AI ALIGNMENT FORUM
AF

2022 MIRI Alignment Discussion

Jun 15, 2022 by Rob Bensinger

A collection of MIRI write-ups and conversations about alignment released in 2022, following the Late 2021 MIRI Conversations.

77Six Dimensions of Operational Adequacy in AGI Projects

Eliezer Yudkowsky

3y

33

147AGI Ruin: A List of Lethalities

Eliezer Yudkowsky

3y

144

74A central AI alignment problem: capabilities generalization, and the sharp left turn

3y

18

98On how various plans miss the hard bits of the alignment challenge

3y

48

47The inordinately slow spread of good AGI conversations in ML

3y

0

67A note about differential technological development

3y

11

54Brainstorm of things that could force an AI team to burn their lead

3y

5

64AGI ruin scenarios are likely (and disjunctive)

3y

3

34Where I currently disagree with Ryan Greenblatt’s version of the ELK approach

3y

7

50Why all the fuss about recursive self-improvement?

3y

0

25Humans aren't fitness maximizers

3y

4

56Warning Shots Probably Wouldn't Change The Picture Much

3y

11

38What does it mean for an AGI to be 'safe'?

3y

5

42Don't leave your fingerprints on the future

3y

10

47Niceness is unnatural

3y

4

47Contra shard theory, in the context of the diamond maximizer problem

3y

3

36Notes on "Can you control the past"

3y

1

45Decision theory does not imply that we get to have nice things

3y

36

46Superintelligent AI is necessary for an amazing future, but far from sufficient

3y

3

52How could we know that an AGI system will have good consequences?

3y

10

26Distinguishing test from training

3y

3

78A challenge for AGI organizations, and a challenge for readers

Rob Bensinger, Eliezer Yudkowsky

3y

17

49Thoughts on AGI organizations and capabilities work

Rob Bensinger, So8res

3y

8