🤓 Hollywood’s AI Obsession Has a Blind Spot

PLUS: Sora soars & sparks copyright chaos; ChatGPT has apps now—and more

✋ Quick note before we dive in:

If you missed the Updates section in last week’s newsletter, it had a full breakdown of OpenAI’s new launches

The changes are already live, and they touch every aspect of every business.

If you lead a team (or are in one), this is stuff everyone should know about, so make sure to read it.

---

Hollywood is obsessed with what AI can make.

Storyboards. Concept Art. Visual effects. Sizzle Reels.

All while cutting production costs.

And that’s understandable.

Generative AI generates content.

And Hollywood is in the content business.

So, it makes some sense that the loudest conversations are about how we can use these tools to create the kinds of content the industry already produces.

But here is the irony:
The flashiest uses of AI may actually be the least transformative.

While we debate whether AI can write a screenplay or generate a convincing explosion… other industries are asking an entirely different question:

How can AI make us smarter at the strategic thinking that actually moves the business forward?

And they’re building capabilities around the answer.

A Different Kind of Power

At the heart of AI progress over the past few years are large language models (or LLMs)—the brains behind tools like ChatGPT, Claude, Gemini, and Copilot.

They’re what power the thinking layer.

They’re the reasoning engine.

They’re what help image and video models get better over time.

They’re what’s driving breakthroughs in robotics, after years of stalled progress.

And they’re what’s allowing other industries to do strategic work that wasn’t possible before—at least not within normal time and resource constraints.

Other industries get this.

They’re using AI to think through problems that would normally require armies of consultants and months of analysis.

💊 Big Pharma uses AI to discover new drug formulations and figure out which will resonate with doctors before spending millions on trials.

🔮 Consumer brands are feeding in decades of behavior data and asking, “What product should we launch 18 months from now that doesn’t exist today?”—then testing it against economic models, demographic shifts and emerging cultural trends.

🛍️ Retailers are simulating Black Friday scenarios across 500 stores to optimize everything from inventory to in-store displays.

📈 Financial firms are mapping competitors and risk scenarios in hours instead of weeks and analyzing risks and opportunities their teams might not have had time to fully explore before.

Even the CEOs of top AI companies are using their models to guide their strategy and high-stakes decisions.

In a recent interview, OpenAI CEO Sam Altman shared that his team has asked their models for strategic advice, and that the answers are genuinely insightful.

They’ve realized that the real power of generative AI isn’t what it creates.

It’s in how it can sharpen our thinking.

What AI Makes Possible

Here are just a few examples of what media and entertainment companies could be doing today:

1. Stress-Test Positioning with More Depth

   Teams already spend significant time researching how to position a property—analyzing the competitive landscape, cultural trends, and audience data to determine whether something should be marketed as a ‘prestige horror’ vs. a ‘psychological thriller’, or whether to lead with star power or concept. AI can extend that work by simulating how different audience segments might respond to various positioning approaches: the core message, genre choice, your tagline or synopsis—so you can spot disconnects, and refine before locking them in.
   

2. Expand and Personalize Brand Partnership Outreach
   AI can take analyze a film or TV show’s themes, core audience, tone, and launch window, and match it with brands whose goals align. It can also research each brand’s current priorities, analyze recent campaigns, and draft personalized pitches that feel tailor-made for them.
   

3. Revive Catalog Titles Without Big Campaigns
   Libraries are full of titles that could find new audiences if positioned against the right cultural moment. But tracking every emerging trend and manually matching it to your catalog isn’t realistic. AI can monitor cultural signals—from social conversation to search patterns to a new nostalgia wave no one saw coming—and flag when an older title suddenly aligns with what audiences are craving. It can also suggest low-lift ways to repackage and promote them without a big spend.
   

4. Help Talent Prep for Interviews
   Before your filmmakers and actors sit down with press, especially in high-stakes moments or around sensitive topics, AI can simulate tough questions based on the outlet’s approach, the journalist’s past work, and current conversations and controversies. It can identify pressure points and help craft responses that are emotionally resonant and stay on message without sounding defensive.

None of this requires new technology. It’s all available today.

The answer isn’t another tool.

It’s learning to think with the ones you already have.

Even when some entertainment teams do use AI, they tend to treat AI like a glorified intern or a task rabbit, used mostly for surface-level tasks.

