On Tuesday, OpenAI dropped Sora 2, the newest iteration of its 2024-launched video-generation tool. Figure skaters with cats on their heads, dog astronauts gobbling tennis balls, unusually agile horseback riders standing astride multiple animals — these were some of the snippets seen in its short video presentation.

Sure, the model still struggles with dialogue — notice that the scenes feature mostly voiceover when there are words at all. And slow down the videos to individual frames and you’ll see plenty of hallucinations. But OpenAI did enough shiny things Tuesday to make users take notice. The company will release — first by invite, and then presumably commercially — a tool that will let users prompt these videos into existence and share on their social feeds. Now instead of dunking on people with off the rack memes or shooting a quick video to troll our friends we’ll just prompt them into existence.

“Sora 2 can do things that are exceptionally difficult — and in some instances outright impossible — for prior video generation models: Olympic gymnastics routines, backflips on a paddleboard that accurately model the dynamics of buoyancy and rigidity, and triple axels while a cat holds on for dear life, “the company said in a statement.

So what will this mean for our content landscape? For Hollywood producers? For all of us entertainment consumers? The news also comes amid reports that the company will launch a Sora-powered social app and that Hollywood studios will need to opt out — read: request takedowns — of any video it finds distastefully close to its own protected IP.  Taken together,  these news developments seem to be hinting at a world of video we’ve been waiting on for a while now. The AI-ification of our TikTok feed is upon us.

The Hollywood Reporter’s senior tech editor Steven Zeitchik and senior features editor Julian Sancton sat down to make sense of what they saw.

**

Julian Sancton: So how do you read today’s announcement? Does OpenAI, which has been having not-so-open talks with Hollywood about encouraging the use of their tools in traditional filmmaking, now give up that approach in favor of allowing the public to make 10-second clips? Or is this just a way to make some cash until the technology to make longer clips becomes viable?

Steven Zeitchik: The $64,000 dollar question. Or I guess $500 billion question, given OpenAI’s valuation: Are they trying to be AI TikTok or do they still have hope of being next-gen Pixar? Of course, maybe we’re being too old-line in our thinking.  Maybe there won’t be a next-gen Pixar — maybe that top down, elite producers-mass consumers binary, if OpenAI has its way, is a model of the past. Maybe what they’re really after is a world where studios are just sort of IP gatekeepers and we’re all creating within it.

JS: It certainly seems like being AI TikTok is cheaper and more profitable! At least for OpenAI. And it’s a lot more attainable now.

SZ: And so where does all this leave high-end Hollywood-type video? There are only so many hours in the day, and if these tools catch on we’ll be spending a lot more of them watching or creating our own slop-y videos instead of the professional kind. Studios and even influencers will feel the squeeze.

JS: I would be very surprised if Hollywood allows this “opt-out” model to stand unchallenged. The studios, as you’ve noted before, would much prefer to control the memeification of their IP.

SZ: Right, doesn’t this all work its way toward a deal? Like right now studios are fighting and OpenAI argues fair use and it’s very line-in-the-sand. But ultimately it seems to me it’s all about setting terms for negotiation. Sora 2 or Google Veo or whatever tool we use to create these videos will come with branded characters, and studios will get a cut of that. They’ll move more from a production model to a rights-management model. Which has kind of been happening anyway.

JS: But do you think studios will really go for that? I do doubt that studios would be happy making movies without having to deal with actors and writers and crews.

SZ: You think they’d prefer to deal with actors and crews?

JS: I think for all their cutthroat tactics and greed, the execs did not necessarily get in this business just to make money. Scratch most of them and you’ll find a failed artist.

SZ: The Matt Remick factor. I guess it comes down to how many of them there are — the quiet idealists in the C-suite. You think there are many. You’re much more of an optimist than I am.

JS: Let’s not forget the studios are also still smarting from the strikes. So they don’t want to openly pursue these meme business models.

SZ: Very smart point: the labor optics aren’t great if they’re making deals with these companies.

JS: And the more AI is used for memes the more solid the unions’ leverage gets. The guild leaders can say “Is that the business you want to be in? Or do you want to be in business with us?

SZ: That’s true. But the boards have a say here too. Pixar — not to pick on them — but Pixar takes on many billions of dollars of risk and many thousands of employees of overhead. OpenAI and SV will essentially be saying “Here’s a much shinier proposition we have for you, Hollywood — give us your stuff and we’ll make you Pixar money without having to build the thing or take flyers on 10-year development cycles.”

JS: I should say I’m a little skeptical of what these tools can do. It’s still just prompting a model and seeing what you get, like a slot machine. That’s what they all are. And there’s also the consumer-appeal factor. ChatGPT’s great innovation was the simple user interface which I haven’t seen here yet.

SZ: That’s true. But I think we both agree that today’s drop does make it easier for us to imagine a world, collectively and individually, where you don’t have to shoot a video to post it. Like it will seem like cheating for a second and of course also frustrating because of all the hallucinations and flying unicorns but it does seem like the road is getting dramatically shorter from where we’ve been to a place where we just speak/type a video and then share it

JS: It’s notable that OpenAI is launching a wide ad campaign for Sora 2 and the app, including print ads. They want this to be the standard. It will make it easier for people to try their hand at it. What takes a little savvy now — the horribly offensive Chuck Schumer deep fake that Donald Trump re-Truthed — will be at anyone’s reach in no time.

SZ: And then toss in the IP factor, assuming those deals are made and legal issues are worked out. I can troll you for not going to that party but instead of just words I can have Luke Skywalker do it. Fun for you, for me, and fun for Disney’s coffers!

JS: It’s just so depressing to see how quickly these AI companies pivoted from this lofty talk of AI’s potential to change the world for the better to just pouring gasoline on the social media dumpster fire.

SZ: You were hoping for AGI.

JS: I’m just amazed by how tech history is repeating itself. It took about 10 years for technoutopianism to curdle into dangerous, brainrotting enshittification with social media. AI has achieved it in less than 3. I guess it does speed up progress! 

SZ: See you’re bleak too. You just garb it in a cloak of optimism. But you’re right of course. Sam Altman and OpenAI have been touting different forms of Artificial General Intelligence — a machine than can reason like a human – for years now. And we just keep getting souped up iterations of ChatGPT. It’s like: “you won’t have an AI Agent that can help you reason through the biggest decisions of your life with utmost clarity but at least you can speak a flying on-screen panda into existence.”

JS: This all reminds me of something I write about in my (shameless plug) upcoming book, which is how treasure hunters are less adept at finding treasure than at keeping investors on the hook with big, outrageous promises. Like most shameless of them all, Robert Marx, who said that he was close to finding the greatest Spanish treasure ever lost, a life-sized solid gold statue of the Madonna and child. He never provided evidence that it was even made.

SZ: So you’re saying Sam Altman is the Robert Marx of Silicon Valley.

JS: Yes.

SZ: I guess it’s better than being the Richard Marx of Silicon Valley. Last thing: How much art do you think survives here? Like, big picture.

JS: I’m an optimist on this. PTA is going to PTA.

SZ: That’s true. If you’re him. But with so much less room for risk, where is the next PTA coming from? That’s what worries me. The automation of production — as opposed to the digitization of distribution, which is what we had in the streaming era — feels like it will lead to so, so much more slop. Obviously there will be exceptions.

JS: So PTA and Slop and little in-between?

SZ: It would be a fitting mirror to the death of the middle class in our broader economy.

JS: Quality inequality.

SZ: At least there’s a phrase AI could never come up with.