The Next AI Unicorns Haven’t Been Born Yet—Will You Be Their Founder?

You are not too late in the AI revolution. Over the next decade we will see unimaginable AI innovations and new business models changing the way our world works.

I am very excited about the opportunities, and the good news is that you are not too late. This is where the entrepreneurs will shine.

You can still become an early adopter of AI… and it starts by educating yourself.

Let’s look at some of the opportunities.

AI Is Like Electricity—But We Haven’t Built the Appliances Yet

Imagine it’s the late 1800s. The first power stations have just been built. Electricity is real, and it works. But if you were living in that era, would it have changed your life overnight?

Not really.

Why? Because while the power grid existed, the appliances didn’t. People were still boiling water over open flames because the electric kettle hadn’t been invented yet.

The real transformation didn’t happen when electricity was discovered—it happened when everyday products were designed to use it.

AI today is in the same place. The “power stations” (large AI models like ChatGPT, Midjourney, and Claude) have been built. They can generate text, create images, and analyse data at an incredible scale. But we’re still waiting for the killer applications—the AI-powered equivalents of kettles, fridges, and washing machines that will make AI truly indispensable in daily life.

We’ve Seen This Before: The Internet Took Decades to Reach Its Full Potential

The internet wasn’t an overnight success. The foundation (networking protocols) was built in the 1960s, but it took decades for broadband, cloud computing, smartphones, and payment systems to unlock world-changing applications like Instagram, Zoom, and Amazon.

AI is at that same inflection point. The foundation exists, but the real transformation will come from those who build the next layer of infrastructure, interfaces, and business models.

We’re now at that moment with AI. The foundational technology exists, but the infrastructure, interfaces, and business models that will define AI’s future are still being built.

AI is at this same inflection point. The foundational technology exists, but the real revolution is still ahead.

Right now, we’re seeing a flood of AI wrapper companies—businesses built on top of existing AI models like ChatGPT, Midjourney, and Claude. But just like the internet didn’t stop at static websites, AI won’t stop at simple chatbots.

The Early Adopter Advantage: Lessons from Electricity & the Internet

History has shown that the biggest winners aren’t those who merely adopt a technology—but those who build the systems that power i

• Electricity: The wealth wasn’t in light bulbs—it was in power grids and electric-powered industries.

• Internet: The winners weren’t basic websites—it was companies that created new business models like Amazon (e-commerce), Google (search & ads), and Facebook (social media).

• AI: The biggest profits won’t come from those who simply automate existing processes—but the ones who reimagine how businesses, industries, and economies work in an AI-first world.

So, where are the unbuilt AI opportunities?

It starts with Education:

AI’s Lack of a Learning Curve: a Risk

Just because you know how to switch a light on and off or boil water in an electric kettle, doesn’t make you an expert in electricity. You wouldn’t claim to understand electrical engineering just because you can plug in an appliance.

The same goes for AI.

Knowing how to type a prompt into ChatGPT isn’t the same as understanding how AI models work, how to get better outputs, or how to integrate AI into real-world business strategy.

Where can you find the opportunities? This is where I get excited…

Where to Look for AI Opportunities: The Missing Piece

Over the next 5–10 years, the biggest opportunities will be in:

1. AI-powered infrastructure → The backbone that supports AI’s expansion (just as AWS did for the internet):

   • AI Infrastructure

   • Data

   • Interfaces

2. AI-native business models → Companies that don’t just use AI, but are fundamentally built around it. Just like Uber wasn’t a taxi company with an app, these businesses will rethink industries from the ground up.

3. Industry-specific AI applications → AI tailored for high-value industries like healthcare, finance, and law.

1. AI-Powered Infrastructure → The Backbone of the AI Economy

The internet didn’t just succeed because of websites—it needed a digital backbone (AWS for cloud computing, Stripe for payments, Google for search).

AI will require its own infrastructure—companies that provide the platforms, tools, and systems that other AI-driven businesses will rely on.

1.1 AI Needs Its "Power Grid"—The Physical Infrastructure

Just like the internet needed AWS, Stripe, and Twilio. The biggest opportunities lie in:

• AI Compute & Energy → AI models demand enormous power. Solutions in energy-efficient AI chips, sustainable power, and on-device AI processing will be critical.

• AI-Optimised Data Storage & Processing → AI thrives on structured, high-quality data. Businesses that streamline AI data pipelines will be invaluable.

• AI Workflow Automation → Just like Make and Zapier automates tasks, AI needs seamless integration into business workflows.

• Data Centers & Cloud AI – AI needs scalable cloud infrastructure to store and process vast amounts of data. Businesses that provide AI-ready computing will be essential.
  On-Device AI – AI currently runs on the cloud, but in the future, it needs to run locally on devices for real-time applications in mobile, robotics, and IoT.

Big Idea: The companies that solve AI’s computing bottlenecks will become the backbone of the AI economy—just like AWS did for the internet.

