HomeVideos

Gavin Baker: AI is too dangerous to centralize

Now Playing

Gavin Baker: AI is too dangerous to centralize

Transcript

34 segments

0:00

I think Anthropic and people in the

0:02

effective altruism movement believe that

0:04

this technology is too dangerous to

0:05

distribute.

0:07

And what Mark Zuckerberg and I think

0:09

Elon and Jensen believe

0:11

is that this technology is too dangerous

0:15

to centralize. And history has spoken.

0:18

And when given the choice, it is always

0:21

better to distribute and decentralize.

0:24

There's really no counter example that I

0:26

can think of.

0:28

And just as an American, but as a human,

0:32

I want an AI

0:34

that looks out for me.

0:36

Looks out for me.

0:37

>> The individual.

0:39

>> Be the individual. My

0:40

>> individual was a key point. Yes.

0:43

>> Yeah, my conception of what's good for

0:44

America, I do not want Dario Amodei or

0:48

Anthropic

0:49

deciding what's in my best interests. I

0:52

just fundamentally don't want that. I

0:53

don't

0:53

>> trust AI Jesus.

0:55

>> I don't I mean I I AI Jesus seems like

0:58

>> He seems like a He seems like a great

1:00

guy. You know, although it is funny,

1:01

somebody said he was cursed with resting

1:04

smiling face. And it is a little weird

1:07

to listen to him talk about this

1:09

terrible vision for humanity

1:11

while he's smiling happily.

Interactive Summary

The transcript discusses a fundamental debate regarding the future of AI development, contrasting the centralized, cautious approach favored by Anthropic and the effective altruism movement with the decentralized, open-source model advocated by figures like Mark Zuckerberg and Jensen Huang. The speaker argues for the decentralization of AI, expressing concern about large companies dictating the interests of individuals.

Suggested questions

2 ready-made prompts