Menu
Sign In Search Podcasts Charts People & Topics Add Podcast API Blog Pricing

David Duvenaud

๐Ÿ‘ค Speaker
1059 total appearances

Appearances Over Time

Podcast Appearances

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

machine companies or whatever is running the economy to employ humans.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

Like humans are pretty unreliable and kind of a pain to employ for a lot of reasons.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

And we don't want to hire, for instance, like 12 year olds, even if they're going to work for a dollar an hour.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

In fact, that's illegal.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

And that's one of the sort of structural forces that we also expect to be operating is that humans are just going to be this unreliable sort of scary thing to involve in anything important.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

And it'll be it'll be seen as like irresponsible to involve them in important decisions once machines are like better alternatives.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

Okay, yes.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

So exactly.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

So I agree with this.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

And again, I think this is something that people have a lot of issues with and they object and they say, but we won't let it get to that point.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

And like one thing we really hammer home is that competitive pressures are really going to force people into that situation.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

So even if everybody in the whole civilization sort of loves humans and would prefer to empower them, anyone doing anything important is really seen as this like irresponsible actor if they're like...

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

putting a human like surgeon in charge of like some surgery.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

It's like, what are you, it's just like, you know.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

It's risking someone's life.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

Right, it's like take your kid to work day and then you're like having them do the surgery.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

That's like how to be seen.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

Right.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

So at this point, there's basically no alpha in humans in the sense that it's not a good idea to invest in human capital in the way that we have like universities and these institutions that are like designed to sort of be human facing and human run.

80,000 Hours Podcast
Why 'Aligned AI' Could Still Kill Democracy | David Duvenaud, ex-Anthropic team lead

And so I think anyone who's like an investor, like the capital markets in general are going to be saying, why are we investing in this stream of human capital that's ultimately going to be less competitive and provide less returns than the sort of more machine centric, fully automated solution.