A crowd watched panel speakers on stage at SXSW Photo Credit- Hillary Lewis

The Current State of the Industry and the Future Impact of AI

The Entertainment Industry: How did we get here and what is the future impact of AI?

A pandemic followed by a recession and twin strikes, add the sudden rise of AI, the over-production of original content, changing viewing patterns, potential bundling of streaming platforms – the list goes on and on. The entertainment industry is in a state of flux that has not been seen in decades. When the smoke clears, where does that leave us? What really is the future of the entertainment industry? How are networks and studios adopting AI, and what does that mean for the rest of us?

CreativeCOW’s Contributing Writer, Hillary Lewis ventured to SXSW 2024 in Austin, TX, one of the country’s largest conferences in tech and film, to get a pulse on these questions. 

Some of the biggest industry leaders were in attendance, including SAG-AFTRA National Executive Director, Duncan Crabtree-Ireland; CTO of Paramount Global, Phil Wiser; CEO and founder of Roku, Anthony Wood; VP Product of Adobe Express, Ian Wang; Vimeo SVP of Product, Zohar Dayan; and notable writers, directors, actors, media analysts, AI executives, and other leaders in the industry.

First up: What is the current state of the industry?

The biggest question on everyone’s mind is…How did we get here? Why is our industry still in such a disarray? The Hollywood strikes ended six months ago, so why does content production remain at a standstill? Will the industry ever recover? 

There were an overwhelming number of panels at SXSW covering this very topic. One specific panel, ‘The Death of Peak TV – A 2024 Post Mortem on Production Downturn’ led by Tyler Aquilina, Media Analyst for Variety’s Intelligence Platform, shed some light on just how we ended up here. 

According to a mountain of data, Aquilina says the industry had been approaching a tipping point long before the Hollywood strikes began. “So the first thing that we should really take note of here is that 2022 was the year that TV finally actually peaked. The term ‘Peak TV’ was coined well before TV output finally crested and then fell.”

‘The Death of Peak TV’ presentation shows the gradual shift from cable and broadcast to SVOD, i.e. streaming
‘The Death of Peak TV’ presentation shows the original series output each year by category of content.

As shown above, in 2023 the industry as a whole saw a 20% drop in original series output across all platforms, plunging from more than 2,000 titles in 2022 to just over 1,600 last year. That came after two decades of almost continuous growth. 

There were, of course, other years with small dips in content – one for the writer’s strike in 2007-08, and one for COVID. However, those dips were never more than 7% – a sharp contrast to last year’s 20% drop. We also saw a spike in production immediately proceeding both of these time periods – another contrast to the stagnant production levels we’ve seen in 2023-24. 

The year 2020 also happened to be when the streaming wars really got going. Disney+, HBO Max, Peacock, Discovery  –  all of these new platforms started popping up beginning in late 2019. And they all decided to follow the model Netflix built, aiming to entice more and more viewers away from cable to streaming by producing massive amounts of original content. More content than anyone could ever dream to watch.

So in 2021 and 2022, we see an unprecedented influx in TV outputs with these new streaming companies throwing billions of dollars into production and racking up massive deficits in the process. The problem is most of these companies weren’t turning a profit. Their main goal, and the only thing Wall Street cared about at the time, was increasing their subscribers. 

That is, until something very important happens in 2022…Netflix loses subscribers in Q1 of that year, the first time they’d ever reported a drop in subscribers. This is what starts a chain reaction and triggers the beginning of the fall of Peak TV. Investors begin to pay a lot more attention to how much money these legacy media companies are spending AND how much they’re losing. 

So while there are many events in 2023, like the twin Hollywood strikes, that bare the brunt of the blame for the current state of the industry, the true death of Peak TV was happening before most of the industry was even aware. 

How are networks and studios responding to these changes?

So what happens now? Has corporate greed ultimately killed the industry? Or are other factors also at play? How will streaming change in the coming years? The guesses are endless but the consensus seems rather bleak for the entertainment industry for some time. 

