As we say goodbye to 2021, and hello to 2022, it seems a good idea to take a look back at some of the
most notable advances of this year in the artificial intelligence systems used to assist in movie production.
The year was packed with huge news stories regarding AI-assisted filmmaking and other advancements in
artificial intelligence systems used in some form or another to enhance the movies that we all love so
much. Let’s get started.
AI Turns 20 Years Old
This year Steven Spielberg’s film AI (2001) turned 20 years old. Hard to believe that this film that
looks like it was made yesterday is now a classic.
https://www.hollywoodreporter.com/news/general-news/ai-robot-cast-lead-role-70m-sci-fi-film-1300068/
If you have not seen it then why not try it this holiday season.
More and More Accurate
Given that AI systems, once designed and launched, chew through data and learn at an incredibly fast
rate, the biggest but most unnoticeable advancement of the year most certainly lies ‘under the hood’ as the
saying goes.
The accuracy of last year’s artificial intelligence systems will have increased in proportion to the amount
that they have been used and trained. Accuracy is defined as “the nearness of a measurement to the
standard or true value.” – Source IGI Global
https://www.igi-global.com/dictionary/how-can-education-use-artificial-intelligence/329
Many of the leading AI-assisted moviemaking companies already entered 2022 reporting average
accuracy rates of between 70 and 85%, on average. Studies indicate that AI programs have the potential
to be at least 95% accurate on a regular basis, their average accuracy levels are usually somewhere around
the 80% mark.
Why is this? Well, it all comes down to the quality of the data that they analyze. (Alprazolam) In short, the more
accurate the data they analyze, the more accurate the results. So, while in 2022, AI movie systems have
greatly increased their ‘perspective’ or, out another way, their level of understanding relating to the tasks
to which they have been assigned, the quality of the data that they have had to work with has been
severely impacted by the ongoing COVID-19 pandemic.
Indeed, as all of you reading this already know, we are living in unprecedented times. Naturally, AI
systems have been waiting for the data to come in to make sense of how the pandemic is changing the
conditions under which they have been trained to make their predictions.
So while we might not be able to see those results at the moment, the real advancements made by AI
systems are likely to really start showing themselves in the new year when the data pools of this year and
are fully analyzed and understood by AI-assisted filmmaking systems. Watch this space for more.
The First AI Actor in a Leading Role
In June of 2020, the Hollywood Reporter broke the news that an artificial intelligence actor was to
star in the leading role of a science fiction film due to be released in 2022.
https://www.hollywoodreporter.com/news/general-news/ai-robot-cast-lead-role-70m-sci-fi-film-1300068/
Created by two Japanese scientists, Hiroshi Ishiguro and Kohei Ogawa, as part of their study of robotics,
‘Erica’ was taught to act and to seem as human-like as possible.
Erica was snapped up for a SCI-FI story created by visual effects supervisor Eric Pham, Tarek Zohdy, and
Sam Khoze, which tells the story of a scientist who creates a program to help perfect human DNA but instead helps the artificially intelligent woman he designed to escape.
Shooting has already wrapped up and the film is now in post-production. Expect its release to create more
headlines in the new year.
Leading AI-assisted Filmmaking Company Launches New Advertising Tool
Largo.ai, one of the leading global artificial intelligence-assisted filmmaking platforms recently added a
powerful new feature to its SaaS platform with which filmmakers and TV professionals can get an AI
assessment of their movie trailers and how different audiences will respond to them.
So, now, not only can filmmakers and producers use their AI platform to gain accurate insights into how
audiences will respond to various aspects of their films such as their scripts, characters, actor choices,
etc., but now they can actually apply the same accurate technology to their film’s trailers to see which
ones work with audiences and which ones don’t.
Going deeper, Largo’s genre recipe tool will actually allow film industry professionals to analyze their
trailers on a scene-by-scene basis just as they could already do with their scripts and video versions. This
will give them powerful insights that can be used to help them to improve their trailers.
This amazing tool was developed by Largo in order to increase the democratization of artificial
intelligence technologies that would otherwise be monopolized by a handful of big movie studios and
video-on-demand platforms.
AI Helps in Restoring Old Movies
The online magazine Wired reported an AI breakthrough where AI engineer Denis Shiryaev had
developed a system that was able to ‘restore’ degraded silent films.
https://www.wired.com/story/ai-magic-makes-century-old-films-look-new/
While there is a huge debate in the film restoration community regarding the accuracy of these
‘restorations’, and therefore, whether these can be called restorations at all, the results are stunning and are
a breath of hope for all those old film lovers who dream of seeing their films in all their former glory.
Shiryaev is the product director of Neural. love, a company that provides AI-driven video enhancements to
films. The results achieved by Denis are astounding.
Turning his company’s AI to a 1906 film called A Trip Down Market Street, Shiryaev was able to not only
‘restore’ the original but to colorize and sharpen the film to 4K resolution and to achieve a 60 frames per
second frame rate via frame interpolation.
You can judge the results for yourself by viewing the completed version of A Trip Down Market
Street via this link.
https://www.youtube.com/watch?v=1Ok_lwYyHWo
Stay connected