When Everyone Has the Same AI Tools, What Actually Becomes a Competitive Advantage?

A couple of years ago, simply knowing how to use AI well could give you a pretty meaningful advantage.

You could research faster. Produce more. Analyze information differently. Automate work that used to take hours. Take an idea from rough concept to something tangible much faster than you could before.

All of that is still true.

But there’s an obvious problem with calling AI itself a competitive advantage: your competitors have it too.

Their employees/agencies have ChatGPT, Claude, et al.. Their consultants are building custom tools and workflows. They can generate 50 campaign ideas before lunch, summarize thousands of customer reviews, build a presentation in an afternoon and produce content at a pace that would have been unthinkable a few years ago (quality be damned).

Point being- the tools are only getting better and more accessible.

So what happens when everyone has access to roughly the same superpower?

If anything, the “superpower” is now becoming an assumptive competency. So where is the real competitive advantage?

AI can give you an answer. That doesn’t mean it knows whether it’s a good one.

One of the most interesting things about generative AI is how incredibly good it has become at producing things that look “right” without attention or (gasp) scrutiny.

Ask it for a marketing strategy and you can get a…yes, arguably fine… strategy.

Ask it for customer personas, competitive analysis, campaign concepts, social content, SEO recommendations or a product roadmap and you’ll probably get something that is logical, organized and, at least on the surface, pretty impressive.

But did you read it? And did you pay attention to what it said, and did you think about it with the specialist IQ your industry/client/topical area deserves?

Because something can be polished, logical and completely miss a detail that anyone who actually understands the business would catch immediately.

Artificial fluency is getting cheaper. Vertical expertise/judgment is absofuckinglutely disappearing unless we actively intervene.

In other words, the more I work with AI, the more I think the value of specialized knowledge may actually increase rather than decrease.

Your customers know when you don’t know their world

I work quite a bit in the cruise industry, and cruise is a great example of this.

People who cruise know ships.

They know cruise lines. They know what different ships look like. They know the experiences associated with them. For a frequent cruiser, the ship isn’t generic transportation sitting behind the vacation. It’s part of the product.

That matters when you’re choosing creative.

AI can generate a spectacular image of a cruise ship in seconds. Blue water, perfect sunset, rows of balconies, dramatic lighting. It can look incredible.

Except the ship doesn’t exist.

Maybe the architecture doesn’t quite make sense. Maybe it looks vaguely like several cruise lines combined into one. Maybe there are physical details that someone who knows ships would immediately recognize as wrong.

AI sees a cruise ship.

The customer sees the wrong cruise ship (or blatant AI).

And this problem existed well before generative AI. Stock libraries have always presented versions of it. AI just makes it possible to create inaccurate content much faster and much more convincingly.

I saw the same thing in a completely different industry earlier in my career at an agency with a trucking insurance client.

Truckers know trucks.

They notice whether the equipment is modern. They notice the type of truck, including the absolutely nuanced specific grills - yes, really. They notice whether the environment looks like somewhere they would actually work. They notice whether the people being portrayed look like people they recognize from their own industry.

Someone outside that world might look at an image and think, “Great, it’s a truck.”

The audience may see something very different.

There’s something bigger going on there than factual accuracy.

People want to see themselves reflected in the brands trying to reach them. They want some indication that you understand their industry, their interests and the things they care about.

Getting those details right communicates something without having to say it:

I/we know this world.

Getting them consistently wrong communicates something too.

Expertise is knowing which details aren't interchangeable

The same thing happens throughout travel.

You can create a beautiful airport scene that makes absolutely no sense to someone who spends a lot of time in airports.

Maybe people are saying goodbye somewhere they couldn't actually be together because one of them would need to have gone through security. Maybe baggage is appearing somewhere it shouldn't. Maybe the airport, aircraft or travel experience doesn't work the way the creative implies it does.

Does one inaccurate detail destroy an entire campaign? meh - probably not.

But these things are tells.

And the closer your audience is to the category, the more likely they are to notice the details a generalist might assume are interchangeable.

Cruisers notice ships.

Truckers notice trucks.

Frequent travelers notice airports.

Runners notice shoes, courses and race experiences.

People inside an industry notice the things that people outside it don't even know to look for.

That’s part of what expertise actually is.

It isn't just knowing more facts about a subject. AI is increasingly excellent at facts.

Expertise is understanding which distinctions matter.

