flâneur

Against ubiquitous alignment taxes - LessWrong

lesswrong.com · 1,721 words · saved by 1 readers

It is often argued that any alignment technique that works primarily by constraining the capabilities of an AI system to be within some bounds cannot work because it imposes too high an 'alignment ta…

x Against ubiquitous alignment taxes — LessWrong AI Risk Alignment Tax AI Frontpage 59 Against ubiquitous alignment taxes by beren 6th Mar 2023 2 min read 10 59 Crossposted from my personal blog . It is often argued that any alignment technique that works primarily by constraining the capabilities of an AI system to be within some bounds cannot work because it imposes too high an 'alignment tax' on the ML system. The argument is that people will either refuse to apply any method that has an alignment tax, or else they will be outcompeted by those who do. I think that this argument is applied t

saved by

related reading