The "AI-Generated Hit Movie" Horror Story
Jennifer Maas
As the CEO of a major streaming player, you'd think Anthony Wood would rather not get Hollywood all riled up. Maybe he's taking a page out of Ted Sarandos' former playbook?
“I have no idea if there’s an investment bubble, but I know that AI is going to be huge, and it already is,” Wood said during a headliner conversation with Variety‘s co-editor-in-chief Cynthia Littleton at the Variety Entertainment Summit at CES Wednesday. “So it’s going to affect lots of different industries, but in our industry, it’s going to lower the cost of content production. I predict within the next three years, we’ll see the first 100% AI-generated hit movie, for example.”
The first part is obviously a reasonable position. Most people think it is an investment bubble, but also that it's perhaps a needed one given the prospects of the technology, which is where I generally fall (outside of a handful of outlandish deals). But Wood is not in the business of investment advice or thoughts. Again, he is in the entertainment business and so not only talking about the prospects of "100% AI-generated" movies, but predicting that they'll be here within three years, producing hits, is, um, controversial.
I also tend to think it's silly. Will we be able to produce 100% AI-generated content in three years? For sure. But we can already do that now. It's what is flooding social network feeds as we speak. What started as AI-generated images has morphed into video. And what started as short clips is naturally morphing into longer ones. But really all you need to do in order to make longer ones is simply stitch shorter ones together, obviously. And so the AI-generated trailers that have been flooding our feeds ranging from fake Avengers movies to fake Zelda movies are just one step removed from full movies. Well, maybe a few steps. But really, it's just time. Anyone can do that right now.
The bigger question is if there will be an audience for that content. And that's the controversial part of Wood's comment. It's not the creation – again, that will happen – but the notion that it would lead to a "hit movie". I have a hard time believing that will happen for a range of reasons, including the current powers that be.
I suppose it depends on how you define "hit". But I take it to mean a movie that millions watch via "traditional" methods, be it in theaters or now on streaming services (or on "regular" television, I suppose). Three years is a long time in AI, but in the grand scheme of entertainment, it's short. Movies are just being greenlit right now that will come out in three years. The mechanisms are already established to distribute those movies. So where is this AI hit movie going to slot in? Certainly not in a movie theater!
Okay, so what about a streaming service? Maybe. But again, the most obvious such service to create hits, Netflix, is in the midst of trying to buy a movie studio and as such, Ted Sarandos is now busy backtracking from his aforementioned previous comments. Certainly Netflix, like Roku, will use AI to enable talent (and yes, drive down costs), but I have a hard time believing they'll go forward with a full AI-generated movie in that timeframe. Also, just imagine the lawsuits?
Other services? Maybe! But it certainly won't be Disney. I don't think it will be Amazon. HBO will either be owned by Netflix or Paramount, and again, it won't be them. So it could be a smaller player, but will they be able to make a movie a true "hit"? Could Roku?
The wildcard would be YouTube. Because of the UGC underpinnings of the service, you'll certainly see movie-length AI-generated content uploaded. Especially from those outside the current systems – that's a good thing! But will any of it actually "hit"? Maybe as a novelty some will get a ton of "views", but it's hard to imagine it would be considered a true "hit" movie. More just like a proof-of-concept type thing, perhaps.
People will point to music and note that there are some AI-generated "hits" popping up. Two things there. First, I still suspect there is a novelty aspect to this. Second, a three-minute song is a different beast than a two-hour movie. Many people are willing to give up three-minutes of their time. Two-hours is a commitment. One other element that our AI-enabled future is going to highlight: time is the ultimate premium.
Now, longer than 3 years for that AI-generated hit movie, all bets are off. But that window Wood gives seems too short. Not because of AI, but because of the way distribution currently works.
I would also just say that I continue to bet that while such AI-generated content is coming, I think once that novelty factor wears off, you'll be hard-pressed to find big hits there in the way you do with human-created movies and television. And, counterintuitively, I think as AI-generated content continues to flood every surface, it will raise the demand (and thus, value) for human-generated content.
I do think there will be "hybrid" content – part human-created, part AI-created – but I think people will prefer to watch work that was created by other people, because they will value the time and care spent on it. We'll slowly realize it's just as much about the input, as the output.
To be fair, Wood does hit on some of this in the same chat:
“I think people underestimate how dramatic that’s going to be,” Wood said. “I mean, obviously I don’t think people are going to get replaced. Humans are still the creative force behind creating content and hit shows, but the cost is going to come down dramatically, and that’s going to change a lot of companies’ business models. So I’m focused on, how do we take advantage of that? That’s a big opportunity for us.”
Again, that's reasonable. The AI-generated movie topping the box office in 2029? Less so.
One more thing: I continue to be intrigued by Roku's "Howdy" play:
For Howdy, Wood describes the $3-per-month Howdy offering as “not designed to replace a major streaming service like Netflix or Disney, it’s designed to be an add-on service.”
“The opportunity for Howdy was, if you just look at what’s going on in the streaming world with streaming services, they’re getting more and more expensive,” Wood said. “They keep raising prices, and they keep adding larger and larger ad loads. So the part of the market where it actually started, low cost and no ads, is gone now. There’s no streaming services that addressed that portion of the market. That’s the opportunity for Howdy. It’s three bucks a month and no ads and it’s doing extremely well. Just like we built the Roku Channel using the promotional power of our platform, that’s what we’re doing with Howdy. We’re using that to grow it. But Howdy has very broad appeal. There’s lots of people in the world that want a $3-a-month streaming service with no ads. So we’ll start on Roku, but we’ll also take it off platform as well. I think it’s going to be a really large business.”
I've been critical of Roku with regard to ads in the past, but this makes a lot of sense to me. And I suspect we'll see others start to copy this model too.
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler