flâneur

Universal Inductors — AI Alignment Forum

alignmentforum.org · 683 words · saved by 1 readers

Now that the Logical Induction paper is out, I am directing my attention towards decision theory. The approach I currently think will be most fruitful is attempting to make a logically updateless version of Wei Dai's Updateless Decision Theory. Abram Demski has posted on here about this, but I think Logical Induction provides a new angle with which we can attack the problem. This post will present an alternate way of viewing Logical Induction which I think will be especially helpful for building a logical UDT. (The Logical Induction paper is a prerequisite for this post.) This post will make several informal claims without proof. Many of the claims are analogous to things proven in the Logical Induction paper, and the math is not substantially difficult. I might extend the paper to have a section talking about the following concepts more formally. A Universal Inductor is an infinite sequence { P n } of probability distributions on infinite bit strings. Satisfying the following two pr

x Universal Inductors — AI Alignment Forum Personal Blog 6 Universal Inductors by Scott Garrabrant 14th Sep 2016 2 min read 0 6 Now that the Logical Induction paper is out, I am directing my attention towards decision theory. The approach I currently think will be most fruitful is attempting to make a logically updateless version of Wei Dai's Updateless Decision Theory . Abram Demski has posted on here about this, but I think Logical Induction provides a new angle with which we can attack the problem. This post will present an alternate way of viewing Logical Induction which I think will be es

related reading