RE:WIRED 2021: Timnit Gebru on AI and the Discipline of Hope

The institute hasn't launched yet,

[upbeat music] but let's say a year from now,

two years, five years from now,

if everything goes well, which I'm sure it will,

how will the way AI is used

or talked about the be different?

Are you hopeful that you can change things?

Well, I have to be hopeful [laughs].

I don't think we have other options

and Ruha mentioned, Ruha Benjamin was quoting someone,

I forget, saying hope is a discipline.

You need to have the discipline of hope.

If you don't have hope, it's difficult to get anything done.

So what I'm hoping to do is first of all,

I'd like to see a lot more institutes,

like the one I'm hoping to build

with different incentive structures.

The option shouldn't just be getting money from the DOD,

the military.

Right now I feel like the money that we get is trying

to figure out how to kill more people more efficiently,

basically automating the defense infrastructure.

And there's a lot of money going to AI

from military apparatus or how to make more money

for huge multinational corporations

that have so much power that you can't dissent in any way.

And so what I want is a different incentive structure

that says,

What is our incentive to do this research?

What is our goal?

It's not to make Google more money

and it's not to help the defense department figure out

how to kill more people more efficiently.

From day one, it's a different sort of goal.

So that's what I hope to do.