preferences from customer choices so new (but relevant) options can be presented for consumption. The sheer bulk of data requires an AI. Sorting and classifying is beyond the capabilities of humans now. The right advertisement in the right place can trigger a sale, but we have to know who will convert when they see it.
The model works well for Netflix, of course, in a contained environment. AIs are now competent (and getting better) at analyzing voices and video images to assess and classify it. Video services use AIs to label something “Family,” “Comedy,” “PG-13”, “Adult,” “Drug, alcohol, or tobacco use,” or “Extreme Violence” because they can watch an entire movie in minutes. They start by reading the “closed captioning” files to analyze dialog, figuring out if there are “bad words,” and then assess actions to identify things like “sexual content” or “violence” with very clear parameters for the rating system. It eliminates human error from boredom or inattentiveness.
There is so much content being generated every day that AI is the ideal candidate to wade through it all. It frees up humans to do more productive work. No sane person wants to (or could) watch three or four movies per day, of varying quality, where you can’t turn it off while rating it for content.

Real World

Applied to ordinary daily business models, an AI can gather digital crumbs of information and create an accurate profile of each consumer. Some may call it a privacy issue, but as long as the data is freely available, or specifically available to your company with appropriate authorizations, it is allowed.
For instance, people on eBay know that when they buy an electric hand-mixer that they will start to see more ads for kitchen appliances. Notably, they won’t see ads for more hand-mixers because an AI will know that they probably don’t need two similar units. The information is recorded, anonymized, and then used for marketing to outside agencies while keeping the PII (Personally Identifiable Information) only in the hands of someone with implicit permission to have it.

Profiles Online

People spread their information far and wide with little regard to their privacy. AIs using neural networks, deep learning (DL), and machine learning (ML), can gather this public data and compile profiles. Once you post something online, it exists forever. Eventually, all our preferences will be almost universally known, and we’ll be targeted with tailored ads all of our lives.
People that possess the profiles will be able to create customized lists and distribute ads to likely customers without revealing the information directly. If the customer contacts the advertiser, they then have the option to identify themselves and conduct business. Whoever owns the lists will have something valuable to sell because they took the time to build an AI to compile it for them.

Fictional Media

Entertainment will continue to be an important market, where we are free to lie creatively and fulfill fantasies. AIs will allow us to embed the viewer into the experience. The will be fully immersive experiences, where the viewer can passively exist within a 3D environment, as a virtual member of a crowd, or take the place of any character, and see it from their perspective. This sort will soon predominate but would be impossible without AI.
In fact, the headgear, loaded with accelerometers, orientation sensors and high-resolution graphics, already exists. Now it is just a matter of creating a product for them in what developers call a “sandbox environment” where there is the main story, but the participant is free to go anywhere in the environment and interact with different features.
This is highly reminiscent of the 1997 video game BLADE RUNNER, which had 13 different endings for the game, depending entirely on your choice of action throughout. It won the Interactive Achievement Award for “Computer Adventure Game of the Year.”

Delivering the Right Content

AIs will help us identify the right day to deliver content; they’ll identify the right time to deliver content; they will make content delivery appropriate for each consumer. Your Saturday activities probably vary widely from your Wednesday activities. You may be more likely to buy grass seed on Saturday but look for a good cultural lunch experience in the city on Wednesday.
Those things are not likely to be interchangeable unless you happen to be in the city on Saturday at noon, or at home in the suburbs on a Wednesday. AIs will be sensitive to the time of day, your location, what sort of activities you conduct there, and even be able to assess your daily schedule to present relevant content.

The Takeaway

AIs will drive the presentation of product to consumers, whether by the class of all items or by a specific product, in a particular situation. The days of a generic delivery system are pretty well over, and the era of customized, personalized content delivery will dominate the foreseeable future. AIs are not going to take over the world, but they can take over the more dreary and burdensome tasks of identifying your ideal customer and then manage the function of automatically serving them content.
If you want to learn more, we provide webinars to show you where the future is going and how you can get there. Don’t shortchange yourself. Join us today!