According to Business Insider, Jules Love, founder of the consultancy Spark AI, has worked with over 60 creative agencies on AI integration. He argues the core challenge is psychological, not technical, and that successful adoption requires deliberate effort. Love insists agencies must assign clear ownership for AI, protect time for experimentation, and provide role-specific training. He warns that using AI merely to work faster risks undermining an agency’s value, pushing them toward a race to the bottom on fees. His advice is to shift focus from doing today’s work more quickly to doing tomorrow’s work better, or risk becoming “old-fashioned” and “uninteresting” by 2027.
The Psychology Problem
Here’s the thing: Love is spot on. The biggest wall isn’t buying a ChatGPT subscription; it’s getting a team of creatives to actually use it without fear or secrecy. When he says a lack of openness is the “No. 1 sign” of a bad AI culture, that hits home. How many people are quietly using these tools, afraid a manager will think they’re cheating or cutting corners? That’s a leadership failure. Creating a space where people can share failed experiments is way harder than rolling out a new software license. It requires a fundamental shift in how we view “work.” If the billable hour is still king, experimentation will always feel like stealing.
Training Isn’t Optional
His Lego analogy is perfect. Handing a team ChatGPT with no training is like giving them that giant, no-instruction box of bricks. They’ll maybe build a wobbly wall, but they’ll never build the cool spaceship. The “picture on the box” is the vision for how a copywriter, an art director, or a strategist uses this tool specifically. Without that, it’s just a novelty. And let’s be real, “training” here doesn’t mean a boring HR webinar. It means practical, peer-led sessions on crafting a creative brief for an AI, not just typing a one-off prompt. That’s what turns a gimmick into a genuine capability.
The Commoditization Trap
This is the most critical insight, and honestly, most agencies are sleepwalking right into it. If all you do with AI is produce mood boards, first drafts, and social posts faster, what’s your value? You’re just a cheaper, faster commodity. Clients will absolutely start asking why they’re paying your hourly rate for work that’s now done in minutes. Love’s push toward outcome-based pricing and fixed-cost projects is the only viable escape route. It forces the conversation away from “how fast” and toward “how good” or “how strategic.” It gives the agency room to breathe, learn, and actually use AI to enhance creativity rather than just accelerate production. But making that shift? That’s a brutal, fundamental business model change.
Who’s In Charge Here?
So, who makes this happen? Love says you need someone accountable, and I think he’s underselling how hard that is. This isn’t a part-time job for the “digital innovation lead” who also has 30 hours of client work. This person needs real authority and protected time—time that, yes, is not immediately billable. They need to be the evangelist, the trainer, the librarian for prompts and workflows. They need to kill the secrecy. Basically, they need to make AI boring and normal, just another tool in the rack. Agencies that treat this as a side project will be the ones Love is talking about in 2027: old-fashioned, expensive, and uninteresting. The clock is already ticking.
