Ready to learn Artificial Intelligence? Browse courses like Uncertain Knowledge and Reasoning in Artificial Intelligence developed by industry thought leaders and Experfy in Harvard Innovation Lab.
Creative genius is 99% perspiration and 1% inspiration. Working in filmmaking, and more generally content creation, this becomes very clear. The story, which itself is creatively refined and tested over time, needs to be translated into a realized work of art. That sometimes involves hundreds of thousands of hours collectively from people behind the camera (or computer screen).
A common complaint I hear from creatives is dealing with tedious tasks that involve organizing, scheduling, managing resources and dealing with process. Artificial Intelligence has been highly impactful in several domains by unburdening this process-heavy work and creative industries are no exception.
So, what does this look like in the film industry?
It all starts with a movie script. Once a script is written, it’s first deconstructed in a process called script-breakdowns, wherein Natural Language Processing (NLP), trained on specific data sets, extracts all of the details pertinent to film producers, such as props and shooting locations. All in all, a grueling process that typically takes several weeks is cut down to a few minutes with 90% human-level accuracy.
Once the elements in each scene are extracted, algorithms and rule-sets can be employed to estimate fairly detailed budget estimates for production. The scheduling process alone—figuring out cast and location availability, cost constraints, and even weather—can lead to millions or billions of combinations to find the optimal timing. This is where AI, compared to manual, heuristics-based ways, can achieve at least 10-15% fewer inefficiencies, translating to significant cost and time savings for productions that sometimes exceed $100 million.
Can AI augment the “inspiration” part as well? The short answer is yes, as algorithms can now be used to analyze character development in a given script and identify plot holes and find continuity gaps at a glance. This is especially valuable for TV series that run multiple seasons.
The scientific community has also made tremendous strides in Natural Language Understanding (NLU) and Generation (NLG), based on deep learning. Advances in autoencoders and hierarchical structures lends more consistency and meaning while taking context into account. One impressive outcome of this is AI generating its own character dialog by utilizing context and producing more meaningful interactions, including humorous exchanges.
What simplifies dialog generation and other aspects of creative augmentation, at least to a point, is that human beings have resonated with the same fundamental plot structures and archetypes since Aristotle’s Poetics. But scoring storylines against known archetypes, and in the future even suggesting narrative changes to improve reader engagement, isn’t possible without human intervention: our years of living in the physical and social world helps us grasp meaning and understanding that is vital to storytelling. AI systems are still a long way from encoding the visceral and emotional knowledge of humans.
Nevertheless, disruptive AI technologies are significantly boosting creative productivity, and it’s not just happening in the movie business: the $450 billion content production industry is ripe for the picking, too. All companies from Fortune 1000s to boutique shops need to create content to connect with their clients. Business owners, content marketers, investors, anyone with a story to tell but little time to tell it, will soon have AI-powered tools to create high-quality content at a much faster rate.
