AI Is Becoming Cheap. Knowing What to Build Isn’t.

A few years ago, simply saying that a product used artificial intelligence could get people interested.

That period is disappearing quickly.

AI is becoming available everywhere. Businesses can access increasingly capable models without spending hundreds of millions of dollars developing the underlying technology themselves.

That’s going to create an extraordinary amount of innovation.

It’s also going to create an extraordinary amount of software nobody really needs.

And I think that distinction is important.

Technology Isn’t the Starting Point

One of the questions I hear constantly is:

“How can we use AI in this business?”

I understand why people ask it, but I’m not convinced it’s the best place to start.

I’d rather ask:

“What’s unnecessarily difficult about this business?”

Where are people wasting time?

Where is information being lost?

What frustrates customers?

What decision repeatedly gets delayed because nobody has the information they need?

What process exists simply because nobody has questioned it for ten years?

Once you understand the problem, then you can ask whether AI can help.

Sometimes it can.

Sometimes ordinary software is enough.

And occasionally the best solution has nothing to do with technology.

Building Has Become Easier

This is one of the most exciting changes happening in technology.

The distance between having an idea and testing it is shrinking.

A founder can create a prototype faster.

A small team can automate processes that previously required considerably more resources.

Businesses can experiment without committing enormous amounts of capital before discovering whether customers actually care.

That democratises innovation.

But it also changes the bottleneck.

When building becomes easier, choosing what deserves to be built becomes harder.

Problems Become the Scarce Resource

There are millions of possible AI products.

Most of them probably shouldn’t exist.

The valuable opportunities will come from understanding real problems deeply enough to recognise where technology genuinely improves the outcome.

That often requires industry knowledge.

You need to understand how people actually work.

You need to understand why the existing process developed.

You need to know which frustrations matter and which are merely inconvenient.

That’s why I think some of the most interesting AI companies will emerge when technologists work alongside people with deep industry expertise.

The technology matters.

But understanding the problem matters just as much.

This Is How We’re Trying to Approach Our Own Projects

With IQ Pitch, we didn’t start with the idea that founders needed another AI tool.

The interesting problem was that potentially good businesses can struggle to communicate their opportunity effectively.

Could technology help founders identify gaps before an investor does?

That’s a problem worth exploring.

Satorian Systems begins with another problem: organisations already generate enormous amounts of information, yet decision-makers can still struggle to understand what’s happening across the organisation.

Again, AI is part of the potential solution.

It isn’t the reason for building it.

I think maintaining that discipline is going to become increasingly important as AI makes software easier to create.

The Best Technology Often Becomes Invisible

Most people don’t care what programming language their banking app uses.

They don’t care about the database architecture behind an airline booking.

They simply want it to work.

AI will eventually become similar.

We’ll stop announcing that everything is “AI-powered” because intelligence will simply become part of how modern software operates.

At that point, businesses will have to compete on something else.

Does the product solve the problem?

Is it reliable?

Is it easy to use?

Do people trust it?

Does it actually improve their lives or businesses?

Those are much harder questions than whether something contains AI.

Looking Ahead

The cost of accessing artificial intelligence will continue to fall.

The ability to build sophisticated software will continue to spread.

That means ideas can be tested faster than ever before.

I think that’s an extraordinary opportunity.

But it also means technology itself becomes less of a differentiator.

The real advantage will increasingly come from understanding problems.

Knowing where technology genuinely adds value.

Knowing when it doesn’t.

And having enough discipline not to build something simply because we can.

AI is making it easier to build.

The difficult part remains deciding what’s worth building.

Contact Us for More Information