CAN AI BE WISE?

Around your nineteenth birthday, your brain clocks one of its fastest laps.

Raw speed – the flat-out, never-seen-this-before kind of thinking – peaks early. From there, it begins to ebb.

Cheering, for a Monday.

But other things keep climbing. Knowledge, vocabulary, patterns.

And ultimately, judgement.

Knowing which precedent matters, when the obvious answer is wrong, when your confidence has outrun the evidence.

Adulthood is a long exchange. Speed goes out. Experience comes in. The proceeds – if you keep paying attention – are wisdom.

AI never makes that exchange.

It arrives fast, tireless and absurdly well read. It crosses disciplines in seconds. It can answer before most of us have properly understood the question.

But nothing has ever happened to it.

It's digested the world's wisdom, but is it wise?

It has never felt the pain of a bad decision. Never heard the awkward silence after confidently saying the wrong thing. Never had to apologise to the person who paid for it.

And so it may hand you a wrong answer with exactly the confidence of a right one.

Or give sage advice one day and contradict it the next.

The machine brings processing power and the library, but for now, people bring the scar tissue: the edge cases, the costs, the instincts built from being wrong and having to live with what followed.

Which leaves the Monday question:

How do we get the scar tissue into the machine before the machine gets into the decision?

Happy Monday.