"We're Not Ready: thoughts on "pausing" and responsible scaling policies" by Holden Karnofsky
LessWrong (Curated & Popular)
00:00
Exploring the Potential of Pausing AI Scaling and its Consequences
This chapter explores the risks involved in AI scaling and discusses the potential scenarios and outcomes of different types of pauses, such as regulation-mandated pauses and temporary bans on large training runs. It also raises concerns about the consequences of no regulation or poorly designed regulation and discusses the potential positive impact of a US-legislated moratorium.
Play episode from 04:21
Transcript
