AI Search, Discoverability, and the Cost of Shortcuts

Not sure about you, but I’ve seen considerable hype lately on my feed, at conferences, and in conversations with executives about AI search.

That interest makes sense. Discoverability has always been one of the hardest problems to solve well.

Every era has had its version of it.

In the early 1700s, discoverability meant having the corner store at the busiest intersection in town. As literacy and printing technology expanded, newspapers, especially their front pages, became the place to be. In the early 1900s, radio and, later, television ushered in a shift toward mass media, in which prime-time exposure mattered more than almost anything else.

Those systems shared a common trait: they favored scale and averages. When access was expensive, only large entities could participate. And when you’re trying to reach as many people as possible, the safest strategy is usually to build something that works well enough for the middle.

That approach isn’t wrong. Average is often essential. But it also limits what gets surfaced.

The Internet, SEO, and the Opening of the Long Tail

The early internet changed that dynamic.

At the beginning of SEO, discoverability didn’t require massive capital. It depended largely on links. The logic was straightforward: if many people took the time to link to something, it was probably useful or important.

In those early days, good ideas often won because curious people on the long tail were actively exploring and sharing what worked. It was chaotic at times, but the system rewarded things that genuinely delivered value. If something helped you, you were happy to tell others about it.

As SEO matured, people recognized this. With enough money, you could pay to manufacture links to drive rankings, optimizing for discoverability rather than credibility built through real experience.

Over time, SEO shifted back toward rewarding scale and averages, rather than the long tail it initially enabled.

The Shift to AI Search and What It Means for the Long Tail

Today, discoverability remains difficult. What’s changing is where the search happens.

More people are moving from traditional browser-based search to language models for answers. In response, many companies are treating AI search the way they treated SEO: as something to be optimized.

The result is a new wave of “AI search optimization,” often built around mass-producing AI-generated, AI-digestible content in hopes of being surfaced more often.

In the short term, this can work. Especially in categories where information is sparse or competition is still low, volume and structure can create discoverability.

But short-term discoverability is not the same thing as long-term trust.

Why “AI-Built” Content Already Feels Wrong

AI systems are built on advanced statistical modeling. By design, they tend to gravitate toward the average.

That’s not inherently bad. Averages are useful. They make systems reliable at scale.

But when you rely entirely on self-generated content optimized for discoverability, something subtle happens: everything starts to sound the same. Intent flattens. Voice disappears. And people notice.

When someone says, “This feels AI-written,” they’re not making a stylistic critique. They’re recognizing a lack of specificity, care, or point of view. They’re responding to content that was generated to exist, not to help.

Average works for infrastructure. It does not work for trust, traction, or building something meaningfully better.

Where “AI-Built” Content Breaks

There are a few structural reasons this strategy degrades over time.

First, AI systems will evolve faster than SEO ever did. As skepticism grows, trust will matter more, not less. Models that infer intent will improve at distinguishing between content created to be useful and content created to extract attention.

Second, data doesn’t reset cleanly. Even if content is deleted, historical data (e.g., the Wayback Machine), embeddings, and provenance still shape how systems interpret behavior over time. You can’t indefinitely optimize for attention without leaving a trail.

And third, personalization changes the game entirely.

AI enables increasingly individualized experiences. Humans are not average; by our very nature, we are unique neural networks. As systems learn preferences, values, and context, being broadly “good enough” matters less than being specifically useful.

This reopens the long tail, but only for those building with intent.

The Long Tail Isn’t for Everyone — and That’s the Point

As personalization increases, the goal doesn’t need to be building for everyone. It can be building deeply for someone.

That requires having a point of view. It means accepting that some people will say, “This isn’t for me.” And that’s okay.

Because the people it is for will care. They’ll trust it. And they’ll come back.

For many individuals and companies, serving a small, committed audience, even a thousand people or fewer, can be enough to create meaningful, long-term value. That’s a very different strategy from chasing short-term discoverability at scale.

How I Use AI (and Why It Works for Me)

None of this is an argument against AI. Quite the opposite.

AI has been one of the biggest productivity gifts I’ve ever had.

My thinking is naturally fragmented. I tend to compress too many ideas into dense, nonlinear drafts. AI has been an incredible editor, helping me structure thoughts, test clarity, and compress weeks of iteration into hours.

It’s also a powerful brainstorming partner. As it learns your intent, it can challenge assumptions, offer alternate angles, and help you explore ideas more broadly.

For well-defined, linear problems (e.g., coding, automation, analysis), it dramatically lowers the barrier to building what would have previously taken months.

The difference is intent.

AI is most powerful when it amplifies thinking, not when it replaces it.

The Cost of Shortcuts

Using AI to mass-generate content purely for discoverability is tempting. It can even work in the short term.

But over time, it comes at a cost.

It trains systems and people to distrust you.

My advice is simple:

Build for something and someone.
Care deeply about their experience.
Treat trust as your most valuable asset. Because it is.

Use AI as a tool to support that work: as an editor, a thought partner, and a task agent. Your care will show up in the work. The trust you build will show up too.

Earned trust still beats volume.
Even when the audience is a machine.