trendswerespotted

[trendswerespotted]

Final thoughts on synthetic actors

Personally, I think that the rise of synthetic actors is inevitable and that it will continue as fans are shocked with each passing of their favorite celebrities. The truth is, synthetic actors are already here, although they haven't been explicitly labeled as such. We've seen them in movies, in video games, and we are now seeing them posting, commenting, and interacting autonomously on social media.

Like I've mentioned before, the technology isn't the limiting factors here. It is already exceptionally advanced and will only continue getting better. Humans are the true obstacles to the ultimate implementation of synthetic actors. Either through legislation or ethical concerns, people may fear this new unknown. Creating AI models that look, sound, and act like us truly does make it feel like robots are going to take over the world and annihilate the human race. Just joking... mostly. In all seriousness, it is scary. What happens when we're not able to decipher between what's AI and what's not? This is an issue that's already happening in schools and on the internet as people spread "fake news" or use deepfakes and AI to post inaccurate/fictional scenarios (see this recent video rapper 50Cent released that uses AI to portray Jay Z and P. Diddy in handcuffs in light of the recent sexual assault allegations against Jay Z. Jay Z has not yet been arrested). With every new invention, people will find ways to abuse it, but that doesn't mean we should stop inventing. Instead, how can we create legislation and guardrails at the same pace that this technology is advancing? These AI actors are a sign of progress and open up so many opportunities for engaging with (customized) content, and may even benefit actors who have busy schedules or have gotten injured during a shoot. Maybe it can even make filmmaking more accessible to smaller studios who cannot afford the real life Dwayne Johnson, but can license his virtual twin... For the most part, I see this technology as a good thing that we should keep developing and exploring, but again, it's crucial that we dot our i's and cross our t's before simply releasing it out into the world.

[trendswerespotted]

The Fashion Industry: another opportunity for synthetic influencers

In one of my previous posts, I talked about AI actors across different industries. I mentioned synthetic social media influencer Lil Miquela, who has actually partnered with fashion brands like Calvin Klein and Prada and even appeared Vogue magazine. This inspired me to add more information about the evolution of AI within the fashion industry, and its exploration/current use of AI models.

Several major fashion brands have already been experimenting with AI, and any brand that's just doing it now has already fallen behind. Italian luxury fashion brand Moncler teamed up with Maison Meta on the brand’s first AI-powered campaign. The AI generated visuals shown below show AI models wearing Moncler inspired, AI designed puffer jackets (one of Moncler's most popular products).

The beloved denim brand Levi's partnered with Lalaland.ai to create synthetic fashion models to wear and advertise their clothes on the brand's website. The campaign was met with public backlash as critics raised concerns about replacing human models for the sake of cutting costs. Additionally, Levi's seemed committed to using AI to develop diverse and inclusive representations of models, however "many critics wondered why Levi’s wouldn’t simply hire real models, rather than generate fake ones to promote diversity" (The Independent). I think this reaction shows that perhaps the technology is ahead of its time. Although AI is capable of generating realistic fashion models, the pendulum is still swinging towards authentic, diverse, and inclusive representation. People would much rather see other humans get paid for this work than witness companies cut costs to experiment with AI. Although I agree with some of the concerns about the implementation of this technology, I don't think AI models as a whole is an inherently bad thing... I think it's just a matter of finding a balance of when and where it is appropriate to use AI models vs human models in the fashion world (and this of course applied to our AI/human actors as well in the long-run). Like what we've spoken about in class, it's a matter of introducing people to something that's new and unfamiliar to them. Company's need to reassure them that even though this technology is here to stay, it will not diminish or replace the value of human work (but rather, enhance it).

[trendswerespotted]

Mass-customization

Mass customization, the process of producing goods and services tailored to individual customer preferences at a scale and cost similar to mass production, is revolutionizing various industries (think customizable Nike shoes or even Chipotle's "build your own bowls/burritos"). The concept of mass customization is nothing new, but we're seeing it converge with media companies like Netflix and Tiktok who are using algorithms to tailor content to our viewing preferences and interests.

