Iterated Amplification

Oct 29, 2018

by paulfchristiano

This is an upcoming sequence curated by Paul Christiano on one current approach to alignment: Iterated Amplification. The posts will be released in the coming weeks, on average every 2-3 days.

Preface to the sequence on iterated amplification

395moΩ 12Show Highlight
0

Problem statement

The first part of this sequence clarifies the problem that iterated amplification is trying to solve, which is both narrower and broader than you might expect.

The Steering Problem

375moΩ 9Show Highlight
4

Clarifying "AI Alignment"

545moΩ 15Show Highlight
47

An unaligned benchmark

275moΩ 8Show Highlight
0

Prosaic AI alignment

365moΩ 10Show Highlight
0

Basic intuition

The second part of the sequence outlines the basic intuitions that motivate iterated amplification. I think that these intuitions may be more important than the scheme itself, but they are considerably more informal.

Approval-directed agents

225moΩ 3Show Highlight
8

Approval-directed bootstrapping

195moΩ 4Show Highlight
0

Humans Consulting HCH

195moΩ 4Show Highlight
8

Corrigibility

395moΩ 9Show Highlight
3

The scheme

The core of the sequence is the third section. Benign model-free RL describes iterated amplification, as a general outline into which we can substitute arbitrary algorithms for reward learning, amplification, and robustness. The first four posts all describe variants of this idea from different perspectives, and if you find that one of those descriptions is clearest for you then I recommend focusing on that one and skimming the others.

Iterated Distillation and Amplification

205moΩ 5Show Highlight
6

Benign model-free RL

105moΩ 3Show Highlight
0

Factored Cognition

345moΩ 10Show Highlight
3
0

AlphaGo Zero and capability amplification

253moΩ 6Show Highlight
23

What needs doing

The fourth part of the sequence describes some of the black boxes in iterated amplification and discusses what we would need to do to fill in those boxes. I think these are some of the most important open questions in AI alignment.

Directions and desiderata for AI alignment

293moΩ 10Show Highlight
1

The reward engineering problem

183moΩ 3Show Highlight
3

Capability amplification

243moΩ 7Show Highlight
4

Learning with catastrophes

263moΩ 7Show Highlight
4

Possible approaches

The fifth section of the sequence breaks down some of these problems further and describes some possible approaches.

Thoughts on reward engineering

243moΩ 5Show Highlight
23

Techniques for optimizing worst-case performance

233moΩ 7Show Highlight
3

Reliability amplification

213moΩ 4Show Highlight
3