RSS

JustinShovelain

Karma: 594

I am the co founder of and researcher at the quantitative long term strategy organization Convergence (see here for our growing list of publications). Over the last sixteen years I have worked with MIRI, CFAR, EA Global, and Founders Fund, and done work in EA strategy, fundraising, networking, teaching, cognitive enhancement, and AI safety research. I have a MS degree in computer science and BS degrees in computer science, mathematics, and physics.

Good­hart Ty­pol­ogy via Struc­ture, Func­tion, and Ran­dom­ness Distributions

Mar 25, 2025, 4:01 PM
32 points
0 comments15 min readLW link

Bounded AI might be viable

Mar 6, 2025, 12:55 PM
17 points
0 comments20 min readLW link

In­for­ma­tion-The­o­retic Box­ing of Superintelligences

Nov 30, 2023, 2:31 PM
30 points
0 comments7 min readLW link

The risk-re­ward trade­off of in­ter­pretabil­ity research

Jul 5, 2023, 5:05 PM
15 points
1 comment6 min readLW link

Align­ing AI by op­ti­miz­ing for “wis­dom”

Jun 27, 2023, 3:20 PM
27 points
8 comments12 min readLW link

Im­prov­ing the safety of AI evals

May 17, 2023, 10:24 PM
13 points
7 comments7 min readLW link

Keep hu­mans in the loop

Apr 19, 2023, 3:34 PM
23 points
1 comment10 min readLW link

Up­dat­ing Utility Functions

May 9, 2022, 9:44 AM
41 points
6 comments8 min readLW link

Good­hart’s Law Causal Diagrams

Apr 11, 2022, 1:52 PM
34 points
6 comments6 min readLW link

How Money Fails to Track Value

JustinShovelainApr 2, 2022, 12:32 PM
17 points
0 comments5 min readLW link

Eval­u­at­ing ex­per­tise: a clear box model

JustinShovelainOct 15, 2020, 2:18 PM
36 points
3 comments5 min readLW link

Good and bad ways to think about down­side risks

Jun 11, 2020, 1:38 AM
19 points
12 comments11 min readLW link

COVID-19: An op­por­tu­nity to help by mod­el­ling test­ing and trac­ing to in­form the UK government

JustinShovelainApr 17, 2020, 5:21 PM
14 points
2 comments2 min readLW link

[Question] Test­ing and con­tact trac­ing im­pact as­sess­ment model?

JustinShovelainApr 9, 2020, 5:42 PM
6 points
3 comments1 min readLW link

COVID-19: List of ideas to re­duce the di­rect harm from the virus, with an em­pha­sis on un­usual ideas

JustinShovelainApr 9, 2020, 11:33 AM
30 points
12 comments7 min readLW link

Memetic down­side risks: How ideas can evolve and cause harm

Feb 25, 2020, 7:47 PM
27 points
3 comments15 min readLW link

In­for­ma­tion haz­ards: Why you should care and what you can do

Feb 23, 2020, 8:47 PM
18 points
4 comments15 min readLW link

Map­ping down­side risks and in­for­ma­tion hazards

Feb 20, 2020, 2:46 PM
23 points
0 comments9 min readLW link

Us­ing vec­tor fields to vi­su­al­ise prefer­ences and make them consistent

Jan 28, 2020, 7:44 PM
42 points
32 comments11 min readLW link

AI al­ign­ment con­cepts: philo­soph­i­cal break­ers, stop­pers, and distorters

JustinShovelainJan 24, 2020, 7:23 PM
20 points
3 comments3 min readLW link