But why should companies even bother with painstakingly customizing their products to individual customers? Doesn't that seem like more work? Well, not really. These consumers are limited to a specific variety of options to pick from. For example, people who are designing their own custom Nike shoes are limited to the company's color choices/combos, shoe styles, parts of the show that are able to be customized. Netflix users are of course limited to Netflix content (whatever is showing on their platform that aligns with their watch/like history). Mass customization simply gives a way to maximize customer satisfaction without exhausting a company's resources. We also live in an age where we think we're special - we want everything from our shoes to our chipotle bowls to our tiktok fyp's to be perfectly curated to our individual identities. Businesses have always wanted to learn as much about their consumers as possible and mass customization has given them an excuse to continue gathering consumer data because if we're going to see ads, we might as see ads for things we're actually interested in. Otherwise, it's a waste of time and money for businesses to be targeting the wrong people.

As this trend continues to evolve, it's intriguing to consider how synthetic actors might fit into this landscape. Will people only be able to choose from a small variety of AI actors/films? Will synthetic media ever be fully customizable?

[trendswerespotted]

Case Studies: Synthetic Actors in Short Films

Although they're not quite ready for the big screen yet, various companies are already creating teasers and short-films to show off their synthetic actors and environments. The examples below are impressive benchmarks that raise the standard of hyper-realism and give us a glimpse of where this technology is currently at in terms of realism/visual capabilities.

1. The Heretic Short Film

Made by Unity Technologies, "the film uses every aspect of Unity’s High Definition Rendering Pipeline, features advanced effects created with the VFX Graph, and showcases Demo Team's first realistic digital human." What truly stands out to me is not the realistic digital human (who, in my opinion, looks more like a video game character), but the amount of detail and realism that brings the scene locations to life.

2. Enemies Cinematic Teaser

Holy smokes. It's hard to believe that the environment and the protagonist aren't real. I'm especially impressed with the close-up shot of the character's face, showing her smile lines, lines around her eyes, textured skin, and even subtly beauty marks. I think the only "critique" I have is that in some clips, her teeth don't look quite as natural as the rest of the character. Because of this, her mouth movement is ever-so-slightly off. Her hair also doesn't fall the same way human hair does, but I love how they got the highlights right! If this is the current state of virtual actors/AI filmmaking, I think within 5 years we'll achieve hyper-realistic synthetic actors. Perhaps within 10 years, they'll be ready to star in feature-length films.

3. Blue Dot Short Film

This film also toes the line between looking like hyper-realistic video game characters or real people. The project was created using Unreal Engine's MetaHuman Animator to accurately capture and replicate real human emotion/expressions.The character was generated by AI, with help from actor Radivoje Bukvić to provide the voice. Perhaps in the near future, there will be a seamless convergence between audio AI and AI actors to create 100% fully synthetic celebrities, both birthed and voiced by gen AI.

[trendswerespotted]

Synthetic Acting vs Traditional Acting

As it often is with the introduction of new technology, people are resistant to replacing the old, tried-and-true ways of doing things. In this post, we'll consider the pros and cons of both traditional (in-person) and synthetic acting. While I don't think synthetic acting will fully replace traditional acting, it will be interesting to see which projects studios elect to use synthetic actors vs traditional actors.

Traditional Acting

Advantages:

Emotional Depth: Human actors bring a level of emotional depth and authenticity that is hard to replicate with technology. Although facial capture is becoming more accurate, nothing beats the real thing. There are certain nuances, quirks and facial ticks that occur as natural responses to our environment/situation.

Creative Chemistry: The interaction between human actors can create a chemistry that is difficult to achieve with synthetic actors. This chemistry can enhance the overall performance and ultimately make the story more engaging. After all, authenticity is key to a good story and performance.