First, networks and streaming companies are taking a step back from over-producing in all areas, only prioritizing their most successful content. More careful considerations for IP-based blockbusters (think The Last of Us and the newly greenlit Monopoly film by LuckyChap and Lionsgate), prestigious awards-based titles (think Oppenheimer or Dune: Part 2), and a strong focus on bringing more sports to streaming (Disney is refocusing a big part of their budget on this).

The projected global spending in 2024 of Sports programming; Disney is set to invest the most in sports out of all other streaming companies

Aquilina also suggests streaming companies will favor unscripted content in general over scripted moving forward, and we’ll see more ‘reruns’ returning to platforms. 

“Streaming has typically been filled more by scripted than unscripted. But that is going to start changing. Streaming will look a lot more like cable, fewer scripted originals being produced. Unscripted is a lot cheaper. In tandem with that, I think streaming will start relying a lot more on ‘reruns’… which is to say, old shows or previously produced content that people watch over and over again. And let’s remember, what was the biggest success story of last year? Suits, right? This means more and more companies will start licensing their shows to Netflix.”

We’ll also see a surge in International TV, or non-English language TV. Successful shows like Squid Game and Alice in Borderland did incredibly well globally and started claiming their share of SVOD releases in the U.S. Unfortunately for the U.S. media entertainment workforce, these international shows can be produced much cheaper and aren’t governed under the same labor contracts.

‘The Death of Peak TV’ presentation shows consistent non-english language share of series until a sharp 9% increase in 2023.

Overall we’ll see shows with a broader appeal to all markets instead of specific markets, shows with less extravagant budgets, shows with longer seasons, and shows with episodes released weekly or in halves. 

This is all partly due to the WGA and SAG-AFTRA deals that were recently ratified – with longer seasons and more episodes per season, writers and actors will have a better chance at staying employed for longer. And partly due to streaming company’s attempts to retain subscribers while implementing more ads on their platforms. The days of binge watching a new release will be a thing of the past.

We’ll continue to see a rise in Fast Channels. What’s a fast channel, you ask? If you haven’t heard of Fast Channels, you’re not alone, but there are many currently available, such as Amazon’s Freevee, The Roku Channel, Tubi, Sling TV, and more. Fast Channels are linear TV channels but on your streaming device, essentially the cable format coming to streaming. Why is this suddenly popular? In a SXSW panel with Roku’s CEO, Anthony Wood says, 

“It turns out people don’t want to decide what to watch, they just want to flip on a channel and let it play. Roku is one of the biggest fast channel distributors in the industry.” 

Beyond all of these shifts, all eyes are currently on the IATSE and AMPTP negotiations, which began in March. The thought of another major strike in Hollywood is a huge concern for industry workers and networks/studios alike. So far reports have been promising that tentative deals are being reached, but we won’t know the outcome until the end of July. 

And then, of course, we have the looming dread of AI. 

How will AI be used in the next 5 years?

We’re all scared. And rightly so. As of the end of March, most of the below-the-line workers in TV & Film industry (both scripted and unscripted) remain unemployed. The networks and studios have played their streaming wars and left carnage in their wake, not willing to bring back their workforce…let alone teach them how to use AI. There’s an industry full of long running careers that feel utterly dead in the water. 

And now on top of all that, the knee jerk reaction to AI only further perpetuates the question, “How much longer do we wait for the industry to return to normal? And will AI have already significantly impacted the jobs that were taken from us? 

There’s a palpable positivity at SXSW around AI in every industry EXCEPT Film & TV, where panelists and leaders are still trying to convince us AI will make our jobs better, not worse. Whether we like it or not, AI is quickly being adopted at industry-leading media  companies. Variety’s Intelligence Platform has several in-depth research papers on this topic, but their most recent study offers a 24-page analysis on AI adoption in Film & TV with a focus on how companies are already adopting near-term and future uses for AI.