“Write me a synopsis.” “Summarize these notes.” “Draft a social post.”

It’s like using a Rolls-Royce to deliver pizza.

To use AI as a strategic thought partner, not just a tool for tasks, takes deep fluency.

And fluency takes more than access. It takes training, experimentation, and iteration.

Why Is This Happening?

Here’s the uncomfortable truth:

The people making decisions haven’t had the chance to use these models enough to understand them.

They’ve seen demos. They’ve read articles.

But they haven’t had the hundreds of hands-on interactions it takes with ChatGPT or Claude to develop real intuition for what’s possible.

And worse, we’re nearly three years in and most companies haven’t made in-depth training a priority.

Last week, I spent four hours with a CEO brainstorming and working through decisions with AI. At the end, he just looked at me and said, “THIS is AI?”

He couldn’t believe he’d been missing this level of thinking, and said it fundamentally changed how he approaches everything.

Meanwhile, that gap in understanding is being filled with expensive mistakes and missed opportunities.

Which is exactly how high-priced consulting firms (you know the ones) have been able to position themselves as experts and have shifted the bulk of their revenue to AI consulting.

Most of their AI divisions are led and staffed by folks who don’t understand how the underlying models work, let alone how to best use them in practice.

It’s the blind leading the blind.

Except one side is collecting millions while the other is left holding the bag: failed pilots, projects that are obsolete before they launch, and months of lost time in a space where every week counts.

Granted, the industry has been focused on real concerns: copyright and IP, union negotiation, budgets

Those conversations matter.

But while they’re happening, the capability gap is widening.

The companies that will dominate Hollywood in five years won’t be the ones with the best AI-generated visuals.

They’ll be the ones who invest in:

• Foundational AI literacy across leadership and teams to better understand the tech‘s capabilities and limitation, and to develop instincts for when AI adds value and when it doesn’t.

• Fluency in prompting to learn strategies and techniques for getting genuine insights instead of generic outputs and “work slop.”

• Developing habits that treat AI as a strategic thought partner—not a task rabbit. Not just for speeding up work, but for thinking through the parts that actually shape outcomes: research, analysis, strategy development, and decision-making.

• A culture that rewards practical experimentation. Teams need time, space, and permission to explore, test, learn, and iterate—because hands-on learning is the only way to figure out what works in practice.

The question isn’t whether AI will transform entertainment.

It’s whether Hollywood figures out the real game before Silicon Valley changes the rules entirely.

📈 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.

---

🎯 Sora is growing. So is OpenAI’s advantage.

OpenAI’s new social video app, Sora, has become the new king of the App Store, rocketing to the top spot on the U.S charts, even as it remains invite-only.

It hit 1 million app downloads in less than 5 days, even faster than ChatGPT did.

But the backlash has been just as fast.

Excited users flooded the feed with AI-generated videos using copyrighted characters.

Studios and talent agencies pushed back.

Agencies are the loudest because they’re the most vulnerable. They don’t own any IP and still don’t seem to understand the basics of this tech enough to know actor “vaults” are not an actual line of defense against AI. 🤯

Last Friday, Sam Altman posted a blog announcing a few key changes:

• OpenAI will shift to an opt-in model for rights holders, meaning studios would need to give explicit permission before their characters can appear in Sora-generated content.

• It’s exploring-revenue sharing models between creators and rights holders. Details TBD.

• It added more granular control over how users’ likenesses can be used in the “cameos” feature.

But Sora’s growth hasn’t slowed.

And that’s the part that should really have Hollywood’s attention.

It already has Mark Cuban’s.

He’s made his Sora Cameo—basically his AI avatar—available for anyone to use.

You know who else gets it? Jake Paul.

His Sora Cameo helped him rack up 1 billion views in just 6 days.

👉 What this means for IP strategy:

Hollywood is understandably focused on protecting its IP.

But if we’re not careful, we risk repeating the mistake we made with YouTube.

Back then, studios spent years trying to shut user-generated content down.

Meanwhile, YouTube became the most powerful entertainment platform in the world.

Tech companies don’t wait for permission.

They build products people love.

They move faster than any regulation and are always a few steps ahead (especially since they’re using their most advanced yet unreleased models to define their strategy).

