1,000+ Hours of AI. Eight Things Nobody Told Me.

Image inspired by Dragon's Lair, animated by Don Bluth.

When I was a kid, my favorite arcade game was Dragon's Lair. You played Dirk the Daring, a knight navigating a castle full of traps and monsters, and the whole point was that every wrong move killed you. There was no manual. No tutorial. You learned entirely by dying, getting back up, adding countless more quarters to the machine, and trying again and again until the path through finally made sense. Learning AI in destination marketing has felt remarkably similar.

Almost every destination marketing professional working today has tried AI tools. More than half use them weekly. But only 16 percent of their organizations have any kind of strategy for it.

That gap is probably the most honest picture of where this industry stands right now. And I think it explains a lot about why so many people feel like they're spinning their wheels.

I spent 16 years working inside destination marketing. Since then I've been pretty absorbed in figuring out where AI actually fits for this industry and where it's mostly just adding noise.

These are the lessons I keep coming back to.

1. The confusion at the beginning isn't a sign it's not for you.

There's a stretch in early AI adoption that nobody describes honestly enough.

You try a few things, the results feel generic or slightly off, and you start wondering if the productivity gains everyone keeps talking about are real or just content. That stretch is normal. Probably necessary, actually. What I've noticed is that clarity doesn't come from more research or better tool selection. It comes from using AI on real work long enough that the pieces start connecting.

That takes longer than most people expect, and longer than most LinkedIn posts would have you believe. If you're in the fog right now, stay with it a little longer before deciding it's not for you.

2. No course will teach you what actually doing the work will.

Training has its place. But there's a real gap between understanding AI conceptually and developing genuine judgment about when it's working and when it isn't. The only thing that closes that gap is using it on actual tasks with real stakes attached.

Not testing it out of curiosity. Actually trying to produce something you need.

A press trip brief, a board summary draft, a batch of social captions for an upcoming campaign (that last one is where many people start, and honestly it's a fine place to start).

That shift from "I'm exploring this" to "I need this to actually be good" is where the real learning happens.

3. You're the expert. AI is the instrument. That distinction matters more than most people realize.

When AI is working well in someone's hands, they're still the one driving. They're applying judgment, reading outputs critically, pushing back when something's off, and making the final call. The AI makes them faster. It doesn't replace the thinking.

When it's not working well, the person hands over the wheel. They accept the first output without really reading it, skip the part where they evaluate it against what they actually know, and end up with something that technically answers the question but doesn't really sound like their organization.

McKinsey published a workplace AI report earlier this year that honestly stopped me when I read it. Only one percent of organizations say they're genuinely mature in AI deployment, meaning it's actually woven into how they work and producing real outcomes.

One percent!

The gap between using AI and using it well is much wider than most people assume, and how engaged you stay in the process explains most of it.

4. Your destination knowledge is what makes the output good.

This is probably the one I'd most want people in this industry to hear, especially those quietly worried that deep expertise in a place, a market, a visitor profile, is becoming less valuable.

It's the opposite.

What separates a genuinely useful AI output from a forgettable one in destination marketing is almost always the quality of context going in. Your understanding of your destination's distinct character, your visitor motivations, your brand voice, your stakeholder relationships, that's what turns a generic response into something you'd actually use. Someone without that context gets something that could be about anywhere.

AI doesn't replace that knowledge. It runs on it.

5. When good people leave, don't let their AI work leave with them.

Staff turnover has always been a reality in this industry. What's new is that it now carries an AI dimension most organizations haven't thought through yet.

If someone on your team has spent months developing prompts and approaches that genuinely work well, and none of that is documented anywhere, it walks out the door with them. The next person starts from zero. That same survey finding that 99 percent of DMO professionals have tried AI also found only 16 percent of organizations have a strategy in place. Which means the overwhelming majority of what's being built right now lives entirely in individual people's habits, not in shared systems anyone else can access or build on.

The teams that build something durable with AI probably won't just be the ones who adopt it fastest. They'll be the ones who treat what they build as organizational knowledge, document it, share it, and make sure it doesn't disappear with the next transition.

6. Pick a small stack of tools and actually get good at them.

New AI tools get announced constantly, and the temptation to keep sampling them is real and mostly counterproductive. The teams getting the best results aren't using the most tools. They're using a focused set consistently enough that they've developed real judgment about when those tools work well and when they don't. That kind of fluency takes time to build, and it resets every time you start over with something new.

You don't need to know every tool on the market. You need to be genuinely capable with the right few. (That's harder than it sounds when something new drops every other week, but it's the only approach I've seen actually work.)

7. How you prompt matters more than which tool you're using.

Two organizations can have access to the exact same AI tools and get completely different results. The variable that explains most of that gap is usually prompt quality.

A vague input gets a vague output.

A prompt that includes real context, your audience, your tone, your organization's specific situation, and exactly what you're trying to produce gets something you can actually use. This is a learnable skill, and it's probably the highest-return investment a DMO team can make right now, because it applies regardless of which tools you're working with.

8. The honest case for this isn't fear. It's capacity.

I'm not really interested in the "learn AI or become irrelevant" framing. It's not that it's wrong, it's that it's not useful. Fear doesn't actually help people learn anything.

Here's what I think is more useful. The DMO teams I've talked with who are getting real results aren't doing it because they're scared. They're doing it because they're genuinely getting time back. Not in theory. In actual hours per week they're redirecting toward the work that actually requires them.

That's the real argument. Not what happens if you don't, but what becomes possible when you do.

The thing about Dragon's Lair was that dying wasn't failure. It was the tutorial. Every wrong move taught you something the previous run didn't. The teams getting real results with AI right now aren't the ones who avoided the hard part. They're the ones who went through it. -Jason

We've built AI tools specifically for destination marketing teams at Swix AI. If any of this resonates and you want to see what it looks like in practice, I'm happy to show you.