Disadvantages:

Cost: Hiring human actors can be expensive, especially for our A-list celebrities.

Limitations: Human actors have physical and time limitations. They can only perform what they are physically capable of and are limited to their schedule availability and labor/union laws. They can also get sick or injured, which can push back filming schedules.

Synthetic Actors

Advantages:

Cost-Effective: Once the initial investment in creating a synthetic actor is made, it can be significantly cheaper to re-use the synthetic versions rather than having to re-hire the human actor for future projects.

Flexibility: Synthetic aren't bound to safety precautions - they can perform impossible and dangerous stunts that are out of the realm of human capability.

Creative Control: Filmmakers have complete control over the appearance and actions of synthetic actors, allowing for precise direction.

Disadvantages:

Realism: Despite advancing rapidly, synthetic actors still lack the full range of human emotions and subtleties. Sometimes they are stiff and unnatural, and it's apparent that they are computer-generated.

Ethical Concerns: The use of synthetic actors raises ethical questions about consent, digital rights, and the potential displacement of human actors.

Technical Challenges: Creating convincing synthetic actors requires sophisticated technology and expertise (and therefore deeper pockets), which can be a barrier for smaller productions

[trendswerespotted]

CGI: the building blocks of AI Actors

So far, I still haven't decided how to refer to these digital beings - AI Actors, virtual twins, synthetic celebrities.... Although it feels very new, it really stems from CGI technology which has been around for decades (its origins are debatable, but many people think that Hitchcock's Vertigo was the first to use basic CGI). There are many ways CGI has been used throughout the history of cinema), but for our purposes I'll focus on a few different milestones regarding the development of this technology:

The first 3D animation of a human hand and face (1972): this would lay the foundations for our modern day realistic digital twins (I would have included a photo of the face... but honestly it's a bit too creepy for me).

First fully CGI character, Young Sherlock Holmes (1985): this film features a short sequence where a computer generated stained glass knight bursts from its window and becomes a part of the live action environment.

2. First texutred CGI, Jurassic Park (1993): "Only four of the 14 minutes dinosaurs appeared onscreen were computer animated, the rest were...modelled by hand and the CGI had to be textured to match."

3. First photo-realistic human actors, Final Fantasy: The Spirits Within (2001): Although photo-realistic has a different meaning today, the rendering of Dr. Aki Ross in the Final Fantasy movie is very very impressive for technology 20 years ago.

4. Motion capture and CG artificial intelligence, LOTR (2009): newly created software transformed actor Andy Serkis into Gollum (the stuff of nightmares)

5. Facial capturing, Avatar (2009): Head rigs worn by actors captured their facial performances to create emotive CG faces (limiting the uncanny valley effect).

As this technology has continued to advance throughout the years, we have come closer to successfully create photo-realistic digital characters whose emotions, movements, and physical details appear more natural.

[trendswerespotted]

Jumping on the AI Actor bandwagon - other industries that will adopt this technology

Let's say this does become a thing. I can already see the major streaming services either buying or investing in whatever company makes these AI actors so that this service is fully integrated and exclusive to their platforms. Considering the contracts that actors would need to sign with these services, maybe this will lead to the rebirth of Hollywood's studio system, where their AI twins can only produce movies with the studios they are contracted with. For example, maybe AI Anne Hathaway will only star in Hulu Originals while AI Meryl Streep is in all the Netflix movies. These services will probably add a section that is solely dedicated to their AI content.

As mentioned in the AI social media influencer post, maybe these digital actors will have their own social media accounts (like on Butterflies AI). They can autonomously post content and interact with fans on social media. Although I have to say, I absolutely love behind the scenes content (sometimes I enjoy it more than the actual movie), and I know it's a big part of press tours and interviews.... so I wonder what these AI actors would even post or how they would promote their movies? How would they do interviews if they technically didn't act? Anyways, I definitely think the social media industry will continue to embrace these AI actors as it already has with synthetic influencers.

