RSS

Stuart_Armstrong

Karma: 17,980

Con­sis­ten­cies as (meta-)preferences

Stuart_ArmstrongMay 3, 2021, 3:10 PM
17 points
0 comments3 min readLW link

Why un­rig­gable *al­most* im­plies uninfluenceable

Stuart_ArmstrongApr 9, 2021, 5:07 PM
11 points
0 comments4 min readLW link

A pos­si­ble prefer­ence algorithm

Stuart_ArmstrongApr 8, 2021, 6:25 PM
22 points
0 comments4 min readLW link

If you don’t de­sign for ex­trap­o­la­tion, you’ll ex­trap­o­late poorly—pos­si­bly fatally

Stuart_ArmstrongApr 8, 2021, 6:10 PM
17 points
0 comments4 min readLW link

Which coun­ter­fac­tu­als should an AI fol­low?

Stuart_ArmstrongApr 7, 2021, 4:47 PM
19 points
5 comments7 min readLW link

Toy model of prefer­ence, bias, and ex­tra information

Stuart_ArmstrongMar 24, 2021, 10:14 AM
9 points
0 comments4 min readLW link

Prefer­ences and bi­ases, the in­for­ma­tion argument

Stuart_ArmstrongMar 23, 2021, 12:44 PM
14 points
5 comments1 min readLW link

Why sig­moids are so hard to predict

Stuart_ArmstrongMar 18, 2021, 6:21 PM
56 points
7 comments5 min readLW link

Con­nect­ing the good reg­u­la­tor the­o­rem with se­man­tics and sym­bol grounding

Stuart_ArmstrongMar 4, 2021, 2:35 PM
13 points
0 comments2 min readLW link

Carte­sian frames as gen­er­al­ised models

Stuart_ArmstrongFeb 16, 2021, 4:09 PM
20 points
0 comments5 min readLW link

Gen­er­al­ised mod­els as a category

Stuart_ArmstrongFeb 16, 2021, 4:08 PM
25 points
9 comments4 min readLW link

Coun­ter­fac­tual con­trol incentives

Stuart_ArmstrongJan 21, 2021, 4:54 PM
21 points
10 comments9 min readLW link

Short sum­mary of mAIry’s room

Stuart_ArmstrongJan 18, 2021, 6:11 PM
26 points
2 comments4 min readLW link

Syn­tax, se­man­tics, and sym­bol ground­ing, simplified

Stuart_ArmstrongNov 23, 2020, 4:12 PM
30 points
4 comments9 min readLW link

The ethics of AI for the Rout­ledge En­cy­clo­pe­dia of Philosophy

Stuart_ArmstrongNov 18, 2020, 5:55 PM
45 points
8 comments1 min readLW link

Ex­tor­tion beats brinks­man­ship, but the au­di­ence matters

Stuart_ArmstrongNov 16, 2020, 9:13 PM
27 points
15 comments4 min readLW link

Hu­mans are stun­ningly ra­tio­nal and stun­ningly irrational

Stuart_ArmstrongOct 23, 2020, 2:13 PM
23 points
4 comments2 min readLW link

Knowl­edge, ma­nipu­la­tion, and free will

Stuart_ArmstrongOct 13, 2020, 5:47 PM
33 points
15 comments3 min readLW link

De­hu­man­i­sa­tion *er­rors*

Stuart_ArmstrongSep 23, 2020, 9:51 AM
13 points
0 comments1 min readLW link

An­thro­po­mor­phi­sa­tion vs value learn­ing: type 1 vs type 2 errors

Stuart_ArmstrongSep 22, 2020, 10:46 AM
16 points
10 comments1 min readLW link