AI Insights: why every producer needs to understand vibe coding
How Zinc Media Group is running its in-house AI label, Cicada
When Zinc Media Group-owned indie Red Sauce received a bumper order of the Channel 5 reality series Bargain-Loving Brits in the Sun a few years ago, it ended up with release forms for 12,000 people.
To manage the heavy administrative load, the company’s in-house technology arm built a “release form engine” that would capture the contributor’s photographs and information and reference them against the finished cuts to work out whether they had appeared in an episode.
Zinc - the production group that houses Putin vs. The West indie Brook Lapping and Rob & Rylan’s Grand Tour producer Rex - was ahead of its time back then. Ultimately, the development costs were too prohibitive for the compliance engine to become a viable solution.
But fast-forward to 2026, and “now, suddenly, we’re in this world where the barriers have been removed,” says Olly Strous, chief technology officer at Zinc. “We had this great idea. We know it’s got value to us, and no doubt, it will have value to others. The time is right.”
Strous, who joined Zinc in 2021, is leading Cicada, a new in-house AI label he describes as a “playground with a purpose.” The division is building tools for both internal and external use. The newly AI-turbocharged compliance platform, for example, analyses video for contributors, checks whether they should be blurred and logs the music that’s used. It then generates music cue sheets and completes post-production paperwork.
At the moment, this innovation is only being used internally, says Strous. But “as we develop it further, it certainly has scope to be a product that we bring to market.”
Zinc is among a wave of producers who are embedding - and legitimising - AI within their operations, rather than quietly letting it run in the background.
Strous accepts there is a stigma around certain areas of AI use, namely generative AI. “We feel the same about the bits of AI that will take people’s creative jobs,” he says. “We all joined this creative industry to be creative… But there’s a whole bunch of things that no one really wants to be doing anyway.”
Why set up a dedicated AI label at all? “We want to be able to share what we’re doing,” explains Strous. “If we give it an identity, then we can shout about it, and people know what it is, rather than it being the Zinc Media technology division.”
In this week’s Indie Hustle, Strous breaks down the ways the production company is approaching AI, and the ways it’s being deployed.
Read on to learn:
Why indies are better served finding AI solutions to specific problems than dwelling on the latest tools and platforms
How vibe coding can help producers work smarter and save money and time
The four types of AI use at Zinc Media Group
Why making money in AI is still more about boosting productivity through cost efficiencies than AI-generated work
Forget the tools. What’s your problem?
Many in the industry are focused on keeping up with the latest AI tools and platforms in order to stay current and relevant. But a better approach for producers may be to think about the specific challenges they’re confronting in their day to day — and what AI can potentially fix.
“You’ve got some companies that are chasing every tool out there and desperately trying to stay on top of it and be seen as remarkable leaders because they’re across every single tool the minute it comes out,” says Strous. “Then you’ve got the other side that sees AI as this huge threat and want to avoid it at all costs.”
Cicada is trying to take a balanced approach “where we’re looking at capability rather than tools, and [asking] what is the problem that we’re trying to fix? How can we use these tools to fix a problem?”
Cicada recently carried out an internal audit of sorts, asking Zinc’s entire workforce to detail problems they often encounter in their roles, and ideas for solutions. Strous and his team are going through the best ideas and seeing whether AI could help to solve them.
Strous doesn’t fixate on what tool is used. “I don’t care whether it’s Claude or ChatGPT or generative AI,” he says. “I’ve heard quite a few people going, ‘Which tools should I use?’ and they are getting bogged down in the wrong question. The question is, ‘What do you want to do?’”
Zinc is using AI to do everything from mundane tasks like automatically coding invoices and extracting the general ledger codes for each of them, to testing development ideas against an AI model of a PSB commissioner, to generating drama reconstructions for documentaries that would have been too expensive to fund with real actors.
“If you can go through business and gain a number of incremental efficiencies, and then gain some really big efficiencies or find new, big revenue streams from the things that we’re doing, then that’s the sort of holy grail of AI,” says Strous.
Takeaways: don’t fixate on the AI platforms and tools; figure out what problem you’d like to solve and look into the ways you can do so using AI; be AI platform-agnostic
Why every producer needs to understand vibe coding
The deployment of AI at Zinc breaks down into four areas: chat, generative, agents and vibe coding. “We’ve managed to use all four of those disciplines across everything that we’re exploring,” says Strous.
Chat involves basic search and critical thinking questions, while agents perform certain tasks, such as the aforementioned invoice coding. Generative is used selectively (depending on the clients and what they are comfortable with) and largely for drama reconstructions in non-fiction production.
Vibe coding, however, has allowed Cicada to innovate with breakthroughs such as its compliance platform. Simply put, if you have an idea for an app, vibe coding enables just about anyone to develop it without needing to understand code or work with a developer (although Strous recommends doing the latter to fine-tune the product).
You describe your idea using detailed prompts, and an AI agent creates it. Other companies that use the practice include Deep Fusion Films, whose CEO Ben Field detailed its use to Indie Hustle earlier this year.
“Historically, you would have gone and spoken to a developer [about your idea for an app], and they would have told you phase one will cost X amount and take Y number of weeks, and you go, ‘I’m not going to see the ROI because that was a cool idea, but it’s never going to come to light. It’s too expensive,’” says Strous.
But with vibe coding, “You take the same idea you would have given to the developer, you put it into a vibe coding app, and it goes and builds it really quickly and pretty efficiently,” adds the exec. “You can get a working app in a number of hours.”
Zinc primarily uses Anthropic’s Claude Code, which costs around £15 per month for a “pro” subscription, and £75 per month for a “max” subscription.
Takeaways: vibe coding allows users to create apps in hours by bypassing code generation and simply inputting English-language prompts into Claude Code; it has allowed the creation of a compliance platform for Zinc
Generating AI-related revenue
Zinc made around £3 million in 2025 from AI-related content. However, Strous underlines that “we would not have made that money if the AI market didn’t exist,” he says.
“It’s not necessarily that someone’s procured a fully generated advert from us that’s exclusively made with AI,” he says.
Instead, Zinc produced an AI trade show event for Abu Dhabi-based AI firm G42, as well as a commercial starring Lewis Hamilton that incorporated some AI.
“The adverts we created for G42 had AI holograms and AI-generated imagery in the background of Lewis Hamilton,” explains Strous. “Once again, the AI industry opened that door to us to create something that is AI.”
The next step for Zinc, however, is to generate £3 million from “actual AI work.”
“That’s the bit we’re trying to navigate,” says Strous. “The bit that everybody’s trying to navigate at the moment is actually generating [money]. We’re more likely to generate £3 million worth of savings through efficiencies with our business than we are to see £3 million worth of generative AI content.”
Takeaways: Zinc made £3 million in AI-related revenue last year, though it was more about cost efficiencies; the company wants to make the same amount from AI-generated content
Hiring? We feature paid job listings, to book a listing, email hello@theindiehustle.com
Sponsorship? If your organisation would like to find out about sponsorship and advertising opportunities, email hello@theindiehustle.com