AI isn't coming, it is here (Part II)

How Generative AI is Reshaping Storytelling

When Matt Silverman talks about storytelling in the age of AI, you can hear both the excitement and the urgency in his voice. As Executive Creative Director at iBelieveInSwordfish, he’s spent decades translating sparks of imagination into visual stories for some of Silicon Valley’s most innovative companies. But today, he says, the creative toolbox looks nothing like it did even three years ago.

“AI is happening now,” Matt stresses. “This isn’t something that’s coming down the pike. It’s not something that might happen. It’s happening, and it’s happening right now. We’re using it daily in our production work. Our clients are using it in their projects, and it’s going to affect everybody.”

From Prompts to Production

For Silverman, the revolution is not just in what AI can do — but how fast it’s evolving.

“Generative AI is really using tons of computers, millions of computers, working together to essentially code on their own. So instead of everybody writing software code one line at a time, they moved into software 2.0.”

That leap has birthed an entirely new generation of tools:

• ChatGPT for writing scripts and even software with natural language.

• Midjourney for breathtaking still images.

• Runway and Google’s Veo 3 for text-to-video that looks shockingly real.

• Eleven Labs for cloning voices and syncing them perfectly to lips.

• Udio and MM Audio for music and sound effects that match visuals automatically.

• Movieflo.ai, which now strings these elements into a single production pipeline.

“In a very short amount of time for very little money, you could get a very high-quality photorealistic video done in hours,” Matt explains. “You can literally have a 30-second commercial made from concept to completion in half a day.”

AI Influencers and New Characters

The changes aren’t just happening behind the scenes. Even the people we see on screen are being reimagined.

“One of the tools I was just looking at is called Lucent Video. What they’re doing is giving you the ability to text prompt and make a person. And that person is your influencer.”

For years, brands have relied on TikTok and Instagram stars to promote products. Now? They don’t need them.

“Revlon doesn’t need that kid anymore,” Silverman says. “They’re making up fake kids. And those fake kids are now convincing. They look 100% real, and they don’t even have to pay those kids now.”

The Fundamentals Still Matter

But despite all this disruption, Matt is quick to remind creators that AI isn’t magic. It’s still filmmaking — and the core lessons never go away.

“Everything that I learned in film school is still applicable. You need to know composition. You need to know lighting. Whether you’re shooting with a 35mm film camera or a video camera or an AI camera, those techniques still come into play. You still need to understand what an establishing shot is. I then want to cut to a close-up. I don’t want to have a jump cut where the framing is too similar. Whether you’re shooting with a real camera or doing it with AI, it’s still very applicable.”

In fact, he argues that AI might even be the best way to learn storytelling today:

“To me right now, I believe that AI filmmaking is probably the best way for people to learn how to make a film. You don’t need to waste money with a film camera these days. You can make a beautiful-looking image right on your computer or even on your phone.”

A Storytelling Revolution

What excites Matt most is how quickly things are changing. Tools are leapfrogging each other weekly, and workflows that felt impossible months ago are suddenly industry-ready.

“It never ceases to amaze me,” he says. “We’ll start a project in AI where we think we know the best quality we can get. And then a week into the project, something new comes out that changes the game. A month ago, something might take me five days to do. And then out of the blue, a new version of Runway comes out and I can now do it in literally two minutes.”

The future of storytelling, in other words, isn’t years away.

It’s already here.

And yet, as Matt reminds us, the spark of creativity is still the same:

“When a client comes in, they’re spit balling ideas. Sparks are firing in your brain. The job is to get those sparks onto paper, and then into a film. That communication I have with artists — that’s the same communication I now have to give to the bots in order to make generative AI.”

Storytelling remains human at its core. But the canvas has never been bigger.

 

 

 

Dynamic 5-second video, via parameters like --motion, --loop, etc. Midjourney
Also “Introducing Our V1 Video Model” from Midjourney confirms their “Image-to-Video” flow: press “Animate” to make images move. Midjourney

 

 

Lucent can generate avatars, voices, UGC video ads from prompts

The Lucent website advertises custom avatars + voices, editing video sequences, building campaign-ready ad videos quickly. lucent.video+1

 

Runway supports text-to-video / image-to-video, generation & improvements

 

Runway’s own product description: they provide multimodal tools for video generation from prompts or images. Runway
Also “Text-to-Video AI-Generated Media: Gen-2 Runway Experiments” describes results using Runway Gen-2 text-to-video. Medium

 

 

·  Runway’s Gen-2 / Gen-4 developments
Runway releases an impressive new video-generating AI model (Gen-4) — this shows how quickly the video generation tools are improving (better coherence, consistent environment / subject behavior). TechCrunch

Also earlier: Text-to-video AI inches closer as startup Runway announces new model (Gen-2) — supports the claim that text-to-video tools are already in rapid evolution. The Verge+1

·  ElevenLabs and voice cloning + ethical / safety concerns
ElevenLabs and the risks of voice-generating AI — confirms that voice cloning is being used, has potential misuses, and that there are efforts (and issues) around consent, verification, etc. TechTarget

·  Virtual / AI influencers replacing “real” people
AI-created “virtual influencers” are stealing business from humans (Ars Technica) — describes virtual avatars like Aitana López (fictional influencer), brands using them, money being paid to fictional characters. Supports the “fake kids” / “AI influencers” part. Ars Technica

·  Speed & quality improvements: “from concept to photorealistic video fast”
Generative AI’s Next Frontier Is Video (Bloomberg) — about Runway’s Gen-2 as a move toward generating short video clips quickly from textual prompts. Bloomberg.com