The music industry is already seeing AI songs become popular (and a few have already gone viral on tiktok). I could see these gen AI platforms working meticulously to craft the perfect synthetic celebrity who is good looking, can act, can sing, can interact online with fans in a charming/personable way, and will avoid PR nightmares.

I think the music industry and social media go hand-in-hand with fame and fortune, so it's natural for them to eagerly and seamlessly adopt this AI actors as soon as possible to attract new listeners and users to their platforms. After all, these AI actors can't complain about being overworked.

[trendswerespotted]

Obstacles and Barriers of AI Actors

Honestly, the biggest barrier I see that can stop this trend from being realized is legislation. There are a lot of people with money and power who would want to prevent any possible misuse of their image and likeness (and rightfully so). Perhaps this is one of those situations where fear will override the advancement/implementation of technology, but maybe it could be a good thing in this case... at least, I think there are absolutely some guardrails and protections that should be legislated in order to protect one's (virtual) reputation, which in turn will inevitably reflect on their offline one.

Another barrier might be that the pendulum will swing back towards desiring true, authentic content faster than we anticipate. Perhaps people will be immediately turned off by this idea of AI actors (aka, actors who aren't even really acting). Anyways, why pay money and let actors make millions off of movies that they didn't even technically star in? I can already hear people complaining about how this takes away from the true art and craft of acting (and I suppose that's true to a degree).

Finally, I this technology might be better suited to the video game industry, where people already show an interest in mass-customization. Many video games not only allow you to fully customize your own characters, but also allow you to control the storyline based on dialogue options and gameplay interactions/choices. Maybe people prefer the interactivity of video games to be balanced out with the non-interactive/static nature of television and film. Perhaps there's an enjoyment in being passive every once in a while - we already make enough decisions in our day-to-day lives, just let me watch some damn tv. If people wanted to choose who to cast all the time, then they would've become casting directors.

[trendswerespotted]

Virtual Twins aren't limited to film

A few months ago, Professor Peruta posted a linkedin article about his experience using the butterflies app - "a social media platform where humans and AI-driven users coexist and interact." This, combined with the rise of synthetic influencers (like Lil Miquela, who has 2.5M followers on instagram), shows that there is an interest and fascination in exploring synthetic influencers on social media.

Of course, the next step to stardom for anyone who finds fame online is either signing them to a record label (like Mason Ramsey, who went viral for yodeling in a Walmart) or having them act in a commercial, movie, or show (like the reality TV show based on the oh so difficult struggle with fame so bravely endured by Tiktok influencers Charli and Dixie D'Amelio). It's natural that this would apply to synthetic influencers as well. Maybe Lil Miquela will be signed to a movie deal soon (or however that works).

Another area I've seen virtual twins is in the medical field. For my SXSW internship, I came across BioTwin, a startup that creates virtual twins for medical testing. It uses an at-home biomarker kit to create a "virtual you in the metaverse allowing to help simulate and predict health outcomes, enabling personalized healthcare and proactive interventions. It helps monitor health, detect patterns, and optimize treatments." Its sister company, TwinMe, allows users to track their health in several key areas over time. These medical biotwins are of course, more focused about what's on the inside of our bodies than the outside - I'm not sure if you can visually customize the physical appearance of your twin to make it an accurate image of yourself, but it'll match whatever health conditions you have going on inside of you. This seems to be a completely different context of "virtual twins," but perhaps all our different definitions of these synthetic beings (including AI actors/influencers and these biotwins) can culminate into fully accurate virtual replicas of real humans. Or, perhaps it can be used to create virtual humans with medical histories and conditions that can autonomously post online and interact with others.

Maybe instead of robots taking over the world, it'll be our very own digital twins o.0

[trendswerespotted]

Taking a step back into the present

Sure, the previous posts have all been hypotheticals and me dreaming big about what this Gen AI/fan casting/virtual twin technology could be. But what does it look like right now? How long will it take for it to do all of the things I've been talking about (create AI versions of actors/individuals who can adapt to different genres, accents, and even an individual's unique acting style)?