1.2 AI Needs Fuel—Better Data Infrastructure

AI doesn’t generate knowledge—it learns from data. And right now, most businesses have messy, fragmented data that AI can’t use effectively.

AI is only as good as the data it’s trained on—but most businesses still rely on unstructured, fragmented, or low-quality data.

AI-driven data management, privacy, and security tools will be in high demand as businesses realise that bad data = bad AI. 

Opportunities in data are:

• High-Quality, Structured Data – AI is only as good as its training data. There’s a massive opportunity in data-cleaning, labelling, and structuring tools.

• AI-Optimised Data Pipelines – AI needs real-time, high-quality data flows to be truly effective. Businesses that can organise and manage AI-ready data will be critical.

• Privacy & Security – As AI becomes more integrated into business, secure, ethical data handling will be a priority.

• Industry-specific datasets – AI needs clean, high-quality data, especially in fields like law, finance, and healthcare.

💡 Big Idea: The best AI doesn’t come from better models—it comes from better data. Businesses that solve the data quality problem will win big.

1.3 AI Needs Better Interfaces—Making It More Usable

Just as electricity needed appliances and the internet needed browsers, AI needs better interfaces to be truly transformative.

Creating better user experiences and interfaces is an opportunity:

• AI "Browsers" & Interfaces – Right now, AI is mostly chatbots. The future lies in seamless AI integration into existing tools and workflows.

• Personalised AI Assistants – Instead of generic AI, customizable AI agents that learn user preferences will be the next wave.

• Multimodal AI – AI needs to go beyond text and images to voice, video, and real-world interaction.

Big Idea: The companies that make AI seamless and intuitive (like how Apple did for smartphones) will define the next decade.. AI’s iPhone moment hasn’t happened yet—who will build it?

2. AI-Native Business Models: The Missing Economic Layer

I love exploring new innovative business models… This is where we can get very innovative, and this is where we as South African can shine (so let’s take advantage of this opportunity - building on top of other’s infrastructure).

The internet created entirely new ways to make money:

• SaaS subscriptions

• E-commerce

• Digital advertising

AI will do the same, but we haven’t yet seen the dominant AI-native business models.

Emerging AI Monetisation Models:

• AI-as-a-Service – AI-powered consulting & decision-making tools.

• AI-driven marketplaces – AI dynamically matching customers and providers.

• AI-powered personal assistants – Subscription-based AI that learns user preferences.

Most companies use AI to streamline tasks, but true breakthroughs come from reimagining industries, not just optimising them.

History shows the difference:

• When the internet arrived, traditional retailers built websites. But Amazon didn’t just create an online store—it built an entirely new way to buy and sell goods.

• Uber didn’t just add an app to taxis—it redesigned transportation itself.

Big Idea: The biggest AI winners will be the companies that don’t just add AI to existing workflows, but build entirely new business models that couldn’t have existed without AI.

If you’re building a business today, ask yourself: Are you designing it for an AI-first world?

3. Industry-Specific AI Applications → The Next Wave of Disruption

AI’s biggest breakthroughs won’t come from generic tools but from industry-specific AI solutions tackling complex, high-value problems.

Do you have deep industry expertise?

Or do you have specific industry datasets?

These can be massive opportunities.

Emerging AI Opportunities

AI-powered businesses will redefine industry workflows, creating efficiencies that were previously impossible.

• AI in construction → Predictive maintenance & autonomous project management.

• AI in manufacturing → AI-driven robotics running entire factories.

• AI in government & policy → AI-assisted legislation and legal research.

• AI in environmental sustainability → AI-powered climate modelling & smart agriculture.

The Next Billion-Dollar Ideas: AI for Industries That Haven’t Been Disrupted Yet

Big Idea: The next wave of AI startups won’t just be tech companies—they’ll be AI-powered versions of traditional industries.

AI Risks and Hurdles

While the future of AI is bright, navigating the path to widespread adoption will require careful consideration of potential roadblocks. A realistic and nuanced view acknowledges these challenges and emphasises the need for proactive solutions. 

Successfully building the "kettles" of AI will depend on addressing these key hurdles:

• Regulatory Uncertainty: 

• Implementation Complexity

• Skills Gap

• Ethical implications 

• Job displacements

• Potential for misuse

Keep the risks and hurdles in mind as you explore new opportunities.

The First Step? Mindset & Education

Technology alone isn’t enough—what matters is how we learn to use it. You need to understand AI, from a First Principles Perspective.

AI is still in its early days, but the time to adapt is now.

The next decade belongs to the builders. The foundation has been laid, but the real work—building the AI-powered world—is just beginning. Invest in your AI fluency. Explore online courses, attend workshops, and experiment with different AI tools. 

Developing this fluency often involves not just theoretical knowledge, but also practical experience and guidance from experts. The future will be shaped by those who understand not just what AI can do, but how to build it. 

Join one of our Executive workshops.

What opportunities are you excited about?