🤓 What This AI Short Film Reveals About Hollywood’s Future – And Who Actually Wins
PLUS: GPT-5.1 is here; Disney is building its own Sora—and more
Last week I said I’d skip this week to focus on client work and experimentation unless something big happened.
Then GPT‑5.1 dropped. So here we are. (Details below👇 in the weekly updates section.)
I’m aiming to take next week off instead—but terms and conditions apply.
I’m still trying to find a rhythm that lets me stay focused and immersed in the work, while keeping this newsletter valuable and sustainable.
That’s the plan, anyway.
—
Most AI content is slop.
Hippos on diving boards. Babies flying planes. Bigfoot vlogs auto-generated by the hour and pushed to TikTok.
Just unwatchable.
Then there’s the other kind.
The kind with stunning visuals but no story, substance, or heart.
Yet this week, I watched a short film called Minnesota Nice (if you’re from Minnesota, you’ll feel this one)—2 minutes 41 seconds—that’s actually… good.
Not “good for AI.” Just good.
Funny. Tightly scripted. Visually polished.
The characters are lovable. It’s the kind of thing you’d want to keep watching.
It’s from a creator named Josh Kerrigan, working under the name Neural Viz (Wired published a feature on him last month.)
The film was written, directed, edited, and performed by him using various tools including Google’s Nano Banana and Veo, Runway Act Two, and Suno.
Skills Meet New Tools
Here’s what makes Kerrigan’s work stand out from the AI slop flooding the internet:
He didn’t just wake up, download some AI apps, and make this.
He spent a decade learning the craft.
He assisted directors.
Made sketch comedy for Funny or Die.
Sold a pilot to Disney in 2021—which Disney now owns.
Directed a low-budget horror feature.
Learned how to play various roles on sets because he had to.
Then Hollywood’s traditional path collapsed.
The streaming bubble burst. Writers’ rooms shrank. Strikes froze work for months.
So, Kerrigan started experimenting with AI tools and building his own cinematic universe, with its own recurring characters, interconnected storylines, and mythology.
He still writes traditional scripts.
He storyboards every shot.
To get natural handheld camera movement, he films his monitor with his iPhone—then maps that real-world motion onto the AI footage.
He performs every character himself using motion capture.
When the AI tools screw up (and they do constantly), he either works around it or folds it into the story.
One character’s verbal tic (starting sentences with extremely long vowels—”Iiiiiiiiiiiiiii came out here”) came from a glitch. Kerrigan kept it as the character’s signature.
As he puts it:
“Everything I do within these tools is a skill set that’s been built up over a decade plus. I do not believe there’s a lot of people that could do this specific thing.”
The tools didn’t replace his skills.
They amplified them.
Who Will Win With AI
Neural Viz became a cult hit on Reddit and Twitter—then a real one, racking up hundreds of thousands of views on YouTube and millions on TikTok and Instagram.
Studios noticed. Kerrigan met with almost all the majors.
One producer told him something most Hollywood players don’t usually say out loud:
“You might not need us anymore. Power has shifted to creators.”
Other big players echoed this.
Kerrigan turned down a studio job to make his own pilot with an independent producer.
Between that contract and YouTube/TikTok revenue, he was able to quit his day job in January.
He owns his work now. Disney doesn’t.
A lot of creatives in entertainment are scared right now. Understandably, so.
They wonder if the skills they’ve spent their whole careers building still matter.
They do—if you’re willing to use them differently.
Expertise actually matters more than ever.
Because when everyone has access to the same tools, only vision, talent, taste and storytelling separate slop from something worth watching.
Or as Wired put it:
“The winners will be the idea people: writers, directors, storytellers. Idea people who can also wield the tools? They’ll be gods.”
The flood of low-effort content will keep coming (just scroll LinkedIn for 30 seconds 🫣).
But so will work like this.
Hollywood built its power by controlling access to production and distribution.
Those gates are lifting.
What each of us build now is up to us.
—
P.S. I shared Minnesota Nice on Threads and the response was… intense.
Plenty of people liked it, but there was also wave after wave of backlash:
“This is what’s killing the industry”
“AI is killing real art”
“This is trash”
“The voice acting is terrible”
“Go fuck yourself with your AI shit”
Most of them had no idea Kerrigan has spent a decade working in film; I’d kept the post short because of the character limits.
But the bigger point is: By the time you say “AI,” the room is already hot.
You’re stepping into a conversation that’s already emotionally charged—one that’s closely tied to people’s paychecks, identities, self-worth, and sense of belonging.
If you’re someone leading, creating, or communicating around AI in entertainment, this is the tension you’re living in every day.
You’re trying to understand a moving target while helping others feel a little less lost in it.
That requires more than vision or strategy.
It requires emotional intelligence.
The patience to meet people where they are.
The ability to hold both the possibility and the grief.
But I’ve seen what that makes possible:
People making meaningful career transitions in an impossibly tough market—finding ways to align their expertise with where this industry is actually heading.
Execs who felt backed into a corner start to design future roles that play to their strengths.
Department heads who found a way to get their teams the right kind of training now have teams that feel prepared and empowered instead of threatened. They’re doing the best work they’ve ever done, and have become the superstars of their companies.
The transition is still hard. It requires genuine leadership, nuance and care.
