Hey,

A few years ago, if you asked people what would shape the future of AI, most answers would sound familiar.

More models.

Better chips.

Smarter products.

Faster automation.

Very few people would have said electricity.

Or financing.

Or transmission lines.

But sometimes the biggest opportunities hide underneath the exciting headlines.

And I think that's exactly where we are right now in Africa.

Everyone wants to talk about the AI applications being built.

Almost nobody wants to talk about the pipes underneath them.

This week, I want to take you beneath the surface.

Because I think some of the biggest companies of the next decade will be built underneath AI.

Let's get into it.

The AI conversation is missing something

When ChatGPT exploded, founders everywhere started asking:

"How do we build AI into our product?"

Fair question.

But I think many founders are asking the wrong question.

Because AI is infrastructure-heavy.

Every AI product depends on:

• computing power
• reliable energy
• cloud capacity
• financing
• internet connectivity
• data systems

Without those things, AI becomes a nice demo instead of a durable business.

And this becomes even more important in Africa.

Because while AI is global, infrastructure is painfully local.

A founder in San Francisco may think about prompts.

A founder in Nairobi may have to think about power reliability, cloud costs, internet access, and customer affordability all at once.

That difference matters.

A lot.

Because constraints create markets.

And markets create opportunities.

Which brings me to something interesting happening quietly in East Africa.

Last year, Microsoft and G42 announced plans around a major AI and cloud data center project in Kenya.

The vision was ambitious.

A geothermal-powered cloud region designed to support AI and cloud demand across East Africa.

On paper, it sounded perfect.

Kenya had:

• growing developer ecosystems
• strong regional influence
• renewable geothermal energy
• increasing cloud demand
• government support

The story looked simple.

Until reality showed up.

Reports later suggested the project encountered challenges involving payment guarantees and questions around energy capacity.

And honestly?

That story may be one of the most important AI lessons founders can learn this year.

Because it exposed something many people miss:

Building AI infrastructure is:

An execution problem.

A financing problem.

An energy problem.

It's a coordination problem.

You can have world-class technology.

You can have demand.

But if the infrastructure underneath is weak, everything slows down.

This matters because many founders are assuming AI growth will happen automatically.

History says otherwise.

The companies that win often build around bottlenecks.

Let's talk about something that sounds boring but matters a lot.

Power.

Data centers are extremely hungry.

AI workloads are even hungrier.

As AI use increases, the amount of electricity needed grows dramatically.

And Africa already has an energy challenge.

Now layer AI demand on top.

Suddenly, power is becoming a technology conversation.

Then financing enters the picture.

Because infrastructure requires large upfront investment.

Someone has to take the risk before demand fully exists.

Governments hesitate.

Private investors want certainty.

Builders want guarantees.

Everyone waits for someone else to move first.

This creates friction.

And friction creates opportunity.

Because founders who understand where friction exists can build solutions around it.

I think many founders are still seeing AI as a software opportunity only.

But increasingly, AI may become an infrastructure opportunity too.

Whenever a major technology shift happens, most people chase the obvious opportunities.

But the biggest companies often emerge one layer underneath.

During the mobile revolution:

Everyone noticed smartphones.

Fewer noticed payments.

Even fewer noticed logistics and distribution.

During the internet wave:

Everyone saw websites.

Fewer saw cloud infrastructure.

Even fewer saw cybersecurity.

We're seeing the same thing happen again.

Here are categories I think become increasingly interesting:

AI infrastructure management

Helping companies manage AI costs, usage, and workflows.

Energy optimization systems

Helping businesses use energy more efficiently as computing needs grow.

Localized AI tools

Models and systems trained for local languages, local regulations, and local business realities.

Compute marketplaces

Making AI resources cheaper and easier for regional businesses.

Vertical AI systems

Industry-specific solutions in:

Healthcare

Agriculture

Financial services

Logistics

Education

Compliance

The winners may simply be the companies solving expensive bottlenecks.

Let's go back in time.

In the early telecom era, connectivity in many African markets was limited.

People focused on the missing phones.

But the bigger opportunity was infrastructure.

Once networks expanded:

Mobile payments emerged.

Digital commerce emerged.

Ride-hailing emerged.

Large technology ecosystems emerged.

The same pattern may repeat with AI.

First comes infrastructure.

Then come platforms.

Then come products.

Then entire ecosystems appear.

This is important because many founders build too late in the chain.

They build where competition already exists.

But the strongest opportunities often exist lower down.

Ask yourself:

What constraint keeps showing up repeatedly?

Where are people losing time?

Where are costs increasing?

What slows adoption?

Those are usually better startup ideas than "AI for X."

If you're building right now, here's something I'd think deeply about:

Don't only ask:

"How can AI improve my product?"

Also ask:

"What infrastructure assumption am I making?"

Because your customers live inside real environments.

Real power systems.

Real financial systems.

Real budgets.

The founders who win will build businesses that work inside reality.

And reality, especially in Africa, has constraints.

But constraints are where opportunities hide.

The next breakout category may already be forming.

Most people just aren't looking beneath the surface yet.

If you're building, scaling, or simply trying to understand where African tech is moving next, join our events calendar here.

We're bringing together operators and founders having the conversations that usually happen behind closed doors.

I also share weekly African founder stories and breakdowns on my personal LinkedIn — not just wins, but why those wins happened.

And on the Smarter SaaS Growth AI LinkedIn page and newsletter, I break down playbooks weekly so you can steal patterns instead of starting from zero.

And one more thing.

If you want your brand in front of 4,000+ B2B operators actively building and scaling across Africa, book a call here.

Sometimes one good introduction changes an entire company trajectory.

See you next week.

-Angela.