AI’s Human Problem: Brilliant Backgrounds, Blunt Decisions

At ARROW VANE, we pride ourselves on testing every shiny new tool that promises to boost agency performance. And let’s be clear: the best of today’s AI is nothing short of miraculous at background research. Want a market snapshot, trend analysis, or a first-draft campaign framework? AI can chew through data mountains in seconds, surfacing actionable insights that used to take our team days.

But here’s the rub: as incredible as that output can be, AI’s true weakness is making nuanced decisions—the kind that shape client outcomes, team progress, and ultimately, project success.

Nike - “Just Do It”

The Framework Is Solid… But The Fork In The Road Is Human

Brands like Nike and Qantas still make waves because somewhere, someone had the courage to back their conviction as much as the data. AI would have handed Nike thousands of slogan options—but would “Just Do It” have survived an algorithmic filter for clarity or risk? Not likely. As for Qantas, the emotional pull of “I Still Call Australia Home” required a leap of faith beyond surveys and sentiment scores.

On the agency side, AI builds out immaculate project trackers, automated Asana assignments, and resource planners. But faced with two good creative routes—or a client nervous about a risky campaign—the decision doesn’t magically appear in your inbox. More often, real people stall, seeking more analytics, more “rational” backup. Paradoxically, AI can amplify decision procrastination: it provides endless research and recommendations, but stops short of helping us actually choose.

Qantas - “I Still Call Australia Home”

Case in Point: Modern Decision Paralysis

Take a look at some well-known brands today. Despite having top-tier AI tools, even giants in e-commerce, software, and FMCG still land in project gridlock when multiple viable paths seem equally defensible. With so much data at their fingertips, decision-makers fall into the trap of seeking “perfect clarity”—a state even the smartest AI cannot provide.

Even inside our own agency, we’ve seen teams circulate AI-powered reports for extra eyes instead of trusting their instincts and moving forward. Automated research is a godsend—but when it’s time to greenlight a campaign or take a creative risk, AI stands politely on the sidelines, and the project slows to a crawl unless a real person steps up.

Cadbury - “There’s a Glass & a Half”

What AI Can And Can’t Do

Let’s give credit where it’s due: AI makes us faster, sharper, and more informed. It uncovers gaps, surfaces trends, and builds frameworks that make our decisions smarter. But the final call—especially when nuance, context, or gut feel matter most—remains stubbornly, wonderfully human.

So, as you integrate AI deeper into your workflow, remember: a faster car doesn’t choose the destination; it just helps you get there, once you’ve decided where to go. The challenge isn’t what AI can do for us, but whether we’re ready to make the call when the research is done and all that’s left is to leap.

—
David Gough
Managing Director, ARROW VANE
david.gough@arrowvane.com | LinkedIn