AI in the creative process: scaling drafts while retaining human control
Explore how AI in the creative process helps scale drafts, boosts pre-production, and maintains human control for brand voice and connection.
Where does AI in the creative process actually help? Brandon Baum at Studio B expected it to speed up post-production. The biggest gains turned out to be in pre-production.
Cameron Adams, co-founder and chief product officer at Canva, said standard language models couldn't generate and edit layered designs with visual awareness. So Canva spent two years building its own: “we've spent the last two years building our own foundational model, which we call the Canva Design Model.”
Why do prompt boxes fail at final creative output?
Adams said a prompt box will get you most of the way and then stall: “It'll give you an 80% solution, but you can't really tailor it just based on typing.” Users end up copying the output into another tool, working on it there, then pasting it back into the AI assistant for more changes.
Adams also warned against handing the human part of the work to a machine. He pointed to fashion houses criticized over disturbing AI ads and publishers releasing hallucinated fiction, and said these failures happen when people are taken out of the creative process.
Where does AI in the creative process speed things up?
Baum admitted his first guess was wrong. He expected the main impact in post-production, where his team of 12 visual effects artists works on graphics and 3D simulations. After pressure testing the tools across the team, he said, “Where we're actually seeing the most uplift in use case in these tools is actually within development, within pre-production.”
Ghazi Shami at EMPIRE said AI opens up production to more people, giving creatives access to virtual session players and instruments they otherwise couldn't afford.
The infinite content problem
Michael Nash at Universal Music Group put the volume in perspective. Four years ago, 100,000 tracks were uploaded to services every day. If you never slept, you would need 1,700 years to listen to all the music ever made, and “you'd have to live 17,000 more years for every year that AI models are cranking out new music. That's the volume of output from just one AI model this year.”
Nash said this explosion of AI content has no market and meets no need. He also argued that a technically better AI music model doesn't make better music, because musical quality comes from artistic intent.
Will AI and human creation merge or split?
Kakul Srivastava at Splice predicted that AI will be taken for granted, just like electricity.
Harvey Mason Jr. at The Recording Academy argued the opposite: traditional and AI creation will stay on separate, parallel paths. He also noted that AI-generated songs increasingly tell emotionally resonant stories that sound like what human artists write.
The premium on human connection
Andrea Czapary Martin at PRS for Music predicted that as AI-generated music becomes abundant, the scarcity of live music and the human side of it will be valued more.
Shami agreed that AI could drive more people to live shows: “I've told all my artists the one thing that fundamentally can protect your art is tour, tour, tour.”
Baum predicted that by 2030, with AI able to generate almost anything, creative success will depend more on human storytelling than on technical execution. That's the line running through every example of AI in the creative process here: the tools draft, and people decide what's worth keeping.
To hear more from the people putting AI into production, see the HumanX 2027 program preview.
Frequently asked questions
What are the risks of over-automation in the creative process, and how do leading brands mitigate them?
Cameron Adams, co-founder and CPO at Canva, warned that delegating the human element to a machine causes output failures. He cited examples of fashion houses facing criticism for unsettling AI advertisements and publishers releasing hallucinated fiction. Adams described this as a problem that occurs, “When fashion houses are criticized over disturbing AI ads or fiction articles are actually complete fiction because an AI just hallucinated them” , noting that creators must remain accountable for the final product.
How is generative AI changing discovery and demand generation for creative work?
Anastasis Germanidis at Runway argued that generative AI allows creators to produce films and games with minimal resources without needing to secure funding first. Anastasis Germanidis at Runway argued that this shift means artists actually need to be less entrepreneurial than before. Germanidis noted that before these tools existed, creators had to convince others to back their ideas to get permission to build them. Germanidis argued that because creators no longer need to secure funding first, they can release their work based solely on the strength of their concepts.