Right now, there are a few platforms (like Tavus and Veed.io) that are capable of creating realistic digital twins. Both platforms can be used to create custom avatars based on recorded audio and video footage/photos. Veed.io even has several template avatars with different clothing styles to choose from. You can also type a script for your avatar to say. Tavus also (allegedly) works in around 30 different languages. While these avatars are impressive, it is still a bit noticeable that they are not human. I think there's a bit of that uncanny valley effect, especially when it comes to their facial expressions while talking and general body movements. It doesn't quite match up, but before we know it these tweaks will be fixed and their motions will seem fluid and natural.

Sora is capable of creating high quality, realistic video from text. Although it cannot currently do virtual twins, it does give us a good idea of how well gen AI can create videos (really well).

I also wanted to mention Google's Notebook LM, mostly because their AI podcast feature is by far the most natural sounding AI voice technology I have heard. The cadence and natural back-and-forth conversation makes it hard to believe these aren't real people talking. This podcast feature is mainly used to summarize/explain notes and other content, so there is less control over the verbatim output.

These are just a few examples of what some Gen AI technology is currently capable of.

answer to the image above: the photo on the left is the Tavus-generated virtual twin, based on the training video on the right.

[trendswerespotted]

Ethical issues of AI twins for entertainment

For this AI fan casting service, it's not the technological capabilities that will be holding it back, it's the ethical implications. As usual, the technology will be advancing faster than the laws can keep up with.

So, what could possibly go wrong with AI generated actors, and using their name, image, and likeness however viewers please??? Here are a few scenarios off the top of my head:

Sexual content: pornographic deepfakes of celebrities and individuals has been a longstanding concern since the genesis of this technology. Since the Taylor Swift Deepfake Controversy, calls for appropriate legislation against creating and sharing deepfake pornography have become more urgent. Does this mean Gen AI recreations cannot be produced for any content featuring sex scenes? What would these restrictions look like on the our non-existent AI fan casting platform? Removing these scenes completely, or having a complete block on any film/television shows that feature sexual content (for example, not having 50 Shades of Gray as an option for Gen AI/customization)?

Child actors: while child actors are heavily protected by child labor laws, could this apply to an AI version of themselves? What if people aren't casting these child actors in age appropriate roles (ultimately abusing the name, image, and/or likeness of these minors)? Personally, I don't think child actors should be available as Gen AI recreational options on this platform. I think it's something actors sign up for on their own accord once they turn 18, so they themselves can determine the terms of their own contracts (instead of allowing someone to do so on their behalf).

Posthumous representation: we've seen CGI/AI audio recordings of late actors/celebrities before (like the AI audio recording of Anthony Bourdain for his Roadrunner documentary), but is it ethical to continue profiting off an individual once they are deceased? Especially if their estates are licensing the individual's NIL to projects they might not have consented to if they were still alive? I think this should be up to actors/celebrities/any public figure to decide while they are still alive... According to this article, the estates of Judy Garland and Burt Reynolds have signed with "AI voice cloning startup ElevenLabs... as part of the company's 'iconic voices' project" so their voices can be used to narrate books and articles... until their voices are (perhaps inevitably) used to say hateful, controversial statements that they have never said in real life...

[trendswerespotted]

It's all about the money, money, money: the monetization of AI Generated Actors

So here's the scenario: an AI fan-casting service has been discovered and has now gone viral. At first, it only had contracts with D and C-list celebrities, but now everyone wants in on the fun. After all, the more celebrities that are interested, the more users/viewers this platform can bring in. It's a great money-making opportunity (as if the rich aren't rich enough), but how will it work?