But the people willing to engage with both the tech and the emotions around it—the ones making space and opportunities for others to learn, adapt, and experiment—are the ones shaping what the next era of this industry will become from the inside.
📈 Work with Me
AI Advising and Consulting: Strategic guidance to build sustainable AI strategies and adapt to shifting audience behavior
AI Training: Practical training that helps teams use AI more strategically, effectively, and safely—so it becomes a trusted thought partner, not just a task assistant
What You Need to Know About AI This Week ⚡
Clickable links appear underlined in emails and in orange in the Substack app.
✨ OpenAI’s GPT-5.1 is here.
It’s supposed to:
💠Have higher IQ and EQ
💠Be better at instruction-following (understanding and doing what you actually ask more precisely), especially for Custom Instructions.
Custom Instructions allow you to share information about yourself and your preferences with ChatGPT, which is then applied to all conversations to give you more tailored, personalized, and relevant responses.
They’re the most consistently applied layer of context and memory ChatGPT uses for every single conversation, and take priority over all other saved memories.
If you haven’t touched yours in a while (or never set them up), take 15-30 minutes to do it this week.
GPT-5.1 also comes with expanded personality presets: Default, Professional, Friendly, Candid, Quirky, Efficient, Nerdy, and Cynical. I recommend sticking to the default mode.
The model will become the new default model in ChatGPT (replacing GP-5), so there’s nothing you need to switch on your end.
It’s rolling out to paid ChatGPT tiers first, with broader access coming soon.
As I’ve mentioned before, it usually takes me a few weeks (and sometimes months) of living with a model to get a real feel for its capabilities and blind spots.
I’ll also need to adjust most of my prompts so they play nicely with its quirks.
I’ll put in the reps and get back to you with my thoughts.
🎢 Disney+ is building its own Sora inside the castle walls.
Disney+ is planning to let paying subscribers generate short-form AI videos starring Disney-owned characters—think Darth Vader, Iron Man, Anna and Elsa—using tools built directly into the app.
Disney’s reportedly in talks with several AI companies to help power the experience.
It’s a very Disney move: keep the magic inside the park.
Instead of letting Sora or other platforms host fan-made remixes of its worlds, Disney wants that creativity—and the engagement and data that come with it—happening on Disney+.
The upside is clear: more time in-app, more reasons to renew, and a new way for fans (especially kids and families) to play inside Disney worlds.
It also doubles as a live focus group: Disney sees which characters, pairings, and scenarios people actually care enough to create with.
The open question is whether Disney can build a real creation ecosystem on Disney IP alone.
Yes, the characters are beloved. And yes, the appetite for remix culture is massive.
But part of what keeps platforms like Roblox and Fortnite interesting is the ability to jump between different worlds and brands.
Disney’s universe is deep, but it’s still one universe.
That’s a major draw for superfans, but will it be enough to keep them creating once the novelty wears off?
The other risk: the most interesting, culture-shaping experiments are still likely to happen elsewhere, on open tools with fewer guardrails and much bigger reach.
I’m excited to see how it turns out.
🎙️ ElevenLabs launched a new marketplace where brands can license AI-generated voices of famous figures—including Matthew McConaughey (an investor), Michael Caine, Liza Minnelli, and even Mark Twain.
The company formalizes deals with rights holders, giving companies a legal way to use recognizable voices in ads and content.
Once everyone has access to the same voices, they stop feeling like creative choices and start feeling like stock music.
Snoop Dogg’s everywhere-ness is a good example. He’s in so many ads that when I see him now, the surprise and novelty are just gone. I find myself tuning it out—like, oh, there he is again.
He and his agents are racking up the money, honey. Good for him.
But for brands, if the reaction is “oh, him again,” you’ve already lost the moment.
📰 TIME just launched a new AI chatbot that lets readers summarize stories, create audio briefings, ask specific questions, or translate content into 13 languages.
At launch, it covers politics and entertainment.
TIME calls this an “agent,” but it’s really a searchable, interactive front-end for their archive.
As audiences change how they engage with media—expecting fast, tailored answers instead of scrolling through headlines and stories—publishers are trying to keep up.
I did a deep dive on this a few weeks ago in Nowhere to Hide, including how this shift leaves brands with less control over their narratives and story angles.
If you work in marketing, PR, or comms and missed that edition, make sure to read it 👇.
Many publishers have launched their own AI chatbots, but I’m skeptical they’ll drive meaningful engagement.
Why ask a bot trained on one outlet’s archive when tools like ChatGPT and Gemini have been trained on the entire internet—and have licensing deals that include content from major publishers and platforms like Reddit?
And the trend continues….
The No. 1 country song in America is AI-generated.
In case you missed last week’s edition, you can find it 👇:
That's all for this week. See you either in 2 weeks or next Friday (if my schedule allows.)
Thoughts, feedback and questions are always welcome and much appreciated. Shoot me a note at avi@joinsavvyavi.com.
Stay curious,
Avi
💙💙💙 P.S. A huge thank you to my paid subscribers and those of you who share this newsletter with curious friends and coworkers. It takes me about 20+ hours each week to research, curate, simplify the complex, and write this newsletter. So, your support means the world to me, as it helps me make this process sustainable (almost 😄).









I listened to that country song a couple of days ago. It sounds like a real person. If you're listening to it, you would not be able to guess that it's AI-generated. This technology is both unbelievably exciting and terrifying at the same time.