For the past few years, the conversation around AI has mostly been about capability. How powerful is the model. What it can do now that it couldn’t do before. Whether it’s getting smarter fast enough.

That conversation made sense when access was limited and technical skill was the bottleneck.

But something has shifted. And I’ve been trying to put words to it.

01 — The Barrier Has Moved

Not long ago, building something real required a specific kind of knowledge. You needed to understand systems, write code, or hire people who could. That created a natural filter — not everyone could participate, so the people who could had an advantage just from being able to show up.

AI has largely removed that filter.

You can now create content, prototype products, design workflows, and communicate ideas at a level that would have taken years to develop, or thousands of dollars to outsource. The ability to build is no longer scarce in the way it used to be.

Which means the gap has moved somewhere else.

02 — The New Gap Is Resources

When a skill becomes common, the advantage shifts to whoever has better inputs.

Better information, earlier. Access to the right people. Proximity to opportunities before they become obvious. These are things that don’t scale the same way AI capability does — they accumulate slowly, through relationships and experience and being in the right rooms over time.

I’ve been noticing this in my own work. I’ll come across an idea — someone sharing how they’re using a new approach to drive traffic, or a framework that’s actually working — and I’ll think about how different people respond to the same information. Some consume it. Some experiment with it. A small number turn it into something real.

The difference usually isn’t skill anymore. It’s whether someone has the context and resources to translate information into action. That’s harder to shortcut.

03 — Trust Is the Hardest Layer

But even resources aren’t the real bottleneck. The hardest thing to build right now is trust.

Here’s the uncomfortable part: AI makes almost anyone look competent. The writing is cleaner. The ideas are more structured. The output is more polished. Which means the surface-level signals that people used to rely on to evaluate credibility are now much easier to manufacture.

When everyone looks competent, trust becomes scarcer — and more valuable.

After I shared an early version of this thinking on Twitter, someone replied with something that stuck with me. They were on day 10 of running a product with 7 AI agents. Everything was working — the pipeline was moving, the product was real. But the sales weren’t coming. Their read on why: “I want this but I don’t know who you are is where every conversation stalls. No shortcut to that one.”

That’s it exactly. The bottleneck isn’t capability anymore. It’s whether someone trusts you enough to take the next step.

I’ve been sharing my actual workflows and thinking on this site, including things that didn’t work and decisions I’m still uncertain about. And I’ve noticed something strange: content grounded in real experience isn’t always easier to trust than content that sounds simpler and more confident. People are often drawn to the cleaner narrative. The faster promise.

That’s not a criticism — I do it too. But it means trust isn’t something you earn automatically just by being honest. You have to earn it in ways that are visible, consistent, and hard to fake.

04 — What AI Can’t Do

AI can make you faster, more articulate, and more productive. It can help you punch above your weight class in almost every dimension of output.

But it can’t build trust on your behalf.

Trust comes from what you’ve actually done. What you’ve failed at. How you think when things are unclear. Whether you follow through. These are things that have to accumulate through real experience, and they can’t be generated by a model.

That’s why two people using the same tools can produce very different results. The tools don’t determine the outcome — the judgment, relationships, and track record behind them do.

05 — What I’m Actually Trying to Build

I used to think about AI mostly in terms of tools and efficiency. Which model to use, which workflow to optimize, how to do more in less time.

I still care about that. But the question I’m more interested in now is whether I’m building something that compounds.

Not just content. Not just output. But actual resources — a body of work that means something, relationships with people who trust my judgment, a way of thinking that gets sharper over time.

AI accelerates a lot of things. But it accelerates what’s already there. The underlying asset still has to be built the slow way.

The Real Question

The barrier to entry has dropped. That’s real, and it matters.

But what creates durable advantage now isn’t access to tools. It’s the resources and trust that tools can’t replace — and that take genuine time and experience to build.

So the question is no longer whether you know how to use AI.

It’s whether you’re building something that AI alone couldn’t have made.