Once the platform has enough A-list celebrities, the platform and its content will be monetized via a tier-based subscription service. The lowest tier will allow for X amount of AI generations per month, with a limited selection of actors. The next tier will grant more access to a wider variety of actors (including A-listers) who are partnered with the Gen AI/fan casting service, in addition to more generations per month. Similar to film studios - you need big bucks if you want big names for your project. The final tier will allow for full customization, allowing viewers to create their own AI models of themselves to star alongside their favorite celebrities. The service will also allow a way for people to share their castings/tv or film recreations with others, and in turn, celebrities will be paid for use of their NIL on these platforms. They can also be paid residuals every time their films are shared/viewed.

Some stipulations may include what type of content they can be featured in (for example, restricting the use of their name, image, and likeness in any sexual scenes).

What about non-celebrity content that goes viral? Well, I suppose depending on the amount of customization, that individual may receive some kind of credit, like director or producer, and receive a small payout. However, it may be more difficult considering that unless this person is an actor themselves, there will be much less material available to train the AI program on in order for their AI twin to accurately mimic their acting style and personal quirks...

[trendswerespotted]

Thesis: Synthetic media will usher in a new age of mass-customization

Apparently, my best ideas strike at the last minute. Inspired by some of my previous posts about Gen AI and its impact on film, I'll be exploring more about this topic for my final essay. With the help of CGI, our current technology has allowed us to digitally render several actors (such as Carrie Fisher and Paul Walker) to reprise their roles posthumously. Within the next 5-10 years, I think that technology will become advanced enough to produce generative AI versions of real-life actors, virtual actors, and even viewers themselves. Combining visual and audio AI technology, viewers will be able to cast their favorite actors in any show or movie they want. Wanna see a young Joseph Gordon Levitt as Marty McFly? Or Amanda Seyfried as Glinda (sorry, Ariana)? Or do you want to play Harry or Sally (...or both?) in When Harry met Sally? I think all of this will be possible within the next decade. Gen AI will take fan-castings a step further, allowing viewers to customize their media content to feature exactly who they want to see, making every movie/tv show a satisfying viewing experience. In doing so, mass-customization will be brought to on-screen experiences (almost like turning fan-fiction into something visual).

My next few posts will delve deeper into the current CGI/Gen AI capabilities, case studies of previous Gen AI use in film and tv, monetizing these services, ethical implications, and more. I think at the heart of this technological advancement lies the question: just because we can, should we?

Feel free to comment who your AI fan cast would be for any show or movie!

[trendswerespotted]

Branded Content in Film

Our previous guest lecture with Adam Pulchalsky covered branded entertainment content and various brands he has collaborated with! This was one of my favorite guest lectures I've seen during my time here at Syracuse. Adam believes that branded content will be the future of advertising/entertainment, and I agree.

Of course, last summer's blockbuster Barbie remains the queen of branded content. I think we will see the film industry move away from prequels, sequels, and sequels-to-sequels, and instead invest in developing stories around existing IP. We've seen this with video-game based movies/shows such as The Last of Us, The Mario Movie, and Sonic the Hedgehog. Now, we're seeing it with toy-based movies (Lego, Barbie). According to this statement from Mattel, the toy company plans to add to their movie universe by developing films based on their pre-existing IP such as Hot Wheels, American Girl Doll, Polly Pocket, Magic 8 Ball, UNO, Rock 'Em Sock 'Em Robots, and many more. In an age where nostalgia sells, this might be the perfect way to bring viewers back to "the good old days".

This is a very obvious form of branded content that tends to be in conversation with originality in film - just because a film is based of off existing IP, is it inherently less original? Not exactly, I think there are still plenty of creative directions to explore (depending on how married the director/producers are to the existing IP's lore).

