G7 的 AI 午餐会：当大国领导，和 AI 巨头同桌

Today we'll talk about a piece of international news worth discussing.

On June 17, in Évian, France,

a very special working lunch was held.

To understand this,

let's first briefly explain the G7.

G7, in Chinese, is called the 'Group of Seven'.

It's a small circle of the world's seven most developed,

wealthiest countries.

It includes the US, the UK, France,

Germany, Japan, and such countries.

Every year, these countries' leaders

find a place to hold a meeting.

Together they discuss the world's big matters.

This year's meeting was held in France.

And this lunch,

what made it most special was:

the ones at the table weren't only national leaders.

There was also a group of tech-company bosses.

And moreover, the companies that, in the world today,

do AI the best.

For example, OpenAI's Altman,

Google DeepMind's Hassabis,

and Anthropic's Amodei.

These names may be unfamiliar to you,

but the products they make, you've most likely used,

or heard of.

So picture that scene:

on one side, national leaders who hold power.

On the other, tech giants who hold the technology.

They sat at one table, eating, chatting.

So what exactly were they talking about?

The core was one thing:

how to let AI develop safely.

Right now AI's capabilities are rising too fast.

So fast that even these AI-makers

are a little worried themselves.

They worry about several things:

for instance, whether AI will be used to do harm;

for instance, how to protect children

from being harmed by AI content online;

and for instance, whether one country's AI

will become too dependent on another country.

But the most interesting part

is their disagreement over 'who sets the rules'.

Several American companies' bosses

put forward an idea:

America should take the lead,

pull together a group of countries into a coalition,

and jointly set AI's rules and standards.

Put plainly, they hope America is the one in front.

Altman also proposed

building an international forum,

specifically to give each country's AI

a unified, impartial safety test.

But Europe, on its side, wasn't comfortable.

European countries, with France as the representative,

are increasingly worried about one thing:

that America is too dominant in AI.

They raise a term: 'tech sovereignty'.

It means: Europe doesn't want to just follow behind America,

listening to America, using America's, obeying rules America sets.

Europe also wants its own AI,

its own voice.

So this lunch, on the surface a friendly meal,

was, underneath, a contest over

'who calls the shots in the future'.

So, did they talk out a result?

Honestly, no.

This lunch

released no specific policy,

and set down no rules.

Everyone just sat down

and put their thoughts and worries on the table.

But even so,

this meal itself is already a signal.

It shows: the matter of AI

has become important enough

that it needs the most powerful people,

and the people who understand the tech best,

to sit at the same table to talk.

Finally, let's think together.

There's a brand-new, powerful technology.

The company that made it says: 'We understand it best, we'll set the rules.'

But other countries say: 'No, we want to take part too.'

Do you think a powerful technology

should, in the end, be managed by whom?

By the people who understand it best?

Or by everyone who could be affected by it,

deciding together?

This question has no standard answer.

But it concerns the future of every one of us.

Finally, I want to ask you.

As AI more and more

enters your work and life,

the thing you most hope someone helps you hold the line on,

which line is it?