ensuring that these experiments are as efficient and safe as possible
Is āsafeā here meant in the sense of ānot accelerating risks from AI,ā or in the sense of ādifficult to stealā (i.e. secure)?
A bit of bothābut youāre right, I primarily meant āsecureā (as I expect this is where engineers have something specific to contribute).
Is āsafeā here meant in the sense of ānot accelerating risks from AI,ā or in the sense of ādifficult to stealā (i.e. secure)?
A bit of bothābut youāre right, I primarily meant āsecureā (as I expect this is where engineers have something specific to contribute).