The New Bottleneck Is the Human

Yesterday, I was on a call with a group of men, and one conversation stayed with me. We were talking about how quickly people are starting to build things with AI.

Vibe coding has changed the feeling of what is possible. You can have an idea in the morning and have something working by the evening. You can build a prototype without knowing how to code. You can test an idea without hiring a developer or raising money first.

This is an incredible shift. But I have started to notice a strange side effect.

The easier it becomes to build, the easier it becomes to exhaust yourself.

I see this in founders I work with. They can now move incredibly fast, but many of them are also becoming overwhelmed incredibly fast.

They build at night. They change the product every day. They try another idea. They add another feature. They ask AI to solve another problem. They open another tool.

The distance between an idea and a working product has become very small. The distance between an idea and another unfinished project has also become very small.

And I think this points to a much bigger change in how we build companies.

For a long time, the biggest constraints in building a product were external. You needed engineers to write code. You needed designers to create interfaces. You needed money to build the first version. You needed infrastructure, distribution, and access to specialized knowledge.

The tools were the bottleneck.

Today, many of these constraints are disappearing. AI can write code. AI can create designs. AI can analyze data. AI can help with marketing, research, customer support, and operations. A single person can now do what previously required a small team.

But removing one constraint does not remove constraints altogether. It simply moves the bottleneck.

And increasingly, I believe the bottleneck is becoming the person.

Not because people are becoming less capable. Because the number of things they can now do has grown faster than their ability to decide what they should do.

This creates a new kind of problem.

A founder can build almost anything, but does not necessarily know what to build. They can launch a product, but may not understand why customers are not using it. They can generate hundreds of ideas, but have no system for deciding which one deserves attention. They can create a company without a large team, but suddenly discover that they do not understand finance, hiring, legal, sales, or operations.

The problem is no longer only execution. The problem is the ability to navigate complexity.

I see this often when I help founders.

If they move fast with AI, they can quickly become exhausted. There is always something else to improve, another problem to fix, another possibility to explore.

If they slow down, they encounter a different set of constraints. They may not know how to structure the finances. They may be afraid to hire the first employee. They may not know how to choose a co-founder or a contractor. They may struggle to understand distribution or positioning. They may have a product, but no clear strategy for what happens next.

So the founder becomes the weakest link in the system.

This is not a criticism. It is a natural consequence of the world we are entering.

When technology becomes more powerful, the human part of the system becomes more important. The ability to build is becoming more accessible. The ability to choose is becoming more valuable.

This is why I think the next generation of founders will need to develop a different kind of skill set.

They will need to know how to use AI, but also when not to use it. They will need to move fast, but also know when to stop. They will need to generate ideas, but become much better at killing the ones that should not be built. They will need to learn how to work with uncertainty without turning every uncertainty into another project.

And perhaps most importantly, they will need to build systems around themselves. Systems for thinking. Systems for making decisions. Systems for prioritization. Systems for learning. Systems for managing their own energy.

Because when one person can now perform the work of five or ten people, the question becomes: can that person also carry the cognitive load of five or ten people?

This is where I think we are going to see a new kind of founder challenge.

The old question was often: "How do I get enough resources to build this?" The new question may become: "How do I manage the resources I now have without losing myself in the process?"

AI can help you build faster. But it cannot decide what is worth building.

It can give you ten solutions. It cannot tell you which problem matters most. It can help you launch a product. It cannot give you a reason to keep going when the product is not working. It can reduce the cost of execution. But it cannot remove the emotional cost of making difficult decisions.

This is why I increasingly believe that the most important skill for founders in the AI era may be clarity.

Clarity about the problem. Clarity about the customer. Clarity about what matters now. Clarity about what can wait. Clarity about what to stop doing.

The irony is that the more powerful our tools become, the more important it becomes to know ourselves.

Our attention becomes a scarce resource. Our energy becomes a scarce resource. Our ability to make good decisions becomes a scarce resource.

And perhaps this is the real shift happening right now.

We spent decades trying to make machines more capable. Now they are becoming capable enough that the limiting factor is increasingly the human directing them.

The next competitive advantage may not belong to the person who can build the fastest. It may belong to the person who can think clearly enough to know what deserves to be built in the first place.