Although I'm excited to see many of these upcoming Mattel films, I'm more interested in subtle forms of branded content and its ethical implications as brands begin to influence the art we consume. Sure, product placement has been a common practice in the industry for a while, but branded content is beginning to blur the lines between advertising and art even more. What if the movie I just watched was a 3 hour life insurance ad? Does it matter if it was entertaining? I think its more important to influence corporations now have over storylines, character development, and even entire franchises, which can potentially compromise creative integrity and artistic vision. Personally, I don't like the idea of brands having even more of an influence on our lives than they already do, but perhaps it's inevitable at this point...

[trendswerespotted]

Pendulum Swings in TV/Film

The idea of the pendulum swing is that society will favor one end of a spectrum then gradually shift to favor the opposite end, like a pendulum swinging back and forth. I alluded to this concept briefly in my previous post, but here I'll look at more specific examples and predictions of pendulum swings in the Film/TV industry.

Bundling and Unbundling

For years, I've heard that "streaming is the new cable." We went from channel bundles on cable with commercials to individual, ad-free streaming service subscriptions to (streaming) platform bundles (usually with ads). Despite its unpopularity with consumers, the tiered subscription model works for many streaming services - people will still pay for whichever plan suits them, even if that means paying way more for an ad-free experience. I'm not sure what would instigate the swing back to unbundling... so I asked Perplexity. Unsurprisingly, it mentioned antitrust concerns and consumer protection laws to promote competition and transparency regarding prices. I liked Perplexity's suggestion of improved recommendation systems (via AI/machine learning technology) as an impetus for unbundling. It suggests that improved technology can recommend better and more personalized content for viewers, encouraging them to unbundle (instead of receiving a diverse lineup of recommendations).

2. No screens to screens

Although we discussed in class that "no screens" originally meant paper, perhaps the pendulum will swing back towards no screens via different alternatives. I think Google's Project Astra is the middle of the pendulum swing right now - the glasses with AR technology are non-traditional screens. Though this sounds silly, maybe at some point in the future we'll arrive at holographic technology that can be displayed and interacted with (just like in the movies). We'll have the ability to produce these 3D holograms anywhere without the limitation of size or space. Unlike screens, these can't be scratched, cracked, or damaged by liquids and other materials.

*The image above is AI platform DreamDeep's creation of the "perfect James Bond"

3. "Authentic" content to generated/"fake" content

I talked a bit about this in my last post when I mentioned the rise AI generated content in film. Because this technology still feels "novel," and we're still pushing its boundaries of creative output, we're still at the "gen AI" end of the spectrum. I think in the future we'll see celebrities license out their own images and voices for simultaneous project creation (without having to physically be on-location) and even posthumous acting careers. With gen AI, celebrities could literally live forever and be immortalized through technology. However, the proliferation of "fake" or "inauthentic" AI generated content would ultimately cause the pendulum to swing back towards "real" and "authentic" content. I can already hear people saying "Brad Pitt actually acted in this movie, not the AI version of him" or "This isn't AI or special effects, they actually filmed this at a real place!" Perhaps we'll be paying actors even more for "real" appearances, instead of AI generated ones.

[trendswerespotted]

Digitalization - the new Golden Age of filmmaking?

Last week's MMI reading was The Second Machine Age by Andrew McAfee and Erik Brynjolfsson in which they highlight 3 characteristics of digital progress and the "Second Machine Age": exponential, digital, & combinatorial. In this post, I'll be focusing on digitalization and some of its current/potential impacts on the film industry.

High Quality Digital Cameras

Today, high quality digital cameras are more accessible and more powerful than ever before. Albeit being expensive, many marketplace cameras have advanced features like 4K/8K recording, sophisticated autofocus, IBIS (in-body image stabilization), and 10-bit recording. Just a few years ago, these features would have been exclusive to professional/industry level equipment. Nowadays, even our phones offer high quality, portable cameras that capture images and videos as well as traditional digital cameras. Such high quality equipment contributed to the "democratization" of the film industry by lowering such a crucial barrier to entry. In 2024, anyone can be a filmmaker. As an added bonus, better in-camera capabilities reduces the time and cost necessary to fix things in post, maximizing efficiency.

