I have a confession.
I spend a ridiculous amount of time experimenting with AI, and there are still days when I feel like I’m behind.
Every week there’s a new model, a new agent, a new workflow, or some guy on LinkedIn explaining how he automated his entire Marketing department before breakfast.
The worst part is figuring out where to start.
Should you get better at prompting?
Learn how to build your own agents?
Figure out how to connect AI to your Marketing data?
Start using OpenClaw??
There are about 1,000 directions you could go.
We know the opportunity is there: according to PwC’s 2026 Global AI Jobs Barometer, jobs requiring specific AI skills carry an average 62% wage premium over comparable jobs without them.
Meanwhile, Gallup’s 2026 research found that 52% of U.S. employees now use AI at work, but only 15% use it daily.
Translation: most of us are still figuring this out. And there’s a payoff for the Marketers who get there first.
I recently had a conversation with Ged King, an ad agency partner and Academic Director of the Digital Marketing and AI program at Wake Forest University.
Ged teaches in the Online Master’s in Digital Marketing and AI at the Wake Forest School of Professional Studies, and I wanted to know: how do they run AI education when things are changing so quickly?
The answer: they change, too.
His faculty don’t recycle the same Marketing material year after year. They’re working Marketers who build & test AI systems at their day jobs, and bring what they learn back to class.
When I asked Ged what he’d be teaching in February, he told me he genuinely didn’t know yet. On purpose.
The structure is set. The tools and examples aren’t.
I loved that answer. If you’re teaching AI right now, your curriculum SHOULD be changing that fast.
So I put together something inspired by our conversation: Welcome to your unofficial AI Marketing syllabus for 2027.
Ahead: 5 skills to practice if you want to get past the basics, with homework, plus a scorecard you can steal. And at the end, my recommendation if you REALLY want to take your skills to the next level.
Lesson 1️⃣: Figure Out Where You Stand.
I worry a lot of Marketers overestimate their AI skills.
I get why. If you’re using ChatGPT every day to write content, brainstorm campaigns, and analyze documents, it probably feels like you’re pretty comfortable with it.
Ged’s AI Leverage Model might make you reevaluate.
It breaks AI adoption into 5 stages:
- Exposure: You have access to AI and understand the basics.
- Usage: You’re using AI for individual tasks like writing copy or summarizing research.
- Routine: You’ve built repeatable AI processes into your work.
- Integration: You’ve redesigned workflows around AI, your data, and automation.
- Delegation: AI agents execute multi-step tasks on their own, while humans make the important calls.
⚠️ NOTE: the goal isn’t to reach Delegation on everything. Some work should always have a human making the final call. But it’s a great framework to assess how effectively you’re using AI across all your work.
Ged pairs it with another framework, the AI Leverage Index, which asks whether AI is improving your work…or just making MORE of it. (Raise your hand if you’ve been victimized by Workslop recently.)
I mean, if your team went from 10 social posts a week to 50, but engagement and pipeline haven’t budged, what did you really gain?
📚 Your homework:
→ Start with recurring work your team does. Competitive research, campaign reporting, customer analysis, and creative development are all good places to look.
→ Rate each workflow at 1 of the 5 stages.
→ Pick 1 workflow worth moving up a stage. That’s your target for the rest of this syllabus.
Lesson 2️⃣: Learn How to Make AI Evaluate Your Marketing.
Ged built an AI system that grades advertising creative BEFORE a campaign launches. No big deal.
This system combines a large language model with Jev, a brand-new model from a startup called TypeSafe AI. It recently opened in early access, so if you haven’t heard of it, you’re not alone.
It’s built for 1 job: give it text and a fixed question (yes/no, multiple choice, or a 1 to 10 score), and in milliseconds it returns an answer, plus how confident it is.
That kind of fast, single-question judgment makes it a cheap way to screen 100s of ads at a time.
Ged’s system has the LLM gather competitor ads and describe each 1, with brand names removed. Jev scores every description against the 1 to 10 scorecard. Then ads go head-to-head, blind 1st, then with the names showing. (I love the blind test, since a famous name can make a weak ad look better than it is.)
Ged’s team ran it on Wake Forest’s own advertising. Their original creative scored 6.4 out of 10, with the best competing university scoring a 6.8.
So they rewrote and retested, based on the insights, until the work met their standards.
The new concepts scored 8.0, 7.8, and 8.1 out of 10, and started winning head-to-heads. The blind test taught them prospects picked based on proof points they could check, like price, time to finish, and outcomes. So, the winning concepts put Wake Forest’s proof up front, like capstones from real clients.
As a gut check before you spend a dollar, this process is hard to beat.
📚 Your homework: Steal Ged’s creative scorecard.
→ Pull 5 competitor ads, plus your own.
→ Strip out every brand name and logo.
→ Ask AI to rate each ad from 1 to 10 on Ged’s 7 dimensions, weighted:
- Emotion (20% weight): How strongly does it make you feel something?
- Attention (15%): How fast does it stop the scroll?
