I’m Betting €1M That AI Can Run My Business in 2026

2026-01-01

TL;DR

In 2026, I am betting (investing) up to €1 000 000 to test one idea: Can a real business run reliably with AI in 2026? Not as a robot CEO, but by using AI as guardrails to give the operators super powers

AI can do everyting better than a human, or will do it better in the near future. But there is one lay of physics that will make sure humans will remain relevant: entropy. Without the irratinal need for humans to burn huge amounts of energy for long stretches of time to acchieve a finctional goal, nothing would relly happen at all. AI has no drive to accomplish anything. We are the reason it exists. The human stays responsible for vision, people, and judgment. The system handles standards, execution, checks, and follow-through.

We combine:

Big BOB: proven consulting knowledge

BASE100: a standard way to describe any business

BOSS: a roadmap to turn companies into AI-powered systems

The goal is simple: turn small businesses into professionally operated companies, and prepare them for exit.

This is not about replacing humans

This experiment is often misunderstood.

It is not about taking people out of the loop. It is about taking chaos out of the system.

Most business failures are not caused by bad intentions or low effort. They are caused by normal human limits.

People get tired. People leave. People make inconsistent decisions. Context gets lost. Standards slowly decay.

In small and mid-sized companies, this compounds fast.

The real problem we keep seeing

As an investor in SME companies, one issue keeps repeating.

Finding the right CEO for the size of the company is extremely hard.

Either:

the CEO is over-qualified, too expensive, and mismatched to the stage
or
the CEO has energy and drive, but lacks experience and pattern recognition

Both cases create risk.

Not because people are bad. But because the business depends too much on individuals.

So the question becomes:

What should a CEO really do, and what should the company do by itself?

The CEO role, reduced to essentials

A CEO should focus on the work that truly makes a difference:

setting direction
building trust
hiring and keeping the right people
creating alignment
making judgment calls where context and taste matter

What a CEO should not have to carry is:

every process detail
every checklist
every control loop
every reminder
every report interpretation

That part should be engineered into the business.

If leadership quality drops slightly, the company should not fall apart.

AI as guardrails, not autonomy

This is where AI becomes interesting.

Not as a replacement for leadership. But as guardrails for it.

Think of it like a technical system. You do not rely on perfect operators. You design the system so failure is hard.

AI can:

enforce standards
check consistency
surface deviations early
execute repeatable work
reduce decision load

The human stays in control. The system reduces the damage of normal human variation.

That is the bet.

Why most AI-in-business experiments fail

Most attempts fail for one reason: drift.

Small changes in inputs cause large changes in outputs.

a prompt changes slightly
context is missing one day
extra information appears the next
the model behaves differently under the hood

For creative work, this is fine.

For business operations, it is unacceptable.

In business, the rule is simple: the same input should produce the same output.

Most agent-based setups today cannot guarantee that. So trusting them with a real company would be irresponsible.

Unless the approach changes.

The top-down approach

Most AI adoption is bottom-up.

automate emails
generate content
answer support tickets
build small tools

Useful, but limited.

We take a top-down approach.

That means:

define what the business is
define how it should behave
define standards and targets
define decision rules
define escalation paths
define feedback loops

Only then do we let AI operate inside those boundaries.

Not via chat. Not via free-running agents.

Through controlled software systems.

The system we use

This experiment combines three components.

Big BOB

A collection of proven consulting knowledge. Boring. Old. Effective.

The kind of knowledge that turns chaos into predictable performance.

BASE100

A standard model to describe any business in a consistent way.

Most companies cannot clearly describe themselves as a system. That makes automation fragile.

BASE100 becomes the shared language between humans, software, and AI.

BOSS

The implementation roadmap.

This is not a switch you flip. It is a sequence.

BOSS defines how a real, messy company is converted step by step into an AI-supported operation, without breaking it.

What Marqui actually does

This is exactly how we work at Marqui.

We step into companies and apply this system to:

professionalize operations
reduce decision load
remove bottlenecks
improve margins
stabilize execution
prepare the business for exit

The result is not a flashy AI demo.

The result is a company that is harder to mess up.

The 2026 experiment

In 2026, this system will be tested on a real business.

Not a simulation. Not a sandbox.

A small company. Near break-even. Full of complexity. Real staff. Real customers. Real consequences.

I am committing up to €1,000,000 to this experiment.

Because opinions do not matter. Results do.

Why follow this series

This is not a motivational blog. It is not a marketing channel.

It is a public engineering log.

You will see:

what we design
what breaks
what works
what does not
how complexity is reduced
how systems replace heroics

If you run a business, this shows what professionalization looks like in practice. If you operate inside one, it shows how structure beats effort. If you invest, it shows what makes companies durable and sellable.

No hype. No shortcuts. No magic.

Just systems.

One sentence to remember

The future is not AI replacing leaders.
It is leaders supported by systems that make businesses reliable.