The £100 AI Business: Why the Next Wave of Entrepreneurs May Start With Almost No Infrastructure

Starting a business used to begin with a list of things you needed.
An office.
Employees.
Equipment.
Software.
Marketing.
Inventory.
Capital.
Artificial intelligence is steadily removing items from that list.
A timely signal comes from recent entrepreneurship coverage. On August 3, Forbes highlighted AI-powered side businesses built around tools such as ChatGPT⁠� and Claude, reflecting how generative AI is lowering the operational barrier for individuals who want to turn skills into businesses. �
Forbes
Meanwhile, a July Harvard Business Review analysis argues that agentic AI can particularly benefit startups because younger companies can design operations around AI from the beginning instead of retrofitting it into decades-old processes. �
Harvard Business Review
Something interesting is happening.
The minimum viable company is getting smaller.
The £100 Thought Experiment
Imagine giving someone £100 and asking:
Can you build a digital business?
Twenty years ago, the answer would probably have been no.
Today, perhaps.
Not necessarily a profitable company immediately.
But the infrastructure?
Absolutely possible.
A domain.
Basic website.
AI subscription.
Payment processor.
Email list.
One useful digital product.
Suddenly you have the beginnings of a commercial system.
The scarce resource isn’t necessarily technology anymore.
It is knowing what to build.
AI Changes the Cost of Experimentation
This may be one of AI’s biggest effects on entrepreneurship.
Consider an entrepreneur who wants to test an idea for a small educational product.
Previously, she might need someone to:
research the market,
write the material,
design graphics,
build a landing page,
prepare marketing copy,
analyse customer feedback.
AI can now assist across all of those stages.
That doesn’t mean the entrepreneur should blindly accept AI output.
It means she can test an idea before investing heavily in it.
That changes risk.
The Rise of Micro-Products
This creates ideal conditions for micro-products.
Not every digital product needs to become a £500 course.
Someone may need:
a £3 checklist,
£5 template,
£10 prompt library,
£15 mini-course,
£20 specialised AI tool,
or £5 monthly subscription.
The internet allows thousands of small transactions to accumulate.
AI reduces the cost of creating and maintaining the products.
The combination could expand the long tail of entrepreneurship.
One Problem Is Enough
Entrepreneurs often make the mistake of beginning with:
“I want to build a business.”
A more useful starting point is:
“What annoying problem can I remove?”
Perhaps freelancers struggle to create invoices.
Parents struggle to organise school schedules.
Creators struggle to repurpose long videos.
Small businesses struggle to write consistent product descriptions.
Students struggle to organise research.
Solve one narrow problem extremely well.
Then expand.
The AI Microbusiness Stack
A modern microbusiness might need only a handful of components:
Knowledge — something useful you understand.
AI — to increase execution capacity.
Website — somewhere you control.
Payment infrastructure — somewhere transactions happen.
Audience — people with the problem.
Product — the solution.
Notice what isn’t necessarily required.
A huge team.
An expensive office.
Millions in venture capital.
AI Agents Could Shrink the Company Further
Agentic AI takes this another step.
Instead of manually prompting an AI for every task, entrepreneurs are beginning to experiment with systems that can carry out multi-step workflows.
A research agent might analyse competitors.
A customer agent might organise common enquiries.
A content agent might prepare drafts.
A human remains responsible for strategy and consequential decisions.
The organisational structure begins looking surprisingly sophisticated.
One human.
Several specialised intelligences.
But Cheap Creation Creates a New Problem
If everyone can create digital products cheaply, the internet will fill with them.
So creation stops being the advantage.
Trust becomes the advantage.
Expertise.
Distribution.
Brand.
Original perspective.
Customer understanding.
Those remain difficult.
AI lowers the manufacturing cost of ideas.
It doesn’t automatically create demand.
Don’t Build 100 Products Nobody Wants
The temptation of AI is abundance.
If you can generate one e-book, why not twenty?
If you can create one application, why not fifty?
Because customers don’t care how many things you can manufacture.
They care whether one solves their problem.
The smartest AI entrepreneur may therefore create less than technically possible.
Test.
Observe.
Improve.
Then scale.
The New Entrepreneurship Equation
The old equation often looked like:
Capital + Labour + Infrastructure = Business Capability
The emerging equation could look more like:
Knowledge + AI + Distribution + Trust = Business Capability
Capital still matters.
People still matter.
But the threshold for beginning has fallen dramatically.
Final Thoughts
Artificial intelligence may democratise entrepreneurship in a way the internet only started.
The internet gave individuals access to global customers.
Cloud computing gave them infrastructure.
Social media gave them distribution.
AI gives them something new:
scalable cognitive capacity.
That doesn’t guarantee success.
It simply gives far more people permission to try.
And sometimes the next serious company begins not with millions of pounds.
It begins with one person asking:
What useful thing can I build with what I already know?