And because of that, they negotiate from a position of power, especially with individual IP holders.

Just look at how content licensing with publishers has played out.

Even The New York Times—arguably the most powerful news brand in the country—is still largely sidelined as AI platforms cut deals and offer visibility to their preferred partners.

Big changes are already underway.

Hollywood can’t afford to keep fighting for the business models of yesterday.

But you can’t shape what you don’t understand.

And right now, too many companies are trying to control a technology that’s complex by nature—and evolving faster than anyone can keep up from the sidelines.

If the goal is to protect and expand your IP, the first step is building deep fluency in how this technology actually works—with people who already live and breathe it—so you can spot opportunities early and set the terms, not just react to them.

This industry already understands audiences better than anyone.

The companies that win will be the ones who combine that intuition with a real grasp of the technology shaping what audiences want next.

P.S. If you’re ready, but don’t know where to start, start here.

---

Meanwhile…

Within hours of Sora 2’s launch, tools to strip its watermark—an adorable, tiny cartoon-eyed cloud logo—started flooding the web.

The watermark was supposed to help audiences tell the difference between real videos and AI-generated ones.

But it didn’t stand a chance. And neither did we, as unlabeled Sora clips started going viral.

One, shared on TikTok, racked up nearly 5 million likes before anyone realized it was fake.

If you don’t know what the watermark looks like, you can spot it in this video 👇 I created and shared last week.

Oh, but there is more.

Just to make things more interesting, someone made a site that adds Sora watermarks to real videos. 🤯

So now, you might think a real video is fake.

That’s the AI Wild West we’re living in.

---

✨ You can now use apps inside ChatGPT.

Apps like Spotify, Expedia, Coursera, Zillow, and Canva can now run directly inside the chat, so you don’t need to switch tabs or re-explain what you’re working on.

That means you can just ask:

🎵 Spotify to build a playlist for your Friday night dinner party

🧳 Expedia to find flights to Austin under $400

🏡 Zillow to show homes for sale near your location

🎨 Canva to help design a one-pager for your pitch deck

The apps can access your full conversation context. So if you’ve been chatting about a birthday party, then ask for a playlist, Spotify will already know it’s for that event.

And if you’re not sure which app to use, ChatGPT might suggest one automatically, based on your conversation.

A new app directory is coming soon to help you explore what’s available, with more big-name apps like Uber, DoorDash, and Target rolling out next.

Ultimately, this strategy doesn’t just make ChatGPT much more powerful. It might help it become the new default homepage for the web.

—

🛠️Other Dev Day launches for developers or teams building with AI:

• A new Apps SDK for building interactive in-chat apps. Monetization tools are planned but details are TBD

• An Agent SDK for automating multi-step workflows

• GPT-5 Pro and Sora 2 are now available in the API

---

---

🧭 Reddit is winning the AI game.

After licensing deals with Google and OpenAI, Reddit is now the most-cited source in AI answers.

That visibility tripled its traffic and turned it into a critical growth channel for publishers. News orgs are building subreddit communities to stay relevant, while Reddit cracks down on AI scrapers and pushes to get paid what it’s worth.

---

🦃 This Thanksgiving, consumers and audiences are asking AI what to buy or see.

Adobe Analytics, which tracks over a trillion visits to retail sites, predicts AI-powered shopping traffic will surge 520% this year.

Thanksgiving Day alone is expected to see a 725–730% jump in AI traffic year-over-year.

• More than 50% of shoppers plan to use AI for product research

• 36% are using it to hunt for the best deals

• 30% are using it for gift inspiration

• About 40% are looking for personalized recommendations

🚨 If you’re a seller of products, services or content, make sure you’ve evolved your marketing strategies to optimize for AI visibility and discoverability.

P.S. One of my favorite ChatGPT use cases is finding coupon codes for online stores. It almost always delivers.

If you have a Plus subscription ($20/month), you can even take it one step further with ChatGPT agent by asking it to verify the code automatically.

👉 To try it: Click the ➕ icon in the ChatGPT message bar and select Agent mode.

Then use a prompt like:

Find discount codes for <store>, then go the site, add an item to the cart, go to checkout and test the discount code to confirm it works.

In case you missed last week’s edition, you can find it 👇:

That's all for this week.

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 😄).