Even as AI-generated video ads go mainstream, many teams still struggle to produce TV-quality creative fast and affordably.
Stephen Parker, Chief Creative Officer at Waymark, explains how structured AI production pipelines are beginning to change that by widening access to CTV ads.
Using AI to generate multiple ad options early can help companies choose a direction faster, spend less on trial-and-error, and produce polished ads at a lower cost.
AI-generated ads have officially arrived. But where the public sees strange visual errors in high-profile campaigns for brands like Coca-Cola, many professionals see something else entirely. For those working at the forefront of this technology, the focus isn’t on a single magic prompt. Instead, it’s on a sophisticated chain of systems working in concert—a process with the potential to quietly reshape the economics of creative production and open television advertising to a new class of businesses.
Stephen Parker, Chief Creative Officer at Waymark, lives in that world every day. With a long tenure in design leadership, including more than 8 years at Squarespace, his perspective offers unique insight into the intersection of design, technology, and advertising. The entire conversation about AI needs a fundamental reframing, according to Parker. Rather than a critique of imperfect final products, he advocates for an appreciation of the intricate architecture running behind the scenes.
Despite popular belief, most professional-grade AI video relies on a multi-layered process called "ensembling," Parker explains. "It's extremely rare to just type a prompt and get a finished result. For us, that means 14 or 15 different model instances are running just to infill a template we've already constructed. We're orchestrating the music, story, script, and voiceover, and analyzing the business's history and assets to bring it all together into something cohesive."
Today, that "ensembling" process unlocks what Parker calls "AI's greatest strength" in the current market: generating creative optionality, for free. In an industry where concepting a single campaign can sometimes cost hundreds of thousands of dollars, AI lets agencies visualize multiple directions instantly.
Safe, funny, curveball: By de-risking the creative process before committing significant budget, organizations can drastically lower the cost of entry for high-quality creative. "We can now generate a safe option, a funny option, and a curveball option. Instantly, we can give that optionality to the customer and ask what ballpark they want to play in with the ad," Parker says. "It's helping us get faster at seeing all the optionality that's there, in a way that's affordable for everyone, not just the biggest brands and agencies in the world."
The new efficiency is poised to unlock a massive, overlooked market. For Parker, the moment represents the collision of two different histories in advertising. Over the last few decades, linear television gave major brands prestige and massive reach, though it couldn't provide granular data. In parallel, platforms like Facebook empowered a new generation of small and medium businesses with hyper-specific targeting, just not in a premium broadcast environment.
The 80% prize: "We're used to the money coming from big brands, but in truth, that's only about 20% of the potential money that's out there. The SMBs have so much to offer." His observation is backed by recent market data showing that smaller brands are adopting generative AI for video at a higher rate than their larger counterparts.
Prestige meets precision: Connected TV (CTV) can finally bring both worlds together, Parker explains. "Connected TV finally has the technology to offer granular tracking and quick distribution. That is happening at the exact same time as the advent of faster, AI-powered advertising. What we're seeing is the merger of those two forces."
Of course, this speed raises a critical question: "Have you certified that the water is safe, for lack of a better metaphor? Is this really a utility we can use? I think that's a big question mark for people," Parker explains. Meanwhile, industry anxiety about the unknown variables in AI training data has many businesses scrambling for indemnification.
From chaos to CAD: A pragmatic interim step is to focus on lower-risk content, such as logo animations and product-focused shots, Parker explains. But a more durable technical solution is emerging from model makers as well. "What professionals want from these models is specificity. In auto advertising, you need to put a specific car model in a scene, and it has to be exact," he says. "AI model makers are now competing on these features to serve professional creatives."
Looking ahead, Parker predicts that the technology will soon evolve into a proactive creative partner. Today, that forward-looking vision guides his team toward what he calls the next frontier of advertising, or "AI as a Service."
The proactive package: With AI tools already being integrated into CRM systems, this is the tangible direction of current product development—to automatically generate spec ads for sales teams. "We should look forward to the equivalent of an Amazon package knocking on our front door, saying, 'Hey, I was just designed for you yesterday. Am I a device you'd like to try out for a couple of days?' That's how I think about it."
Ultimately, for AI to evolve from a controversial gimmick into a trusted, transformative tool, the industry can’t wait for a perfect solution to emerge. The path forward, Parker concludes, requires a dual commitment: from creatives to abandon the sidelines for hands-on experimentation, and from major brands to stop chasing easy PR wins and start leading the difficult moral conversation that will build lasting trust. "I don't know how we get there, but I really wish somebody would do that."