2. AI/Gen AI

Although I don't think screenwriters will lose their jobs to generative AI anytime soon, I'm not sure if I could say the same for cinematographers or special effects editors. As gen AI platforms continue to develop and humans get better at prompting, I think we can expect to see more and more AI generated content in films - especially for complex scenes that require risky stunts or unique/fictional locations. Perhaps we could even see actors license their image and voices to AI platforms so that they can star in several projects at once, without having to physically be in that filming location. That's more of a "Black Mirror" scenario. Realistically, I see generative AI initially being used to cut costs of filming complicated scenes. I also think it will be used as a back-up to prevent delays in production in case the lead actor/actress is sick or injured or if the weather is unideal. There are already some partial or fully AI generated films, but it's just a matter of time until these types of movies become the next blockbusters. The pendulum will swing towards "fake" AI generated content, and then a few years later it will swing back to "authentic" content with "real actors" filmed in "real locations."

3. New business models

Of course, streaming has introduced new methods of distribution and exhibition, with many films often skipping theatrical releases and going direct-to-streaming instead. Digital distribution also allows for a greater global reach, reaching new and diverse audiences. With such convenient and cheap digital accessibility, we're also seeing the death of DVD's/physical copies of media. Sometimes, I imagine that if Netflix or Hulu lost their entire original catalogue in one day, so much art would be lost (like a modern day burning of the Library of Alexandria). Maybe we should protect and preserve piracy sites... as back-ups, of course.

[trendswerespotted]

Supertrends in Film: Increasing deculturalization

When it comes to cultural representation in film and writing stories about a specific culture, you can't play the "global audience game." By that, I mean that these cultural stories can't be written with everyone in mind (i.e. give in to the pressures of deculturalization and globalization for the sake of profit), otherwise they lose their authenticity and the quirks that make these cultures unique. When you try to represent everyone, you represent no one.

The effects of deculturalization significantly decrease the impact of these films. One example that stands out to me is Disney's Raya and the Last Dragon. Before its release, it was rumored to be a story with filipino representation. However, I didn't feel represented at all. This story was an an amalgamation of several southeast asian cultures, combined into one movie and served on a silver platter with Disney saying "Here's the representation you asked for. You're welcome." This is deculturalization in action.

We also see it in TV episodes with "no sabo" characters (i.e. hispanic or latino characters who don't speak Spanish). This is often seen when actors have poor pronunciation or when characters use bad grammar (due to non-spanish speaking writers). The hispanic and latino community often poke fun at bad spanish in film and tv, especially in situations where a character uses a random spanish word (particularly one that english speakers can understand, like abuela) to remind audiences that they are hispanic/latino while still being understood by english-speaking audiences. Why don't they just say a whole sentence in spanish and add english subtitles? Or better yet, hire spanish speakers in the writers' room to maintain cultural authenticity and not white-wash these characters?

A personal pet peeve of mine is when characters are speaking a different language and the captions read [speaking in foreign language] or [speaking in Spanish]. I think this reinforces deculturalization and establishes the idea that the audience doesn't even need to know what these non-english speakers are saying; it's not in english, so it doesn't matter.

A film that I think combats deculturalization really well is Everything Everywhere All at Once. The characters speak Mandarin and Cantonese (intentionally switching between dialects) and there's a true understanding of the unique dynamics within an asian immigrant family. Even the set decoration feels like stepping into a real Chinese family's apartment.

It's easy to tell when filmmakers are using diversity as a money grab. Simply featuring a token minority who says one word in their native tongue doesn't count as "representation." To combat deculturalization, it's important for filmmakers and storytellers to write stories for a specific culture, instead of simply writing about them. This can be remedied by emphasizing diversity off-screen, not just on-screen. By writing for a specific cultural audience, we will naturally tell better stories. If we tell authentic cultural stories that resonate with these groups, the rest of the world will want to listen.