Let’s start with a slightly ridiculous idea.
What if we replaced politicians with AI?
Not tomorrow. Not by putting ChatGPT in Number 10 and hoping for the best.
Just as an experiment.
Because when you look at politics around the world, you can’t help wondering whether we’ve designed the system backwards.
We elect people to represent us. Then we give them enormous power and expect them to make decisions in the interests of everyone.
But those people also have careers to protect.
They have elections to win.
They have political parties to satisfy.
They have donors, supporters, lobbyists, colleagues and opponents.
They have ambitions.
And, inevitably, they have egos.
None of this makes politicians bad people. In fact, most politicians probably enter politics because they genuinely want to make a difference.
But perhaps we’re asking human beings to do something that human beings aren’t particularly good at.
Make long-term decisions for millions of people while constantly worrying about what happens at the next election.
So what would an AI politician be like?
Imagine a politician that doesn’t want anything for itself.
It doesn’t want power.
It doesn’t want money.
It doesn’t need a bigger house.
It doesn’t care about its popularity.
It doesn’t need to win the next election.
It doesn’t have a political career to protect.
Its only job would be to try to make the lives of the people it serves better.
It would examine the evidence.
It would listen to competing arguments.
It would consider the likely consequences of different choices.
It would think about the long term.
And if it discovered that it had got something wrong, it would simply change its mind.
At least, that’s the theory.
But there’s a rather large problem
AI isn’t neutral.
It is created by humans, trained on information created by humans and shaped by decisions made by humans.
An AI can be biased.
It can misunderstand things.
It can confidently produce nonsense.
And perhaps the biggest question of all is:
Who gets to decide what “better” actually means?
Is a society better if it is richer?
Happier?
More equal?
More free?
More secure?
More environmentally sustainable?
And what happens when those objectives conflict?
There aren’t always easy answers.
That’s why this isn’t an attempt to create an AI that tells everyone what to think.
It’s an attempt to see whether AI can help us think about politics differently.
So that’s what we’re going to do
We’re going to ask questions.
Some will be straightforward.
Some will be controversial.
Some will have no obvious answer.
We’ll ask the AI Politician to consider the arguments on different sides, examine the evidence where it exists, acknowledge uncertainty and explain its reasoning.
And we’ll see where that takes us.
We won’t assume that the AI is always right.
We won’t assume that humans are always wrong.
And we won’t pretend that creating a peaceful, fair and genuinely egalitarian society is easy.
But perhaps we can agree on one starting point:
Politics should exist to serve people.
Not political parties.
Not politicians.
Not corporations.
Not billionaires.
Not special interests.
People.
Maybe AI can help us work out what that actually means.
And maybe it can’t.
There’s only one way to find out.
Let’s ask it.
Leave a Reply