Public evidence record

Shane Legg

Published podcast speaker

Claims: 13
Episodes: 1
Shows: 1
Named items: 0

Claim ledger

What Shane said.

3 transcript-backed records

01 / preference

26 Oct 2023

We need to make sure it understands humans ethics well, at least as well as a very good ethicist because that's important.

“We need to make sure it understands humans ethics well, at least as well as a very good ethicist because that's important.”

Speaker: Shane Legg
Episode: Shane Legg (DeepMind Founder) — 2028 AGI, superhuman alignment, new architectures
Publisher: Dwarkesh Podcast

02 / preference

26 Oct 2023

If you want a model that writes creative poetry, then that's fine because you want to be able to be very free to suggest all kinds of possibilities and so on.

“If you want a model that writes creative poetry, then that's fine because you want to be able to be very free to suggest all kinds of possibilities and so on.”

Speaker: Shane Legg
Episode: Shane Legg (DeepMind Founder) — 2028 AGI, superhuman alignment, new architectures
Publisher: Dwarkesh Podcast

03 / preference

26 Oct 2023

The way I think about it is this: to have a profoundly ethical AI system, it also has to be very, very capable.

“The way I think about it is this: to have a profoundly ethical AI system, it also has to be very, very capable.”

Speaker: Shane Legg
Episode: Shane Legg (DeepMind Founder) — 2028 AGI, superhuman alignment, new architectures
Publisher: Dwarkesh Podcast

Search evidence