Stop hiring experts. Sponsor a learner instead.
The highest-leverage talent play nobody's running right now.
A colleague of mine, Mike, posted in a group last week. He’d seen one of those AI-generated videos (the kind that actually looks good now) and wanted to know: who can make these for my business?
I answered something different.
Don’t hire the expert. Sponsor the learner.
Here’s what I told Mike, and what I’d tell any business owner watching AI reshape a skill they need.
In Summary, during technology transitions, founders should sponsor talented people with craft skills to learn new tools, rather than paying premium prices for “experts” whose tool knowledge is rapidly depreciating.
The Expert Premium Is a Mirage
When a new tool hits the market, something predictable happens. Demand for people who know the tool spikes. Supply is tiny. Prices go through the roof. Founders start hunting for “the expert.”
But here’s the thing about expertise in a brand-new tool category. It’s mostly a head start.
The person quoting you $5K to produce an AI video probably learned the tool six months ago. Maybe nine. Maybe yesterday! That’s not deep expertise built over a decade of practice. That’s someone who got there first. And the tools underneath them are shifting so fast that what they mastered in January might work differently by June.
Think about the pace. Twelve months ago, AI video was mostly short, artifact-heavy clips that looked like a student project. Six months ago, tools like DaVinci Resolve 20 shipped over 100 AI editing features. Then in the first week of February alone, two major models launched: Kling 3.0 (native 4K with synchronized audio) and Seedance 2.0 (multi-shot storytelling from a dozen input assets at once).
Six months. That’s how fast “AI generates a clip” became “AI generates a multi-shot video with synchronized dialogue and character consistency.”
The person who was an expert in August is re-learning their tools today. Right alongside everyone else.
This is what I call the expertise decay curve. In mature tool markets (Excel, Photoshop, WordPress), expertise compounds over decades. Every year you use the tool, you get more valuable. In emerging tool markets, expertise decays in months. The software updates faster than anyone can master it.
Hiring for expertise in a decaying market is buying a depreciating asset at peak price.
So what actually makes a great AI video? Not settings knowledge. Not prompt tricks. Taste. Editing instinct. Storytelling sense. Knowing when silence says more than dialogue. Knowing how to build tension through pacing. Recognizing the moment a scene needs to breathe.
These are craft skills. They transfer from traditional production, traditional design, traditional storytelling. And they don’t decay when the software updates. The tool knowledge is the easy part. The craft is the hard part.
The Play: Sponsor a Talented Learner
Here’s what I suggested Mike to do instead.
Find someone with taste. A video editor. A designer. A creator who understands pacing and narrative, but hasn’t gone deep on the new AI tools yet. Maybe they can’t justify the subscriptions. Maybe they’ve been meaning to explore but haven’t had a reason. Maybe they just need a project.
Make them an offer.
You cover their AI tool subscriptions ($200 to $500 a month for professional-tier access to the best models on the market). You fund a course or two if they want it. You give them a real project to work on: yours. In exchange, they produce your content on a professional contract at a rate that’s fair for someone building a new skill set.
Both sides win.
They get paid education, real-world practice, a growing portfolio, and a client who’s patient while they learn. You get emerging capability at 30 to 50 percent of what an “expert” would charge. You get a relationship with someone who’s growing into the role. And you get first-mover access to a skillset before their rates triple.
Bonus points if there is a way to layer in outcome based compensation as a bonus. That extra income could help them build their own business around their new skills.
The math is simple. That $5K-per-video expert? A talented editor you sponsor might cost $500 to $1,500 per video during the learning period, plus $200 a month in tools. Even if the first few aren’t perfect, you’re building a capability. Not renting one.
Note: These numbers are guesses, it’s about about the proportion between 100% expert to apprentice in training. It’s difficult to know what your real numbers would be but I’d think the 30-50% concept should hold up well.
You’re not doing charity. You’re running the highest-leverage talent play available right now.
This Isn’t Just About Video
The sponsorship model works anywhere the tools are new and moving fast. AI video is the example. The principle is universal.
AI-assisted design. AI content operations. AI sales and outreach. AI customer support. Every domain where the software is powerful but raw, and the craft skills still determine whether the output actually works.
We’ve seen this pattern before.
In the late 1990s, companies needed websites. There were no “web designers.” So agencies hired talented print designers and paid them to learn HTML.
The bet: aesthetic judgment transfers. Technical skills are learnable. The agencies that made that bet early dominated the first wave of the web.
When Apple opened the App Store in 2008, it launched with 500 apps. Companies that found talented developers early and sponsored their learning built category-defining products before the market professionalized.
The pattern repeats. New technology creates a capability gap. No formal training exists yet. The companies that sponsor talent early get loyalty and below-market rates, because the talent is grateful and building their portfolio. Then three to five years later, the market catches up. Training programs emerge. The talent pool normalizes. Rates equalize.
The window for this arbitrage is 18 to 24 months. Maybe less. AI video tutorials are already appearing. Certifications will follow. The person you could sponsor today for $500 a video will be charging $1,500 soon.
For a founder running a business under $1M annual revenue, this matters. You can’t afford the $10K-a-month video agency. But AI tools at $200 to $500 a month, paired with a talented person you’re investing in? That’s 10 to 20 times the output at a fraction of the cost. And you’re building a capability your competitors will be paying a premium for later.
Two Questions Before You Hire
Next time you’re about to hire for an “AI skill,” run it through two filters.
Is this person’s value in the tool, or in the craft? If their expertise is primarily in the software, that knowledge is depreciating. Fast. If their expertise is in the underlying craft (storytelling, design, strategy, editorial judgment), that transfers no matter what the tools do next.
Would a talented person with the right craft skills learn this tool in 30 days? If yes, sponsor the learner. Cover the tools and education, give them a real project, and let them grow into it. If no (genuinely deep technical work with real consequences for getting it wrong), hire the expert.
Most AI skill gaps fall into the first category. The tools are powerful and getting more intuitive every month. The bottleneck isn’t software knowledge. It’s taste, judgment, and craft.
The best founders don’t just buy capability. They grow it.