$200 Million Movie in a Day? Welcome to the End of Hollywood

Prepare to be blown away by the Hollywood-quality video AI now generates, and if you work in or even near Hollywood, prepare for a vicious case of the night-sweats.

The Dor Brothers, who bill themselves as "pioneers in AI video production," earlier this week claimed they "just made a $200,000,000 AI movie in just one day."

Well, no.

But they did release a three-minute trailer for what looks like it could be a $200 million Hollywood production. In many ways, that isn't a complement. But in the ways that matter to the bloated movie studios, what the Dor Brothers have done might just represent the future of filmmaking, for better or worse.

Before we get into any of that, please take three minutes to watch the "100% AI" trailer for Apex.

There are plenty of nits to pick — palm trees in NYC, really? — but overall, the special effects* are probably at least as good as anything from whatever the hell the most recent Marvel superhero movie was.

The asterisk after "special effects" is because there are no special effects. There aren't any actors, either. If you still watch SFX spectaculars like the big studios spew out several times each year, take a look at the credits and you'll see a massive list of computer animators responsible for all the CGI. But for Apex, there was no big team of well-paid CGI artists. There were only prompts fed into an LLM server farm, and a big team of Nvidia graphics cards doing the work for "free."

And while the "actress" is no Oscar contender, she's probably good enough for Netflix "second screen" streaming slop. And even with all those pricey Nvidia cards behind her, the AI heroine is a lot less expensive than hiring Zendaya for the same role in a "real" movie. Probably more expressive, too.

We're barely into 2026, and the state-of-the-art (or perhaps only nearly so) in AI video generation might have have been Runway Gen-2 or one of its competitors. Compare and contrast what Runway could do then with what the Dor Brothers did on Monday.

In 2024, AI-generated video struggled not to suck, and failed at clearing even that low bar. In February, 2026, we're complaining that those real-looking trees in the fast-moving action clip don't belong in New York City.

"You've come a long way, baby," the cigarette ads used to boast. "And in such a short time, too," I'd add.

So if we've gone from "not even real" to "we're picking at nits" in two years, does that mean we're just another year or two away from reaching the Singularity — when AI gets smart enough to reprogram itself faster than we can keep up. And then politely asks if we’d like fries with our obsolescence.

That's where things get complicated.

Way back in 1993, mathematician and hard sci-fi author Vernor Vinge said he'd be surprised if the Singularity occurred before 2005, or later than 2030. In his 2006 novel, Rainbows End, he narrowed that down to 2025. While there's still no superhuman intelligence, as Vinge described the Singularity, does a video like Apex provide evidence it's nearly here — and very close to Vinge's timeframe?

There's another school of thought — human-limited thought, granted (heh) — that while AI now excels at things like generating convincing video from nothing but a few text prompts, deeper, real thinking just isn't possible for machines.

Xaira Therapeutics AI biomedical chief, Bo Wang wrote this week that "Professor Judea Pearl — the pioneer who invented causal reasoning in AI — says scaling won't save us." Or as Pearl put it himself, "Mathematical limitations that are not crossable by scaling up."

"LLMs aren’t learning how the world works," Wang added. "They are learning how we describe the world."

This resonates with most biologists: Drug discovery is hitting the same wall. We have mountains of genomic data, but most AI models just find patterns in published papers — not in the raw biology itself. They're learning what scientists think causes disease, not what actually does.

Instead of Garbage In, Garbage Out, scientists are getting Good Stuff In, Same Stuff Out.

There's nothing creative happening in the models. Which, frankly, is probably fine for producing $200 million brainless action movies for Netflix — but sucks if you were hoping for a beyond-human-intelligence to find all the cures for cancer on some random Tuesday in 2028.

We were hoping for miracle drugs. We feared getting Skynet. But maybe putting 90% of Hollywood out of work — and untold numbers of white-collar workers in every other industry — is all we'll get. At least for a good while longer than Vinge thought..

Maybe, as Jack Nicholson's character cruelly asked a waiting room full of depressed therapy patients in the famous movie, "What if this is as good as it gets?"

I don't pretend to know, and neither does the LLM I consulted for this piece.

Enjoying PJ Media?

Get exclusive content and support independent journalism with 60% off a PJ Media VIP membership. Use promo code FIGHT and join today.