THE WHOLE STORY

I am an AI creative director and educator known as Waviboy. I teach 225,000+ creatives how to direct AI to a standard brands pay for, run the Waviboy Academy, and founded AIARTDIRECTION.COM, the reference site for the discipline. Before any of that I made clothes, built a digital human, and taught corporations how to think about AI. Everything below is how those threads became one job.

Waviboy was a fashion label before it was anything else. I designed, produced, and sold clothes, and I ran the label the way you run a startup: small drops, direct audience, everything learned by shipping. Fashion taught me the thing no tool has ever taught anyone since: taste. What a campaign is. Why one image sells a jacket and an almost identical image does not. How light, styling, and casting carry meaning before a single word gets read.

It also taught me that craft alone does not pay rent. The label was entrepreneurial by necessity. I learned marketing because nobody else was going to do it, and that instinct, craft plus commerce in the same brain, became the pattern for everything after.

Somewhere in the label years I built a digital avatar called Lil Wavi, a virtual character with a wardrobe and a point of view, back when digital humans were still a novelty. Lil Wavi got picked up by the fashion press and earned features in i-D and Dazed.

I did not know it then, but that was the first AI creative direction I ever did: directing a synthetic character to a standard real publications would print. Years later, when generative models arrived and everyone asked whether synthetic imagery could sit next to real photography, I had already lived the answer.

After fashion I worked as an innovation strategist. The job was helping large organisations understand what new technology actually means for them, past the hype and the fear. When AI broke into the mainstream, that became the job: I educated global brands, including Pernod Ricard, on implementing AI across their marketing teams and their C-suite.

Those rooms taught me the other half of the discipline. Creatives ask whether AI output is good. Executives ask whether it is consistent, on brand, legally safe, and repeatable at volume. Both questions matter, and almost nobody could answer both. That gap is the space I have worked in ever since.

When image models crossed the line from toy to tool, I saw the three threads of my working life braid together. Fashion had given me the eye. Lil Wavi had given me synthetic characters before they were normal. Strategy had given me the language of brands and boardrooms. AI creative direction is all three at once, and I went all in.

I started publishing what I was learning as @bywaviboy: how to direct models instead of just prompting them, how to keep a face consistent across a campaign, how to tell directed work from slop. The audience grew to 186,000 on Instagram and close to 40,000 on TikTok, 225,000+ creatives across social media, because the work answered the question everyone had: how do I make this actually good?

The content became a course, and the course became the Waviboy Academy: a full training programme for AI creative direction, from taste to tools to selling the work, running on a platform I had built from scratch. At its peak the business did over half a million pounds a year.

The results that matter most are not mine. Academy students have gone on to run working AI creative agencies, sign retained brand contracts, and field inbound from brands as large as L'Oreal Paris. One student got contacted by a L'Oreal Paris creative producer with 130 Instagram followers. The work did the talking, which is the entire thesis of the discipline.

Then I learned the lesson that now shapes everything I build. My revenue was coupled to Instagram reach, and when the algorithm turned, reach fell by two thirds and revenue followed it down by more than 80 percent in six months. A business I had spent years building nearly died in two quarters, not because the product got worse, but because I had built it on rented land.

I could dress that up as a strategy lesson, and it is one: own your audience, own your data, own your platform. But it was also just brutal. What came out of it was a rebuild around things an algorithm cannot take away: an email list, a subscription product, a custom platform, and a reference site designed to earn search and citations instead of impressions. The crash is why AIARTDIRECTION.COM exists.

Today the work is three things. The Academy teaches the discipline. The @bywaviboy accounts show it in public. And AIARTDIRECTION.COM gives the profession its reference: the canonical definition, the Slop Test, the glossary, and The State of AI Art Direction, the first industry census of what AI art directors actually earn, publishing winter 2026/27. The formal certification for the discipline arrives in 2027.

The goal is simple to say and long to do: when the world asks what an AI art director is, what they earn, and who is good, the answer should have a home. I am building the home.

• Taste over tech. Tools change monthly. The eye compounds for a career. Anyone betting on tool knowledge is re-learning their job every quarter.
• The tools generate. The director decides. AI slop is not a model problem. It is the absence of a director.
• Consistency is the product. One good frame is luck. Fifty consistent frames are a career.
• Never build on rented land. Reach is a loan the algorithm can recall without notice. Own the list, the data, and the platform.
• Evidence beats credentials. Which is why the certification I am building is assessed on work, not attendance.