A full list of current adopted uses of Gen AI by major media companies; Page 9 of Variety’s Intelligence Platform “Generative AI in Film & TV”.

In a panel with Paramount CTO Phil Wiser, he listed several similar Gen AI use cases below in which Paramount is already utilizing in various stages of production. 

  1. Pre-Production: Script Breakdowns for script coverage & comps, Scene Summarization, Character & Structure Analysis. Not pushing so much to post-production and VFX and more pre-visualization up front.
  2. Production: On-Location Performance Capture, Set Scanning, Metadata Gathering. Utilizing the Holodeck in more set productions.
  3. Post-Production:
    1. Localization – Subtitling & Captioning, Dubbing & Lip Sync, Cloned & Synthetic Voices 
    2. Video Generation & Editing – Runaway AI / Pika Labs / Sora related Video Generations, Live Event Facial Replacements & De-Aging
  4. Marketing & Distribution: Compliance Editing, Regulatory Analysis, Nudity & Profanity Detection, Audio Descriptions

The big use cases above are relatively still in their infancy and very costly to create, such as video generation (Sora and other models) and deep fake models being used for live event facial replacements & de-aging. But others are maturing at a much faster rate, such as dubbing & lip sync and cloned & synthetic voices, where you can replicate anyone’s voice and dub in multiple languages in seconds, retaining the same tone and expressiveness.

Top creative tasks polled from entertainment workers show AI-generated sound effects, auto completing code, AI-developed artwork, and AI dubbing to be at the top of the list; Page 10 of Variety’s Intelligence Platform “Generative AI in Film & TV”.

Beyond these production uses for Gen AI, networks and studios are focusing AI efforts on improving business operations and targeted demographics. This includes 1) using AI recommendation machines for targeted advertising on streaming platforms,  2) understanding audiences better by utilizing AI-trained panels in different demographics that watch content at scale and give feedback, and 3) correctly labeling and organizing assets and data in a valuable way to prepare for future AI models. 

Paramount has also been working with Adobe Firefly teams for proprietary research and development in small, specific open source models. These models are trained to learn their IP, i.e. franchises like Spongebob Squarepants, attempting to generate new content within those fan-favorite shows.

Paramount’s CTO Reimagines the Modern Media’ presentation shows how prototypes are broken down into elements to feed into their proprietary franchise model.

Phil Wiser says these models still need a lot of ‘babying’, but are promising tools for creators to generate content more efficiently.

“First we hired a team to label all the IP appropriately and then trained an AI model to auto-label. So now we’ve got a model built up, but our teams needed to get better at coaxing that model to generate algorithms to do it. It’s a real learning process. And if you play around with image models, it can be frustrating, because you’ll enter a text prompt, you’ll get something that’s pretty close, you tweak it a little bit, and then it goes off.”

As large companies push forward with these creative uses, there’s an ever-present concern of responsibility around Gen AI and the moral obligation it presents. 

When the internet was first invented, people were just as fearful about its potential. But imagine telling someone today, ‘You can make this movie, but you can’t use the internet’ – it would sound completely unreasonable. We’re at that same tipping point with AI. We have to recognize that AI is here to stay, and that safeguards and regulations are necessary, like any new tech.


A SXSW Guide to Earning Your Access Instead of Paying For It
Do you ever feel like entertainment industry conferences, festivals, or networking events …
Avengers: Endgame – Creating the Last Battleground
Avengers: Endgame hit theaters on April 26th, and went on to break …

Enjoying the news? Sign up for the Creative COW Newsletter!

Sign up for the Creative COW newsletter and get weekly updates on industry news, forum highlights, jobs, inspirational tutorials, tips, burning questions, and more! Receive bulletins from the largest, longest-running community dedicated to supporting professionals working in film, video, and audio.

Enter your email address, and your first and last name below!

Sign up:

* indicates required

Responses

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy