Apparently This Has Already Been Decided - When Certainty Arrives Too Early, Part One

AI Generated Illustration

So this week, we’re going to talk about certainty.

Not the useful kind, like knowing which road you’re on. The other kind. The sort that turns up early, sounds relaxed, and settles itself before anyone’s finished their complementary biscuit.

You’ll recognise the moment

• Someone mentions AI.
• Someone else responds.
• Both sound completely sure. They don’t agree, but that doesn’t seem to bother either of them.

The interesting thing isn’t the disagreement.

It’s how comfortable everyone is with it.

No one slows down and checks what’s meant. The conversation just carries on, as if confidence itself has done the job.

This isn’t really about AI

AI is just the thing people reach for when they want to talk about the future without getting specific. It’s a convenient container. You can pour certainty into it without having to answer awkward questions.

You hear the same tone around strategy, investment, restructures, and anything else where leaving things undecided would be irritating.

A small, ordinary example

This particular version of it happened in a café, but I'm sure you have heard and read it everywhere too.

Two people disagree about AI. One says it would take all the jobs. The other it wouldn’t.

Both speak as if this was settled. As if the thinking had already happened somewhere else and today was just about repeating the conclusion.

Then they conclude in a stalemate. Take a sip. Move on.

The questions that never get asked

• No one asked which jobs.
• No one asked when.
• No one asked what “take” actually meant. Those questions would have slowed things down. And slowing things down is exactly what certainty is designed to avoid.

What this looks like inside organisations

This is what these arguments look like once they wander into organisations. They present themselves as technical disagreements, but they behave like something else entirely.

Two positions. Strong convictions. Very little shared definition.

One side talks about AI like it’s already rummaging through the filing cabinets. Jobs gone. Skills hollowed out. Control lost while everyone was in a meeting.

The other talks about it like a steam engine. Progress. Inevitable. Complaining treated as a character flaw.

Both insist they’re describing reality. Neither bothers to say which version they’re using.

When disagreement turns into identity

That’s because the argument skips the dull but necessary bit and goes straight to identity.

• Sensible people think this
• Serious leaders say that.
• If you disagree, you’re awkward rather than analytical. That’s not how engineering arguments work. Engineers argue about limits, tolerances, and what breaks first. This is something else.

What certainty does to evidence

Once certainty takes hold, evidence stops behaving properly.

New information doesn’t get tested, it gets recruited. Slides turn into props. Forecasts turn into things you’re apparently supposed to respect.

AI just happens to be the current backdrop.

• If you feel commercially exposed, it sounds dangerous.
• If you feel structurally secure, it sounds useful.
• If you feel bored, it sounds interesting.

What people are really saying

None of this is strange. It’s just people being people. But it explains why organisations can talk for months and still feel certain they’ve been clear.

So when someone says "AI will take our jobs" or "AI won’t change much", don’t hear a prediction.

Hear a preference.

They’re telling you which version of the future they’d quite like the organisation to commit to so they can stop worrying about it.

Tomorrow, we’ll look at how truth gets decided in organisations, usually after the outcome, not before.