The State of AI in Post-Production

Our CEO and Director of Post Production got together to weigh in on the state of the largest change in our industry in decades.

Consider Jurassic Park in 1993. Their initial plan was to film models as had been done for decades. When the first CGI dinosaur test was shown, its advantages were obvious. A threshold had been crossed. Many model and prop makers adopted the new tools and picked up their art in the new medium. This is where AI is at today for the media production industry. As media professionals, we need to adapt to all changes in the industry, AI being the current and largest development by far.  We see this as a sea change  of the industry, bigger than even  film to digital or digital to cloud. 

As a quick example for the uninitiated, with just the AI tools brought to market in the last 18 months we’re now able to:

1. Do 50-80% of the audio cleanup work in minutes that would have previously taken twice as long or more.
2. Replace, remove, or create visuals by describing the changes with text. This allows for rapid style development and testing ideas much quicker than before. 
3. Correct eye line when people are reading offscreen text.
4. Upscale most low-res footage to high-res.
5. Brainstorm video treatments in minutes instead of hours.
6. Speed up rough storyboards from days to hours.
7. And more, higher quality tools are being created weekly, if not daily.

The Gartner hype cycle is a good point of reference to consider. Our personal assessment is that for many AI tools we are dancing around the peak of inflated expectations.  Lots of R&D is still being done and there are mountains of garbage out there to sift through. Other AI tools have reached some version of the plateau of productivity and are here to stay.

To be honest, it often feels like too much to keep up with. These tools are being released at such a high cadence, it’s like drinking from a fire hose. Which ones are worth testing, and which ones are just a re-skinned chat-GPT? Which ones move the needle meaningfully in the right direction and on top of all of that, what ethical and environmental boundaries do we need to consider as we navigate our team’s adoption of these truly next generation technologies?

From a practitioner standpoint, we believe the best course of action is to test and learn. Learn all you can about these tools, and explore their strengths to avoid their weaknesses. With so many tools in the market, it can also be hard to sift. Matte Wolfe, Mike Seymour, Michael Kammes, Katie Hinsen, Two-Minute Papers, and others, are great people to watch, to begin to filter through the news. 

A helpful lens we’re using to digest all this is progress. With each new tool presented, consider if it is a better progression from what you're doing now, or is it just a new feature to market? Is the end result better than not using this? For many (not all) of these tools, the answer is yes. The work we’re doing is faster and easier by adopting these things rather than ignoring them. Ultimately - we all want to know - will using these AI tools help give me more free time to spend with loved ones or pursue passions? That’s the goal, right?

As we enter this new medium, consider if each new tool allows you to pick up your art and propel it forward, don’t ignore the possibilities. From both a business and ethical perspective, pay attention to standards. Initiatives like the Coalition for Content Provenance and Authenticity are helping to establish some much-needed guardrails for this new territory. The new AI Act in Europe seems like it could pave the way for legislation worldwide, much like it did with GDPR. The 2030 Vision Initiative from MovieLabs is gathering inertia from industry leaders. Keeping track of these high-level decisions helps set a good framework to place some of these other new developments. AI is a double-edged sword. As professionals, we need to be shepherding these tools to create great work while being mindful that powerful tools have powerful consequences if used incorrectly. 

Questions to consider:

What tools are you using to make your work/life easier?

What is AI doing in your industry now? How might that shift in the next 2 years?