Why are media and technology companies like Netflix able to compete with the Hollywood studio establishment? One of their core competitive advantages is artificial intelligence (AI) — and it’s only going to get bigger. This is one of the questions that is core to what Accenture Interactive is examining as innovation partner at the StudioLAB at Walt Disney Studios, where we explore new entertainment experiences and production capabilities using innovative technologies.
AI is becoming more mainstream across the entire TV and film ecosystem. From the front stage (such as recommendation engines) through the creative process, scripting, shooting, post-production, and all the way to the back stage (such as meta-tagging and distribution), AI is enabling industry newcomers to leverage new business opportunities.
TV and film executives should take heed. AI can be used to eliminate current business inefficiencies, releasing trapped value which can be redeployed to pursue new areas of growth. Many highly repetitive manual tasks could be replaced by AI. Processes and technologies need to be fundamentally reengineered or developed to catch up with digital natives who don’t have the burden of legacy systems, business models and technology debt.
AI in Content Development
For example, AI can serve as an extra pair of eyes to provide script coverage, offer talent suggestions and as an extra pair of hands to generate visuals of the scene and sketches of characters to storyboard the situation. AI could even scout the next big trend. These tools are at too early a stage for studio executives to fully embrace yet, but there are some quick efficiency wins AI can already deliver during the creative process.
Simply by uploading the PDF of a script, Scriptbook.io can provide a detailed analysis on characters, detecting the protagonists and antagonists etc., within minutes. This can serve as a validation tool and as an additional voice at the table when creative executives are making decisions to green-light a TV series or film.
Beyond creative assistance, AI can also be used to scan a script for IP clearance and infringements, pull in image references for development or ensure brand compliance. IP flagging is particularly valuable as it would not only aid swift legal clearances upfront but could also preemptively avoid costly legal risks in later stages of development.
AI in Production
During principal photography for live action films, computer vision can come in handy to ensure continuity and quality control in the framing of shots through object recognition; e.g., spotting a boom mic in the frame, while NLP-driven capabilities could potentially help in dialog training and table reads.
For animation, AI solutions provide assistance in generating and completing sketches, poses and movements of characters, and background scenes. Done in collaboration with artists, this potentially frees them up for higher value creative tasks.
Some of the AI startups in this space could have very interesting implications for voice coaching and training for actors. One day their innovation could lead to simple virtual characters for new forms of media like XR.
AI in Post Production
Post-production is where AI can deliver immediate impact at scale by taking out much of the drudgery, reducing the cycle times and the multiple handoffs.
Using computer vision instead of manual checks means that hours spent by humans in quality control — perhaps to spot dead pixels or fix the aspect ratio in videos — can be reduced to a few minutes. AI solutions can also be applied in color correction, touch-ups, etc., in much more cost-effective ways.
AI tools have already been embraced by the world of VFX, a critical stage in post-production which can become a bottleneck, causing cost overrun. For live action films that rely heavily on VFX, advanced AI solutions such as Arraiy are already in the works to manipulate and extract images at the capture stage without the need to use green screens during the shoot. This delivers savings in time and cost during the manual and lengthy rotoscopy stage.
Advanced machine learning algorithms enable life-like creation and simulation of characters and scenes, reducing the hours of manual tasks. For scenes that require a massive crowd, VFX companies such as Mackevision can create and orchestrate the crowd just using AI-driven VFX work technologies.
For mastering, there are a plethora of AI solutions to choose from to generate metadata, subtitles and localization. These enhance data sets and discovery around media content. In the age of personalization, meta data becomes critical to understanding the genome of each content and match it to viewers’ interests.
For distribution services, based on the outlet the content is delivered to — such as hotels or airlines — AI tools can assist in ensuring contextual compliance, such as scanning for gore or nudity, and localization, using advanced machine learning techniques.
One day AI will be able to assist in editing the scene based on the dialogue and scene breakdown in a script.
AI in Marketing
Decision support is a critical area in marketing where AI can deliver the most value. When A-B testing, AI solutions such as Affectiva could help in intention reading and empathy understanding, using facial recognition and gesture reading techniques. Some are already augmenting cameras with ML capabilities to understand human body language.
For campaigns and promotions, AI-enabled programmatic buys can give studios better control in responding in real-time to social media scans. AI can complement targeted placements and marketing assets such as synopsis or one sheet, by appealing to the targeted user’s intentions.
It’s even claimed that AI now has the capability to create artwork and trailers by just watching the movie. IBM Watson created the first-ever AI-made movie trailer for Morgan.
AI in Distribution
There are many AI-centered solutions — like Pilot — that predict performance, such as box office numbers or TV ratings, for specific distribution windows. In addition to using historical industry data as a basis for analysis, these AI solutions take into account several external factors, like seasonality or socio-political-economic climate, to finesse their performance prediction algorithms.
In the future, AI-driven agents could become the intermediaries between the studio and the viewer, ultimately determining what content should be delivered to the user based on the viewer’s moods, emotions or activities. Studios must recognize this upcoming issue of becoming beholden to the algorithms and prepare to gain control over their distribution end point.
AI As a Core Part of Business
It’s time for traditional TV and film companies to consider AI as a linchpin in their evolution and transformation to the future. AI-driven initiatives are not to be treated as outsourced services, but instead should be understood, owned and operated as a core part of their day to day business.
Remember, during the early stages of digital video, many Hollywood studios decided to take a vendor-based approach for video services instead of taking ownership and investing in the infrastructure.
While the scope and promise of AI can be overwhelming, as a first step, entertainment companies should ease their way into AI by educating and evangelizing their key stakeholders about the art of the possible. They should embrace AI by going after opportunities that are clearly laborious — the low-hanging fruit in delivering savings.
In a cash-driven business such as content production and distribution, delivering clear savings is where the proof of the pudding lies. This will excite the rest of the organization to fully leverage AI to drive smarter decisions and launch new initiatives.
The sooner companies start embracing AI, the sooner they’ll be future-proofing themselves against the oncoming storm of competition.
Siva Natarajan is director of business design and strategy atFjord, Accenture Interactive’s design and innovation consultancy. Jeff Bauer is group director of client innovation at Fjord.
The television industry's top news stories, analysis and blogs of the day.