Ask A Genius 1447: AI Video, Urban Myths, and the Future of Simulated Reality
Author(s): Rick Rosner and Scott Douglas Jacobsen
Publication (Outlet/Website): Ask A Genius
Publication Date (yyyy/mm/dd): 2025/07/13
Rick Rosner and Scott Douglas Jacobsen discuss the rapid evolution of AI-generated video tools like MidJourney and VEO-3, exploring how probabilistic models simulate reality. Rosner critiques right-wing portrayals of cities as “shitholes,” defending urban vibrancy, diversity, and rising property values as signs of desirability—not decay—in cities like New York and London.
Scott Douglas Jacobsen: What are your thoughts on AI right now? I sent you the link to MidJourney’s collection of today’s best videos, right?
Rick Rosner: I don’t know if you watched it, but you understand the landscape. With just a few sentences, people can create incredible worlds using MidJourney’s video tools. The videos are about three seconds long.
I mentioned this on PodTV this morning. A guy who knows more than I do said, “That’s nothing—go to VoVo-3,” or maybe he meant Google’s AI video platform—I’m not sure of the name. They offer eight-second clips, and yes, they’re impressive.
All the major AI companies are now in the video space. Two years ago, they focused on still images, which were already impressive. Now it’s about video. These companies are preparing to launch real-time, explorable environments—generated worlds you can walk through freely.
You won’t get an entire planet’s surface. I imagine it will use some spatial wraparound—if you walk far enough in one direction, you loop back around. The initial worlds will likely be small, but I expect them to grow over time. Eventually, it will be like navigating Google Maps: use arrows to move around in real time.
Want to be in a ’90s disco full of supermodels? You can. Want a world full of capybaras? Done. Whatever you want.
At first, you’ll likely be limited to pre-made environments. But soon, you’ll be able to design your own. The surface area—or three-dimensional volume—of these worlds will expand as the tech improves.
That’s the first point.
Second, reality is turning out to be simulatable.
AI-generated worlds often resemble reality. Where they fall short, the companies fix the errors, or the AI learns. Remember when AI struggled with hands? You’d get weird fingers or extra digits. That issue has largely been resolved.
There are still flaws. Fast motion—like martial arts—can break realism. AI struggles with precise, high-speed combat sequences.
But in many other areas, it’s getting realistic. Hair movement, for instance, is surprisingly good, likely because the models are trained on large datasets of hair moving under various conditions.
When a character in an AI-generated video turns their head, the hair often moves realistically.
AI, of course, does not “know” anything in a conscious sense. It works through probability networks, filling in gaps based on patterns. If the hair is in one position and the character moves, the system infers how the hair should behave.
It is not applying actual physics, at least not formal physical equations. It uses a Bayesian probability map based on statistical likelihoods drawn from its training data.
That turns out to be good enough. With sufficient data, it can generate a convincing world, with some exceptions, but those can be fixed. So if you are making a 15-second ad, you can develop the whole clip with AI.
Then you can tweak specific issues to improve realism, or wait six months, regenerate the ad, and the AI will have advanced enough to handle those flaws automatically.
This is not so different from how our brains work. We do not consciously know the physics of hair, but we know how it should move because we have seen it move all our lives.
If you watch a video and the hair moves oddly, it feels off. David Lynch used that effect deliberately. In Twin Peaks and some of his other films, he reversed the footage to make scenes feel unsettling. A character might walk forward, but it was filmed backward.
One of the giveaways is the hair—it moves before the motion that should have caused it. The swing and snap are reversed. That violates our intuitive sense of motion, and we notice it instantly, even though we do not consciously calculate it.
AI works on the same principle. It simulates reality convincingly by pulling from massive datasets, just like our brains. Often, on a first viewing, nothing feels out of place.
That is mostly a good thing. It means we can create immersive, convincing worlds with AI. In terms of entertainment, it is excellent. People can make content that feels natural and believable.
But it is also troubling. People are losing jobs.
L.A. has a lot of local news stations—probably seven in English, four in Spanish. There are Vietnamese and Cantonese channels, too. So, many people work in local news.
But now, teleprompter jobs are disappearing. That used to be someone manually typing scripts into the system. Now, a news director dictates the story, and AI formats the entire segment. That is one lost job—maybe two. You used to need a day shift and a night shift operator.
