How to know if a problem is big enough to build a startup in the AI era

Building software has never been easier.

New artificial intelligence tools allow founders to generate entire applications from a simple prompt. Platforms such as Lovable, Replit and Cursor are driving a new development approach often referred to as vibe coding.

At Nomu we are seeing this shift firsthand. Through hackathons and events we host with Lovable, such as BETA DASH, teams are able to build functional prototypes in just a few hours.

This dramatically lowers the technical barrier to building software.

But it also introduces a new risk for founders and product teams. When building becomes extremely easy, many startups begin with the product instead of starting with the problem.

That reversal is often the reason many projects never become real companies.

The impact of vibe coding on startup creation

The rise of vibe coding is democratizing software creation. Today, founders can launch applications without large technical teams.

This shift is positive. It allows teams to validate ideas faster, experiment with new products and reduce the cost of launching an MVP.

However, it is also creating a new pattern.

Many people are building platforms simply because they now can.

Technology is no longer the main constraint. The real challenge becomes identifying a problem important enough to build a company around.

Problem size vs market size

One of the most common startup mistakes is focusing first on market size.

A large market does not guarantee a real business opportunity.

The key question is different.

Is the problem you are solving important enough within that market?

AI tools now allow people to build platforms for almost any niche imaginable. But if the problem you solve is not a priority for users, traction will be very difficult.

A small problem inside a large market is still a small problem.

When building a startup, the intensity of the problem matters far more than the size of the market.

The intensity of user pain

Successful startups usually solve problems that create clear frustration for users.

There is a big difference between a product that would be nice to have and a product people truly need.

In many hackathons we see technically impressive ideas that emerge directly from what the technology makes possible.

The strongest startups work differently.

First there is a clear problem. Then technology is used to solve it.

When a problem is real, clear signals appear:

• Users are already using imperfect solutions

• They are losing time or money trying to solve it

• There is constant frustration around the process

  
  

If no one is trying to solve the problem today, it is probably not important enough to build a company around it.

Real signals of urgency

One of the best ways to identify a meaningful problem is to observe urgency.

When a problem is strong, users actively search for solutions. They pay for incomplete tools or adapt internal processes to deal with it.

When a product appears and removes that pain, adoption happens naturally.

Many applications created with AI today follow the opposite logic. The tool is built first and only later people try to figure out what it could be used for.

This often produces interesting products, but rarely sustainable companies.

The real challenge of building startups with AI

For many years the main barrier to building a startup was technical. Software development was expensive and slow.

That context is changing quickly.

Artificial intelligence tools now allow software to be generated in minutes.

The real challenge is no longer building.

The real challenge is deciding what is worth building.

At Nomu we focus on that part of the process. Before building technology, we help teams understand whether the problem they want to solve is important enough to build a product around it.

Software can be generated with a prompt.

Finding a real problem is still the difficult part.