Artificial intelligence continues to drive the news cycle even as most media executives remain unsure about how exactly to fully leverage it across the enterprise. In many ways, this week was yet another reminder that AI moves fast – really fast – and execs that survive the next five years are probably the ones learning the AI ropes now. Later will be too late. Of course, top of mind this week was Alphabet’s Google I/O 2025 annual developers conference, which birthed a flurry of AI announcements ranging from Chrome’s AI mode (publishers aren’t happy) to Gemini 2.5 Pro (out of preview next month) to its Android XR smart glasses (watch out, Meta) to Project Astra (AI assistants) to Project Mariner (agentic AI that can complete tasks using specific tools) to Google Beam, which turns video conferencing into a 3D-like experience. And don’t forget that in December, Google introduced its “Willow” quantum computing chip that represents huge advances in error correction and performed a benchmark calculation in under five minutes—one that would have taken a modern supercomputer about 10 septillion years to complete. You heard that right. And when Google and other players can combine reliable quantum computing with AI tools, it’s off to the races for humanity (utopia or apocalypse is anyone’s guess).
All of this is a big change from a year ago when Google’s AI efforts attracted more derision than praise as smaller startups like OpenAI ran circles around it. Not anymore. And for the media industry, none of the announcements hit quite as hard as Flow, Google’s new “AI filmmaking tool” that leverages Google DeepMind’s most advanced models including Gemini, Imagen, and its newest video generator Veo 3 to create ultra-realistic 1080p videos that include advanced camera movements, sound effects, and even spoken dialogue.
Early users have been posting clips all week, and if they’re not turning plenty of heads in Hollywood, they certainly should be. Of course, all this power now comes at a hefty price: Full use of Flow with advanced features like Veo 3 requires the new Google AI Ultra subscription, which runs $250/mo and greatly exceeds the standard $20/mo Pro subscription. The cheaper Pro subscription allows up to 1,000 monthly AI credits while the 15X more expensive Ultra plan enables 12,500 AI credits per month. High-quality video generation uses massive compute, so it’s unlikely anyone will ever get unlimited usage – and that’s important because creating AI content may in fact involve copious trial-and-error generations to get something that remotely passes for a movie or TV show. Don’t worry: We assume more expensive enterprise plans are coming. Traditional content creators once scoffed at the notion of AI generating entire pieces of content with no human actors (and most still resist the idea), but Veo 3 and other cutting-edge models have shown it’s already possible and will only get better, cheaper, and easier going forward. First, in September 2024, Lionsgate inked a partnership with AI video startup Runway to create a customized AI model trained on its own library, presumably to leverage its own intellectual property to fuel future AI content. Then last month, production company Fremantle launched Imaginae Studios to directly leverage AI tools in filmmaking – an idea that would have been completely taboo only a couple years ago. And then this week, we learned that ex-Xumo CEO Colin Petrie-Norris has created Fairground Entertainment to create fully AI generated TV shows by year-end. Do you see the progression here?
Next? In only a few months, the industry has gone from dipping its feet into the AI waters to what could soon become a marathon swim across an ocean of potential. It’s not hyperbole to imagine a world only two or three years from now in which fully generated AI videos compete handily for ad dollars and viewer eyeballs with the likes of “The White Lotus” and “Squid Game.” And no, you don’t have to like it. And while Fairground wants to help the “creators” who are making their own IP and distributing it through platforms like YouTube (yes, it all circles back to Google), it’s also looking to create a central platform that can help connect AI creators to the broader entertainment industry, including streamers like Netflix, whose co-CEO Ted Sarandos just told investors during the company’s Q1 earnings call that he wants to expand the net beyond traditional channels and find new voices outside the Hollywood system. Perhaps the aptly named Fairground could be a literal middle ground between the innovative creators and the Hollywood forces quite intrigued by the idea that AI could lower costs and, in some cases, even enable better set pieces outside the physical budget. Why build a $250,000 set when AI can create one with a $250 prompt?
The bottom line: advances are happening faster than most realize. Also this week: Anthropic announced Claude 4 models that can work on complicated coding tasks for several hours at a time. Meanwhile, reports this week suggest Amazon plans to greatly step up its contextual advertising capabilities using AI, suggesting that AI video generation is only a small part of a story that likely will infuse AI across all aspects of production, marketing, and overall monetization. Scrappy content and tech company Cineverse also this week launched the Cineverse Technology Group “focused on rapidly scaling its technology monetization efforts and accelerating AI-driven innovation across the global entertainment ecosystem.” That includes AI-
assisted project greenlighting, marketing campaigns, audience targeting, rights optimization, advanced content valuation, AI-powered localization, and ad optimization, among other goals. Whether it’s content generation or tools designed to market and distribute that content, AI has arrived. This week was just another example of why anyone who buries his or her head in the studio sands likely won’t survive the next five years. Don’t be that person.
