Prof. Yuval Noah Harari x Legislator KO Ju-Chun on 〈NEXUS〉

Prof. Yuval Noah Harari x Legislator KO Ju-Chun on 〈NEXUS〉

2024-10-03

原始音檔： QmT8rmm1ZichLxVtJFREFzupUMBxK8tQgMTiPNKf8sykja

Legislator KO Ju-Chun pointed out Taiwan's slow progress in AI development and expressed concerns about the transparency of the government's "sovereign AI," calling for public participation. Israeli historian Yuval Noah Harari emphasized that the AI revolution should not be a competition between good and evil, advocating for global cooperation to ensure safety and suggesting the creation of an international body to provide information.

How can we ensure the development of sovereign AI is transparent and involves public participation?
The rapid growth of U.S. social media has negatively impacted democracy.
There is a need to establish an international institution to share AI safety information.
Create an international body without regulatory power to provide AI information.

Ref. 「Harari x Audrey 2021」

Legislator KO Ju-Chun

Hello,

I'm legislator Ko.

I'm the only geek legislator

taking charge of tech

policymaking in Taiwan.

We are all fully aware that

China may develop dangerous AI

or authoritarian AI,

and Taiwan is kind of slow

in developing our own AI.

Recently,

our government may push

for a sovereign AI.

However,

without external oversight,

they only have

their self watching and make

sure it's safe by themselves.

So,

and also

they're possibly blocking

certain AI in the near future.

So I want to know that how can

we ensure a development

of sovereign AI is

transparent enough and have a public

involvement and

doesn't become another

authoritarian AI?

Prof. Yuval Noah Harari

Mmm.

Yeah,

I think a key message about

the AI revolution is not to

think that,

you know,

it's a race between

bad guys and good guys.

If you,

there are dangers all over

the world that even

the people you think are

currently the good guys,

if they move too fast,

if they

move without any thought for

safety,

this could result in disaster.

Like you think about social media.

So you know,

the United States won

the race on social media.

United States developed

the most powerful social media

platforms more than

10 years ago.

Facebook and YouTube and.

Twitter,

long before we saw the

spread of Chinese social media

around the world like TikTok,

and

The result was catastrophic That

what really undermined

democracies all over the world,

including the United States,

were not Chinese

Social media platforms.

It was Western social media platforms.

So this I think is is a

lesson that we should we

should take to heart.

and,

on the other side There are at

least some areas of

potential cooperation,

Global cooperation also with

China also with other regimes,

because as I said before,

About certain

problems of AI safety,

like the problem of control,

The Chinese are as

worried as the Americans.

So,

you know if we If one side

finds a way,

a theoretical breakthrough,

or a technical breakthrough,

how to ensure control of AI,

this is something that should

be shared with everybody.

This

should not be a state secret only for our side,

because it

would be bad for

everybody if another country

produces

AI that goes out of

control.

Now I think what we most

need is an institution,

an international institution,

which is not necessarily

even a regulatory institution.

First of all,

it's an informative

institution that simply

tells people around the world,

in different countries,

and tells different

governments,

what is the current state of AI

development and what are

the potential dangers.

You know,

if you think about regulation as teeth,

we need eyes

before we have teeth.

Otherwise,

we don't know what to bite.

And at the present moment,

most people around the world,

and even most governments,

are dimly aware,

if at all,

of the true state of

development of AI

and of the dangers.

If you think about

the government,

I don't know,

of Indonesia,

or of Nigeria,

or of Argentina,

how are they supposed

to know what is the

cutting edge of AI

technology and what are

the potential threats?

Because this information is

mainly in the hands of a

few corporations

in basically just two countries,

the US and China,

and these governments.

And what we need is first

of all a trustworthy

international

institution that could

perhaps be established

and financed by

international agreement,

which has no teeth,

just eyes,

just you know,

enough resources to pay

for some of the best talent.

in the world

and to have cutting edge

technology

just to inform people around

the world what is happening,

what are the dangers.

and now let's

think about what are the

regulations we need.

Thank you.