Assistant directors, set managers—a ton of roles in entertainment are vanishing.
That said, people should still check out MidJourney and Google’s VEO-3, their video-generation platform. The video quality is incredible. None of the people are real. None of the voices is real. The accents are generated. The faces are entirely synthetic.
It is all AI, built on enormous training datasets. That is where we are. And it reflects how we, as humans, understand the world by constructing internal probabilistic models. We form these intuitively over time. The rules of perspective, gravity, and light reflection—we internalize them just by living in the world.
You do not need to study perspective formally to grasp it. You do not need a physics degree to understand gravity at a functional level. Our brains have absorbed enough examples to generalize the rules.
Some say New York is a shithole like London. Or that it is going to be a shithole because the likely next mayor is a Democratic Socialist, a Muslim, with some progressive policies. He wants to establish maybe five grocery stores across the city with price controls for low-income residents—something like that.
The MAGA crowd is quick to label cities as “shitholes.” But I have a few reasons why that is nonsense.
First, crime statistics.
When Carole and I lived in New York in the 1980s, it was rough. Still exciting—but crime was two, three, sometimes four times higher than it is now.
Today, New York is safe. It is expensive, but it is a good place to live.
Also, here’s the contradiction: Fox News constantly calls New York a “shithole,” but they have had their headquarters in New York City for 29 years, since they began.
They’ve never had to leave. The people working at Fox News live and work in New York. They might complain about aspects of the city, but they like it enough to stay—because New York is a great city. London is, too.
Here’s the obvious clue: the cost of living. The more expensive a city is, the more people want to live there.
San Francisco? Great city. L.A.? Great city. You can tell by the demand and the cost of housing. Real estate prices are through the roof.
My kid and her husband are house hunting in London. Houses are going for nearly £1,000 per square foot—about USD 1,500. And these aren’t mansions. They’re Victorian-era row houses—those narrow homes built in the 1890s for working-class families. They’re packed together in long rows.
And people still pay a fortune to live in them.
From the outside, many do not even look nice by modern standards. They’re small—around 700 square feet. Bedrooms are seven by eight feet. You can barely fit a bed and a bureau.
The floor plans are fucking ridiculous. Some of those houses don’t even have an indoor bathroom. If you want to use the toilet, you have to go out the back door to a separate room that was added in the 1930s.
Initially, they had outhouses in the yard. And no one ever figured out how to retrofit the cramped interiors to include a proper bathroom you could access from inside.
And these places—700 square feet, what they call a “house” but only half of one—might cost $750,000 in a decent neighbourhood.
That’s terrible if you’re trying to buy a home. But it shows what people are willing to pay to live in a desirable part of London—because London is fucking nice.
Yes, you’ll see people wearing headscarves. So what?
The people?
You don’t get a bad vibe from non-whites in London. You don’t get it from Muslims. What you get, most of the time, is people going about their business.
If you’re talking to them in a shop, if they’re serving you, they’re not radiating some hostile “fuck whitey” energy. They’re just part of a diverse, vibrant city. Maybe they’re not thrilled to be working retail, but who is?
Most people aren’t walking around angry, trying to dismantle society.
Last updated May 3, 2025. These terms govern all In Sight Publishing content—past, present, and future—and supersede any prior notices. In Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons BY‑NC‑ND 4.0; © In Sight Publishing by Scott Douglas Jacobsen 2012–Present. All trademarks, performances, databases & branding are owned by their rights holders; no use without permission. Unauthorized copying, modification, framing or public communication is prohibited. External links are not endorsed. Cookies & tracking require consent, and data processing complies with PIPEDA & GDPR; no data from children < 13 (COPPA). Content meets WCAG 2.1 AA under the Accessible Canada Act & is preserved in open archival formats with backups. Excerpts & links require full credit & hyperlink; limited quoting under fair-dealing & fair-use. All content is informational; no liability for errors or omissions: Feedback welcome, and verified errors corrected promptly. For permissions or DMCA notices, email: scott.jacobsen2025@gmail.com. Site use is governed by BC laws; content is “as‑is,” liability limited, users indemnify us; moral, performers’ & database sui generis rights reserved.
In-Sight Publishing 