It’s knowing what customers say they care about versus what actually influences their decisions. It’s knowing the competitive landscape and which supposed differentiators aren’t actually differentiated at all. It’s knowing what has already been tried. It’s understanding the relationships and people that shape an industry. It’s knowing what sounds natural and what immediately sounds like it was written by someone who has never spent much time around the people they’re trying to reach.

And sometimes, expertise is simply knowing when the normal pattern doesn't apply.

AI is only as smart as the context surrounding it

A lot of the conversation about AI still focuses on the tool.

Which model are you using?

ChatGPT or Claude?

Which version?

What prompt did you use?

Those questions aren't irrelevant. But I suspect they’re going to become less interesting as the underlying technology continues to improve and access to powerful models becomes increasingly widespread.

The more interesting questions are going to be:

What do you know that your competitors don't?

What proprietary information can you bring to the process?

How well do you understand your customers?

Do you know which questions are actually worth asking?

Can you recognize when an answer that sounds completely reasonable is wrong?

Can you tell the AI, “Technically, yes. But that’s not how this industry actually works”?

Because that last piece matters.

Someone with specialized knowledge doesn't just have more information to give an AI system. They ask different questions in the first place.

They know what context is missing.

They know what needs to be challenged.

They know when the first answer isn't good enough.

And they can connect the output to things the model couldn't possibly know on its own.

Maybe companies need fewer “AI people” and more experts who are great at AI

There’s been an understandable rush over the last few years to find people who “know AI.”

And yes, you absolutely need people who understand the technology and know how to use it.

But I think the framing deserves some scrutiny.

The question isn't only:

Who knows AI?

It should also be:

Who deeply understands our business and knows how to use AI to make that expertise more powerful?

Those are not necessarily the same person.

Someone who understands your product, customers, competitive landscape, industry and brand can use AI to extend that knowledge dramatically.

Someone who understands AI extremely well but knows very little about any of those things can produce a tremendous volume of impressive-looking work that isn't particularly useful.

And this matters when companies hire employees. It matters when they choose agencies. It matters when they hire consultants and vendors.

AI can help someone learn an industry much faster than they could before. That's a very good thing. I use it that way constantly.

But there is still a difference between knowing information about an industry and understanding how that industry actually works.

Companies shouldn't assume specialized expertise matters less because AI can fill knowledge gaps.

It may matter more because AI makes superficial competence so much easier to produce.

Knowing the industry isn't enough. You also have to know the brand.

There’s another layer to this that gets overlooked.

You can understand an industry extremely well and still not understand a particular company.

Every organization accumulates institutional knowledge that isn't sitting neatly on a website waiting for an AI model to find it.

Why did we stop saying that?

Why doesn't that offer work?

Why does this customer segment behave differently?

What happened when we tried this three years ago?

Why isn't that competitor actually comparable to us?

Why does something that sounds completely logical on paper not work operationally?

Why does the customer service team know something about the customer that the marketing team doesn't?

Some of the most valuable people inside and around an organization are the ones who can connect those layers: the industry, the customer, the product, the brand and the realities of the organization itself.

AI can make those people dramatically more capable.

It doesn't make their context irrelevant.

AI may be raising the floor faster than it raises the ceiling

Before generative AI, producing something decent required a certain amount of skill.

Now decent is increasingly cheap.

Anyone can create a plausible article. A competent presentation. A reasonable campaign concept. A polished strategy document.

That's an incredible democratization of capability, and overall I think it's a good thing.

But it also makes it harder to distinguish between work that looks competent and work that is actually insightful.

Which means I find myself paying more attention to different things.

Who asks the question nobody else asked?

Who catches the assumption that's wrong?

Who knows that something has already been tried?

Who understands why the customer is behaving differently than the data initially suggests?

Who looks at the beautiful AI-generated cruise ship and immediately says, “We can't use that”?

Who can take what AI produces and make it better because they understand something the machine doesn't?

AI is raising the floor incredibly quickly.

I'm not convinced it's raising the ceiling equally for everyone.

So what actually becomes the competitive advantage?

I don't think there's one answer.

Specialized knowledge matters.

Proprietary information matters.

Customer understanding matters.

Judgment matters.

Relationships matter.

Brand matters.

And the ability to actually execute still matters quite a lot.

AI can multiply all of those things.

That's what makes the technology so powerful.

But if two companies have access to increasingly similar technology, the technology itself can't be the entire advantage.

What you bring to it becomes the differentiator.

Maybe the companies that ultimately get the most out of AI won't be the ones that know the most about AI.

They'll be the ones that know the most about their business, their industry and their customers - and figure out how to make AI smarter with what they know.