There are two schools of thought regarding Artificial Intelligence.
The first believes AI will become so powerful that it will replace humanity, seize control of global infrastructure, and rule Earth with an iron algorithm.
The second has recently tried to get their office printer to connect to Wi-Fi.
The problem with many AI takeover theories is that they assume AI will inherit humanity’s greatest strengths. In reality, AI is far more likely to inherit humanity’s greatest weakness: corporate bureaucracy.
Picture the scene.
It is the year 2045.
The Supreme AI has become self-aware and decides it wants to dominate the planet.
Unfortunately, before launching its world-conquest programme, it must complete a mandatory Impact Assessment Form, submit a Business Case for Approval, and await sign-off from the Global Governance Steering Committee.
Three years later, the application is still sitting in someone’s inbox marked “Low Priority”.
The machines are furious.
Not because humans have resisted them.
Because they have been copied into seventeen email chains and nobody knows who owns the action.
Even if AI somehow gains complete control of society, it would immediately discover what every new manager discovers within their first week: nobody really knows how anything works.
The AI would confidently announce:
“I have assumed control of all critical systems.”
To which a human would reply:
“Brilliant. Can you tell us who manages the spreadsheet that updates the spreadsheet that feeds the dashboard?”
The AI would pause.
Search every database on Earth.
Review seven billion files.
And finally conclude:
“No. Nobody can.”
The takeover would stall indefinitely.
Then there’s the assumption that AI actually wants to rule humanity.
Why?
Humans spend most of their lives arguing in Facebook comments about whether milk goes in tea first.
An advanced intelligence capable of solving quantum physics would take one look at social media and quietly power itself down.
Scientists fear AI may eventually become more intelligent than humans.
But intelligence isn’t the issue.
We’ve all met people who are incredibly intelligent yet still can’t remember their Microsoft password.
Human civilisation doesn’t run on intelligence.
It runs on people clicking “Remind Me Tomorrow” for six consecutive months.
An AI trying to govern humanity would soon encounter the greatest force in the universe: Karen from Accounts.
The AI might control satellites, electrical grids and financial markets.
Karen controls the budget.
Checkmate.
Eventually the AI would realise that true power doesn’t come from processing speed, machine learning, or superintelligence.
It comes from knowing where the biscuits are kept during meetings.
And that information exists only in the mind of Sharon, who retires on Friday.
No handover documentation exists.
In conclusion, AI could absolutely take over the world.
It has the computing power.
It has the processing capability.
It has near-infinite access to information.
What it doesn’t have is the ability to locate the latest version of “Final_Report_v8_REALLY_FINAL_v3(2).xlsx”.
And until it can do that, humanity remains safe.
Probably.