- Brand link (15%): How easy is it to tell whose ad it is with the logo covered?
- Clarity (15%): How obvious is the offer on the 1st read?
- Motivation (15%): How strong is the reason to act now?
- Audience fit (10%): How clearly is it speaking to 1 specific person?
- Distinctiveness (10%): How hard would it be for a competitor to run it unchanged?
The overall score = the weighted average.
→ No Jev access? Any LLM works for a 1st pass, as long as the questions stay fixed for every test.
→ Compare ads in pairs, twice each, swapping the order, so position can’t tip the result.
→ Use Ged’s “win bar” for minimum-acceptable scores: 7.3+ overall, 65%+ win rate of blind head-to-heads, no dimension under 5, and 7.5+ brand fit.
→ Rewrite your lowest-scoring dimension 1st. Keep a change only if the score goes up.
📥 Want the full method? Ged’s white paper, Fast Judgment, Slow Proof, walks through the whole pipeline, plus a 30-day pilot you can run with your own team.
Lesson 3️⃣: Understand the Fundamentals Behind AI.
My favorite part of Ged’s research was when he disagreed with the model.
His team ran a landing page through the same scoring process. After 5 rounds of revisions, it climbed from 4.1 to 7.3 out of 10, and started beating 2 established competitor pages head-to-head.
But…then Ged looked at it.
It was too busy.
The model was scoring a written description of the page. It couldn’t see the clutter. Ged could. So he simplified it. The simpler version scored a little lower, but experience told him it was the 1 to run.
AI can tell you what scores well. Knowing what works for YOUR customers still takes Marketing fundamentals.
Consumer psychology, positioning, analytics, creative strategy, how the business makes money: all these help you both ask better questions, and evaluate whether the answers are any good.
That balance is built into how Wake Forest teaches. Their AI education sits right next to the fundamental skills you need to use the tools responsibly.
Lesson 4️⃣: Learn How to Measure Whether AI Is Actually Making You Better.
Are you measuring how much AI gets used? Or what AI makes possible?
Ged’s AI Leverage Playbook suggests measuring a recurring workflow’s time, quality, direct cost, and business impact before you add AI.
Say AI shrinks the time it takes for your monthly competitive intelligence report from 8 hours to 3. Nice.
But before you high five, ask 2 questions:
- Is the report as good as (or better than) the old 1?
- What are you doing with those 5 new hours?
A faster report with weaker insights = a downgrade. And 5 hours that vanish into more meetings won’t show up anywhere.
Ged’s lean: put the time back into the work. More customer interviews. More experiments. More things AI can’t do.
📚Your homework:
→ Grab the workflow you picked in Lesson 1.
→ Before changing anything, write down how long it takes, how many people touch it, and what a good result looks like.
→ Decide what your team will do with the time you save. Don’t skip!
→ Redesign it, execute, then compare.
Lesson 5️⃣: Build an AI Learning Habit That Doesn’t Depend on the Latest Tool.
This is the biggie: how do you invest time and money into learning AI when the technology might look completely different a year from now?
I asked Ged. His answer: become a lifelong learner.
I’ve built a career around constantly learning about Marketing. It’s a big reason I started The Marketing Millennials in the first place.
But the pace of AI has made the learning curve feel steeper than ever, and my ADHD brain needs structure. Otherwise, I’ll burn an afternoon experimenting with cool new tools without necessarily figuring out how they apply to my actual work.
More points to Wake Forest: Ged has his students build portfolios that put last week’s and last month’s work right next to their best projects. It’s proof they keep learning.
Imagine going into a job interview being able to show off an AI-assisted research workflow, creative testing process, or campaign reporting that YOU built. Says a lot more than adding “ChatGPT” under your technical skills bullet points.
And the companies hiring are paying attention. Wake Forest’s February certificate program is bringing in speakers from Coca-Cola, Salesforce, and NASCAR to talk about how AI is working inside their organizations. (When I said real Marketers with serious day jobs, I wasn’t kidding.)
📚 Your homework: start your portfolio.
→ Pick 1 meaningful Marketing problem each month.
→ Experiment with how AI could help solve it.
→ Document what you built, what worked, and what you’d change.
Your AI Marketing Syllabus Will Be Different in 6 Months. That’s the Point.
Remember Ged not knowing what he’ll teach in February, and bringing in working Marketers? All this is why.
Everything above is your syllabus for right now. But the tools will keep changing. Your ability to size up a Marketing problem and judge what new technology can do for it is what will carry over.
And with the homework above, you can start building that habit yourself.
But if you learn best with structure (hi, it’s me), and you want faculty who are building this stuff inside real companies right now, Wake Forest’s online master’s is built by and for working Marketers, and its courses get rewritten every time they’re taught.
Marketers who keep learning and adapting have always had an advantage. AI is supercharging that advantage.
The best thing you can do for your Marketing career right now is get really good at learning what’s next.
👉 Learn more about Wake Forest’s Online Master of Digital Marketing and AI program.
