AI Creative Direction Is Here: Jamey Gannon
Summary
AI creative direction isn't magic—it's a learnable tool that's spawning a new class of designers who blend human authorship with AI production. The key isn't getting perfect outputs in one prompt; it's using references, starting conditions, and iterative refinement to solve blank canvas problems.
Key Takeaways
- Stop evaluating AI tools by whether they pass your bar in a couple of prompts. Instead, assess whether they can meaningfully solve blank canvas problems and serve as a starting point you can refine—a fundamental shift in evaluation criteria.
- References are underutilized in AI prompting. Beyond image uploads, include mood boards, style references, lighting references, and even reference copy or authors—this does more work than trying to describe everything you want in text.
- AI can encode taste by understanding strategic frameworks (like Seth Godin's thinking patterns) and applying them to outputs. Mid-journey mood boards and model selection demonstrate taste is partially codifiable, even if not fully.
- Model selection is critically important but often overlooked. Different models handle different tasks (image vs. copy vs. strategy), and choosing the right model is as important as the prompt itself.
- The practitioner class of AI creative directors aren't debating philosophy—they're shipping work and building new workflows. This pragmatic approach, not theoretical arguments, defines the future of design.
Related topics
Transcript Excerpt
This is like another misconception people have is like AI is like different in some way than like any other tool which like obviously it is but in terms of like learning it, mastering it, bring it into your process. It's like it's not really magic. Uh it feels like magic, but it's not. >> I'm seeing a pattern with creative directors. I saw it at Iverson's studio with individuals like Brett from DesignJoy. I saw it at Hunter Hammond's offen agency. There's this growing class of AI powered creative directors who are merging human authorship with AI production work. And the more you see it, it makes all the bickering amongst designers on the internet start to feel performative. And these designers aren't wasting their time on the opinions of others. They're too busy delivering to clients, bui…